With the global emphasis on environmental protection and sustainable energy, energy conservation and emissions reduction have become critical issues of the present era. Against this backdrop, new energy vehicles, with electric vehicles as a prime example, have experienced rapid development [1, 2]. In the electric vehicle industry, rapid development is accompanied by intense market competition. Driven by the market economy, electric vehicles must not only extend their cruising range and enable rapid charging, but also advance the development of enhanced riding comfort and intelligent functionality. The former necessitates a large-capacity battery pack and ongoing technological advancements, while the latter may result in the addition of more electronic control units within the vehicle. Both factors can contribute to an increase in the vehicle's overall weight, which may affect its cruising range in turn. Therefore, it is necessary to reduce the weight of other components in electric vehicles. In the electrical systems of electric vehicles, a significant amount of conductive metal is required [3, 4]. These power-transmitting components, like the vehicle body, are regarded as critical components with substantial potential for weight reduction. Aluminum and copper are two widely utilized conductive materials. Copper exhibits superior electrical conductivity; however, it is characterized by a higher density and greater cost. In contrast, aluminum, while offering lower electrical conductivity compared to copper, provides advantages in terms of lighter weight and reduced cost. In the electrical systems of electric vehicles, substituting aluminum for copper can contribute to weight reduction and simultaneously lower manufacturing costs. However, in certain relatively confined spaces within electric vehicles, the cross-sectional dimensions of conductors are restricted, necessitating the use of copper as the conductor material to ensure adequate electrical conductivity [5, 6]. In this manner, the cross-application of aluminum and copper in the electrical systems of electric vehicles inevitably involves the joining of these two materials.

Due to the poor metallurgical compatibility between aluminum and copper, the application of fusion welding techniques—such as electron beam welding [7] and laser welding [8–10]—to join these materials results in the formation of a significant amount of hard and brittle intermetallic compounds (IMCs) within the weld seam, which adversely affects the joint's mechanical properties. In light of this, certain solid-state welding techniques characterized by relatively low welding temperatures—such as friction stir welding [11–12], electromagnetic pulse welding [13], and ultrasonic welding [14–16]—have been extensively investigated for the joining of aluminum and copper. Although the application of solid-state welding technology in joining aluminum and copper can partially inhibit the growth of IMCs at the interface, these techniques are subject to certain application-specific constraints and limitations. In the electrical systems of electric vehicles, the connection between copper busbars and aluminum busbars in the form of thin plate lap joints is more suitable for resistance spot welding (RSW) [17].

RSW involves the use of Joule heat generated by electric current passing through the workpieces to locally melt the base metal and create a joint. However, two major technical challenges arise when applying this process to weld aluminum and copper. Firstly, the IMCs layer formed at the interface influences the properties and reliability of the joint. In most cases of welding dissimilar metals, the formation of IMCs at the weld interface is both inevitable and essential. This is because metallurgical bonding can only be achieved through the formation of such IMCs at the interface [18, 19]. Secondly, the superior electrical and thermal conductivity of aluminum and copper—particularly copper—hinders the generation of sufficient heat during the welding process, making it challenging to achieve a sound joint. For the former, relevant research has primarily concentrated on controlling the interfacial metallurgical reaction by introducing an interlayer at the interface. This approach aims to either form IMCs with reduced brittleness or restrict the mutual diffusion of atoms, resulting in a thinner IMC layer. Consequently, the overall performance of the joint can be enhanced. For example, Xiao et al. performed RSW on 1 mm thick T2 purple copper and 2A16 aluminum alloy plates using a 0.5 mm thick FeCrNiAlCu high-entropy alloy powder interlayer [20]. The results indicated the formation of an Al-coated high-entropy alloy particle layer at the interface, along with an Al₂Cu layer on the copper side adjacent to the interface [20]. Through research on the resistance spot welded joint between Ni-P coated aluminum plate and copper plate, it was found that the interface bonding morphology of the joint was significantly improved, and the occurrence of cracks and porosity within the weld was effectively suppressed [21]. For the latter, electrodes fabricated from materials with high resistivity are primarily utilized to generate additional heat during the welding process, thereby facilitating the formation of a larger nugget. Chen et al. employed W-Cu alloy electrodes to weld two copper plates with a Ni-P-coated Al transition layer by using RSW, and found that metallurgical bonding was achieved in the welded joints under three kinds welding schedules with varying heat inputs [22]. Employing electrodes with increased resistivity in RSW can enhance heat generation within the weld zone [23]. However, due to the relatively low melting point of aluminum, electrode sticking tends to occur on the aluminum side when high-resistance electrodes are used on both sides during RSW of aluminum and copper. In the preliminary research, direct RSW of aluminum and copper was performed utilizing a tungsten electrode on the copper side and a CrZrCu alloy electrode on the aluminum side [24]. The results indicated that the nugget size was relatively small, which led to reduced joint performance [24].

In this study, RSW of aluminum and copper was performed using zinc foil as an interlayer, with the aim of achieving a larger nugget. The low melting point of zinc and its capacity to undergo a eutectic reaction with aluminum at relatively low temperature were leveraged to promote aluminum liquefaction during the welding process. The microstructure and mechanical properties of the resulting joints were systematically investigated to provide a foundation for subsequent research.

The RSW experiment utilized two thin metal plates: a 1 mm thick T2 copper plate and a 2 mm thick A1060 aluminum plate, both of which were machined into dimensions of 100 mm × 30 mm. The respective chemical compositions of these materials are presented in Table 1 and Table 2. Additionally, a 0.1 mm thick pure zinc foil was employed as an interlayer and machined into square specimens measuring 30 mm × 30 mm.

Before welding, the surfaces of the T2 copper plate and A1060 aluminum plate were ground with sandpaper to remove the surface oxide film, and then cleaned with anhydrous ethanol and air-dried. The surfaces of the Zn foil were wiped with alcohol and dried. The components were cleaned and subsequently assembled in the longitudinal direction, with a lap length of 30 mm. A zinc interlayer was incorporated between the two plates. Welding tests were carried out using the DM-200 inverter-based medium-frequency DC RSW machine. An electrode made of CrZrCu alloy with a tip diameter of 10 mm was used on the aluminum side, whereas a flat tungsten electrode with a tip diameter of 11 mm was employed on the copper side. During welding, with the electrode pressure maintained constant at 3 kN, the welding current and welding time were systematically varied. When the welding time was kept constant at 400 ms, RSW was performed at 1 kA intervals across a current range of 25 kA to 29 kA by varying the welding current. When the welding current was maintained at 28 kA, RSW was carried out at intervals of 25 ms within the range of 325 ms to 425 ms by varying the welding time. Seven specimens were welded under each set of welding parameters. Of these, two were prepared for microstructural observation, while the remaining five were subjected to tensile-shear testing.

After welding, the resistance test of the resistance spot welded joint between the A1060 aluminum and T2 copper with a zinc interlayer (henceforth called RSWed Al/Cu joint) was conducted using a resistance meter (RM3545A) at room temperature. During the testing process, the test probe was positioned at both ends along the longitudinal direction of the RSWed Al/Cu joint for measurement. Each joint was measured three times, with slight adjustments made to the probe position each time. The average of the three measurements was then calculated and recorded. For comparison, test sheets of A1060 aluminum and T2 copper were prepared with dimensions of 170 mm × 30 mm × 2 mm and 170 mm × 30 mm × 1 mm, respectively. The length of the test sheets was set to 170 mm, which was determined based on the length of the RSWed Al/Cu joint (100 mm + 100 mm − 30 mm = 170 mm). The resistance testing methods applied to the test sheets were identical to those used for the RSWed Al/Cu joint.

At room temperature, a precision universal material testing machine was employed to perform a tensile-shear test on the RSWed Al/Cu joint at a loading rate of 1 mm/min. On the other hand, the RSWed Al/Cu joint was sectioned along its weld diameter, perpendicular to the welding interface, and subsequently underwent grinding and polishing of the cross-sectional surface. The optical microscope (IE500M), scanning electron microscope (SEM) equipped with an energy-dispersive spectrometer (EDS) and an electron backscatter diffraction (EBSD) system were employed to examine and analyze the characteristic regions of the RSWed Al/Cu joint.

Figure 1(a) presents a representative cross-sectional image of the RSWded Al/Cu joint. This joint was fabricated under the conditions of a welding current of 28 kA and a welding time of 400 ms. A nugget was observed on the aluminum side of the joint, but no nugget was found in the copper adjacent to the interface. This phenomenon was attributed to the inherent physical properties of copper, including its excellent electrical and thermal conductivity, as well as its high melting point. From this, it can be seen that the essence of the RSWed Al/Cu joint formation lies in the spreading and adhesion of molten aluminum onto the surface of solid copper, thereby achieving a welding-brazing effect.

As illustrated in Fig. 1(a), the nugget diameter was observed to be larger in proximity to the interface and progressively smaller with increasing distance from the interface, forming an overall disc-shaped cross-section of the nugget. The geometric configuration of the nugget cross-section was primarily influenced by the heat distribution within the welding zone. In terms of heat generation, more Joule heat was produced during the welding process due to the presence of contact resistance at the interface. With respect to heat dissipation, the primary pathway was through the water-cooled copper electrode. Additionally, heat was conducted along both the length and width of the aluminum plate, benefiting from its high thermal conductivity. Although the copper on the opposite side of the joint also exhibits excellent thermal conductivity, the use of a tungsten electrode with relatively high electrical resistance resulted in the welding zone on the copper side not only retaining heat during the welding process, but also receiving additional thermal input due to the increased heat generated at the electrode. Although the temperature on the copper side did not reach the melting point, it ensured that the nugget formed on the aluminum side adjacent to the interface had a relatively large diameter.

As shown in Fig. 1(a), no interlayer was observed in the interfacial zone between the nugget and the copper of the joint. In particular, no residual zinc foil was observed in the interface zone outside the nugget (referred to as the L zone in Fig. 1(a)). This suggests that the zinc interlayer melted during the welding process. Furthermore, electrode indentation was observed on the aluminum side of the joint. The nugget diameter, nugget thickness, and indentation depth of the joints were measured as illustrated in Fig. 1(b). The measured results are shown in Fig. 1(c) and 1(d).

As illustrated in Fig. 1(c) and 1(d), an increase in welding current or an extension of welding time resulted in a corresponding increase in the nugget diameter of the RSWed Al/Cu joints. According to Joule's law, increasing the welding current or extending the welding time can generate more heat, which in turn increased the amount of molten metal and promoted the formation of a larger weld nugget. Within the welding current range of 25 kA to 28 kA, the rate of increase in nugget diameter was relatively high, whereas it became lower when the welding current exceeded 28 kA, as illustrated in Fig. 1(c). A similar trend can be observed in Fig. 1(d), where the nugget diameter increased more rapidly within the welding time range of 325 ms to 375 ms, while the growth rate slowed when the welding time exceeded 375 ms. The surface area of the nugget (i.e., the interface between the molten and solid metal during welding) increased with the nugget diameter, which implies an expanded heat dissipation path. Consequently, as the nugget grew beyond a certain size, its growth rate began to decrease due to the enhanced heat dissipation effect. For the RSWed joint of aluminum alloy, the nugget diameter (D) should satisfy the C-level requirement, which specifies that D ≥ 3.5H^1/2, where H represents the thickness of the aluminum alloy plate [25]. In this study, the nugget diameters of the Al/Cu joints welded within the selected welding parameter range satisfied the required specifications. Compared to the nugget diameter of the direct RSWed joint reported in reference [24], the nugget diameter of the RSWed joint with a zinc interlayer was significantly larger. For example, the nugget diameter of the joint with a zinc interlayer was approximately 7.5 mm (as shown in Fig. 1(a)), which is about 1.44 times greater than that of the direct RSWed joint, where the nugget diameter was 5.2 mm under the same welding current of 28 kA [24]. This suggests that the utilization of zinc foil as an interlayer in the RSW of aluminum and copper has effectively facilitated the growth of the weld nugget. The melting point of zinc is approximately 419°C, and the eutectic transformation temperature of the Zn-Al system is about 382°C. Both values are lower than the melting point of aluminum (approximately 660°C) and the eutectic transformation temperature of the Al-Cu system (approximately 548°C) [26]. During resistance spot welding of aluminum and copper with a zinc interlayer, a eutectic transformation occurred between aluminum and zinc in the interface zone upon heating to 382°C, resulting in the formation of a liquid phase. As the heating process continued, the amount of the formed liquid phase gradually increased. Upon reaching temperatures above the melting point of aluminium, it also melted, resulting in the formation of a mixed liquid phase. Concurrently with the liquefaction of zinc and aluminium, copper began to dissolve and diffuse into the mixed liquid phase. Following subsequent cooling and solidification, the nugget was formed. Due to the relatively low temperature at which metals began to liquefy in the interfacial zone, a larger amount of molten metal was produced during the welding process, leading to an increase in the final nugget size.

Similar to the nugget diameter, the nugget thickness (h) increased with higher welding current or extended welding time, as illustrated in Fig. 1(c) and 1(d). For RSW, the penetration rate λ (defined as λ = h/H, where h denotes the nugget thickness and H denotes the plate thickness) should exceed 20% [25]. The aluminum plate utilized in this study has a thickness of 2 mm. Accordingly, the nugget thickness must attain a minimum of 0.4 mm in order to comply with the specified standards. As shown in Fig. 1(c), when the welding time was set to 400 ms, the penetration rate of the joints welded using a welding current greater than 28 kA met the standard requirements. Additionally, when the welding current was maintained at 28 kA, the penetration rate satisfied the standard requirements for welding times of 375 ms or longer, as shown in Fig. 1(d).

As shown in Fig. 1(c) and (d), the depth of electrode indentation remaining on the aluminum side of the RSWed Al/Cu joint increased with either a higher welding current or an extended welding time. Nevertheless, under the welding parameters selected in this study, the electrode indentation depth does not exceed 0.4 mm, which corresponds to 20% of the thickness of the aluminum plate. During RSW, the welding zone experiences plastic deformation due to the application of electrode pressure, leading to the formation of indentations on the joint surfaces. As the welding current increased and the welding duration extended, the quantity of molten metal also increased, leading to a reduced amount of solid metal in the electrode-clamped zone during the welding process and a diminished capacity to resist plastic deformation. In addition, no significant electrode indentation was observed on the copper side of the RSWed Al/Cu joint. This phenomenon is attributed to the larger diameter of the electrode tip (11 mm), as well as the fact that the copper side has not attained a high-temperature plastic state.

Figure 2(a), 2(b), and 2(c) present the SEM images of the interface zone of the RSWed Al/Cu joint, which correspond to locations A, B, and C marked in Fig. 1(a), respectively. Shallow gray precipitates were observed in the nugget zone of the joint. Figure 2(d) presents the EDS results obtained along the MN line depicted in Fig. 2(a). The EDS results show that the content distribution of Al and Cu in the nugget zone has some fluctuations. Where the Al content was slightly lower, the Cu content was slightly higher, and vice versa. This may be attributed to the influence of precipitates, which also suggests that the precipitates present in the nugget zone are of the Al-Cu phase type. At the interface zone between the nugget and the copper base material, the formation of a reaction layer was observed as shown in Fig. 2(a) and 2(b). The EDS results presented in Fig. 2(d) demonstrate that the Al content in the reaction layer decreased, whereas the Cu content increased, in comparison to the nugget zone. Based on the morphological characteristics and contrast of the reaction layer, it can be distinguished into two distinct layers: the U₁ layer, which is adjacent to the copper side, and the U₂ layer, which is closer to the nugget. The thickness of the U₁ layer was significantly greater than that of the U₂ layer. Specifically, at the interface located in the central region of the weld (Fig. 2(a)), the thickness of the U₁ layer was approximately 100 µm, whereas that of the U₂ layer was approximately 24 µm. In comparison, at the interface within the peripheral zone of the weld, the U₁ layer was thinner, while the U₂ layer exhibited a slightly increased thickness. More precisely, the thickness of the U₁ layer was approximately 43 µm, and that of the U₂ layer was approximately 30 µm, as illustrated in Fig. 2(b).

In addition, as shown in Fig. 2(c), the reaction layer was also observed to form at the interface outside the nugget. The reaction layer in this context also consisted of two sublayers, U₁ and U₂. It is evident that the U₂ layer is slightly thicker than the U₁ layer. Specifically, the average thickness of the U₂ layer is approximately 25 µm, whereas that of the U₁ layer is approximately 17 µm. In the interface region of section L shown in Fig. 1(a), neither the formation of a nugget on the aluminum side nor the presence of a residual zinc interlayer was observed; however, a reaction layer predominantly composed of Al-Cu-type phases was still formed. This suggests that the temperature in the region exceeded the eutectic transformation temperature of aluminum and zinc (382°C), yet remained below the melting point of aluminum (660°C) during the welding heating process. During the eutectic transformation of aluminum and zinc in the interfacial region, where a liquid phase was formed, copper also dissolved and diffused into the liquid phase. Upon subsequent cooling, a reaction layer predominantly composed of the Al-Cu phase was formed.

Figure 3 presents the magnified SEM images of selected regions on the cross-section of the RSWed Al/Cu joint. EDS analysis was performed on several characteristic zones, and the results are summarized in Table 3. Figure 3(a) presents an enlarged view of location D as indicated in Fig. 2(a). The nugget zone was predominantly characterized by a dark gray matrix phase and a light gray precipitated phase, clearly distinguishing it from the adjacent aluminum base material. Figure 3(b) presents high-magnification SEM image of a localized region within the nugget. The light gray precipitate phases exhibited a net-like distribution pattern. The EDS results indicate that the precipitated phase (at location B₁) contained a higher concentration of Cu compared to the matrix phase (at location A₁). A trace amount of Zn was identified in both the matrix phase and the precipitate structures. Figure 3(c) presents an enlarged view of location E as indicated in Fig. 2(a). It can be observed that the interface between the U₁ and U₂ layers was not smooth, but instead exhibited a jagged, irregular pattern. Figure 3(d) presents a high-magnification SEM image of a localized region within the U₂ layer, revealing the presence of black phase interspersed among the gray columnar structures. This observation suggests that the U₂ layer comprised at least two distinct phases. The EDS analysis results from location C₁ indicate that the U₁ layer was predominantly composed of the Al₂Cu phase. Al, Cu, and a small amount of Zn were detected in the gray columnar structure within the U₂ layer (at location D₁ in Fig. 3(d)). The Cu content significantly exceeded its solid solubility limit in Al, indicating that the U₂ layer primarily consisted of the Al₂Cu phase and the Cu solid solution in Al (referred to as α-Al).

Additionally, as illustrated in Fig. 3(c), certain blocky reaction products were observed within the U₁ layer adjacent to the copper side. Figure 3(e) presents an enlarged view of location F as indicated in Fig. 3(c). It can be observed that a relatively thin reaction layer, referred to as the U₃ layer, was formed between the U₁ layer and copper. Some of the blocky reaction products within the U₁ layer were connected to the U₃ layer, whereas others remained isolated and exhibited an island-like distribution within the U₁ layer. It can be observed from the contrast that these blocky reactants and the U₃ layer were composed of a Cu-rich Al-Cu phase. Observation results indicate that the U₃ layer formed in the central region of the weld was relatively thick, and a greater quantity of blocky reaction products was present at the interface in that area. In contrast, the U₃ layer formed at the interface in the peripheral zone of the weld was very thin, and fewer blocky reaction products were observed. Moreover, no such reaction products were formed at the interface in the L segment outside the nugget region as shown in Fig. 2(c).

Figure 4 displays the EBSD detection results at the boundary of the nugget (indicated as location P in Fig. 2(a)). Figure 4(a), 4(b), and 4(c) illustrate the distributions of Al, Cu, and Zn within this region, respectively. Al was mainly detected in the base material zone, while Al, Cu and Zn were detected in the nugget zone. This also confirms that the copper dissolved and diffused into the liquid phase, while the zinc liquefied and mixed with the aluminum melt during the welding process, both contributing to the formation of the nugget. Figure 4(d) illustrates the phase distribution within the region. The nugget zone was predominantly composed of two phases: α-Al and Al₂Cu, whereas the base material outside the zone consisted solely of the α-Al phase. It can therefore be concluded that the precipitates of the Al-Cu type phase within the nugget mentioned in the preceding text are composed of Al₂Cu. However, the amount of Al₂Cu precipitated in the nugget zone was relatively small, accounting for approximately 7.8%. The majority of the precipitated Al₂Cu was distributed in a particulate form along the grain boundaries of α-Al. These diffusely distributed IMC particles are believed to contribute positively to the enhancement of the mechanical properties of the nugget zone [27]. Figure 4(d) presents the grain orientation distribution within the region. Compared to the grain size of the base material zone (approximately 10.89 µm), the α-Al grains in the nugget zone exhibited a larger average diameter of approximately 18.01 µm. In contrast, the Al₂Cu grains precipitated within the nugget zone were significantly finer, with an average diameter of about 3.90 µm. From the perspective of grain morphology, the α-Al grains in the peripheral region of the nugget exhibited a columnar structure, with a width (W₁) of approximately 110 µm. The grains in other regions of the nugget exhibited an equiaxed morphology. As shown in Fig. 4(e), the grains within the nugget exhibited a variety of colors, indicating a random distribution of grain orientations. Figure 4(f) presents the kernel average misorientation (KAM) map of the region, which reflects the relatively high dislocation density in the base material zone. This can be attributed to the fact that the aluminum base material sheet was manufactured through a rolling process. In contrast, the nugget zone was formed via remelting followed by solidification, leading to a noticeable reduction in dislocation density within this zone.

During the cooling and solidification process, the liquid metal adjacent to the solid aluminum base material solidified first, due to the high thermal conductivity of aluminum and the electrode employed on the aluminum side. Under these conditions, non-spontaneous nucleation occurred at the solid wall external to the molten nugget. During grain growth, certain grains exhibited accelerated growth in the direction opposite to the heat dissipation of the liquid melt, ultimately forming columnar crystals. As the solidification and crystallization process progressed toward the interior of the liquid phase region, an increase in nucleation particles led to greater grain formation, while heat dissipation became more isotropic. These factors contributed to the development of equiaxed crystals in the central region of the nugget. During grain growth, solute atoms segregated to the grain boundaries. Due to the relatively high solubility of zinc in aluminum (e.g., approximately 16.0 at.% at 275°C), which varies gradually with temperature, and the comparatively low solubility of copper in aluminum (approximately 2.5 at.% at 548°C), which decreases sharply with decreasing temperature [26], Cu atoms were segregated to the grain boundaries, whereas Zn atoms dissolved into the matrix and contributed to the formation of α-Al grains during grain growth. As the temperature at the grain boundary decreased, a eutectic transformation took place, leading to the formation of Al₂Cu and α-Al phases.

Figure 5(a) presents an image of the welding interface zone, captured at location Q as indicated in Fig. 2(a). Figure 5(b), 5(c), and 5(d) illustrate the distributions of Al, Cu, and Zn within this zone, respectively. Compared to the U₁ layer, the Cu content detected in the U₂ layer was slightly reduced. Additionally, the zinc content measured on the copper side adjacent to the interface was found to be higher than that detected within the nugget region. During the welding process, the copper adjacent to the interface remained in a solid state. The diffusion rate of zinc atoms into copper was lower than that into liquid aluminum. However, due to the relatively high solid solubility of zinc in copper, the diffused zinc atoms primarily accumulated on the copper side near the interface. On the other hand, although the majority of the zinc interlayer was liquefied and mixed with the liquid aluminum, the liquid phase within the welding zone underwent thorough convection and mixing under the influence of electromagnetic forces and other factors. As a result, zinc was distributed throughout the nugget, leading to a relatively low concentration of Zn in that region.

Figure 5(e) illustrates the phase distribution within the region. The nugget zone was also predominantly composed of two phases: α-Al and Al₂Cu, whereas the copper base material consisted solely of the α-Cu phase. Although the precipitated Al₂Cu in the nugget zone near the interfacial reaction layer was predominantly located at the α-Al grain boundaries, its volume fraction (approximately 8.9%) was slightly lower compared to that in the nugget zone farther from the interfacial reaction layer. As illustrated in Fig. 5(e), the U₁ layer was predominantly composed of the Al₂Cu phase, whereas the U₂ layer primarily consisted of both the Al₂Cu and α-Al phases. Within the U₂ layer, the volume fraction of the Al₂Cu phase was found to be approximately 63.2%. Figure 5(f) presents the grain orientation distribution within the interfacial region. Compared to the coarse Al₂Cu grains observed in the U₁ layer, the Al₂Cu grains in the U₂ layer exhibited a finer morphology, with an average diameter of approximately 5.96 µm. The grains within the nugget zone and copper base metal zone exhibit a wide spectrum of colors, suggesting a random grain orientation. The grain orientation distribution of Al₂Cu in the U₁ layer is predominantly characterized by the color red, which indicates a prevalence of < 001>-oriented grains. However, the statistical significance remains relatively limited due to the coarse grain structure and the small sample size of detected grains

During the welding heating process, the concentration of copper in the liquid metal near the copper side was relatively high. Upon cooling, solidification of the liquid metal initiated from the region adjacent to the copper base material where the copper's higher thermal conductivity facilitated faster heat dissipation, at the same time as solidification began from the area adjacent to the solid aluminum base metal. When cooled to approximately 591°C, the Al₂Cu phase precipitated initially in the copper-rich liquid phase region adjacent to the copper-based substrate. As the Al₂Cu grains grew, the aluminum in the liquid phase was progressively enriched at the solidification front. As solidification proceeded, the concentration of aluminum in the liquid phase at the solidification front gradually increased. Upon cooling to approximately 548°C, an eutectic transformation occurred, resulting in the formation of a eutectic structure composed of Al₂Cu and α-Al. In other words, the Al₂Cu phase precipitated first, forming the U₁ layer, whereas the eutectic structure composed of Al₂Cu and α-Al formed the U₂ layer. Here, it is believed that a small amount of zinc has been dissolved in them.

Figure 6(a) presents a high-magnification image of the interface zone between the U₁ layer and the copper base metal. Figure 6(b) illustrates the phase distribution within the region. The blocky reaction products precipitated within the U₁ layer were of the Al₄Cu₉ phase. It can be observed from the magnified view in Fig. 6(b) that the U3 layer, formed between the U1 layer and the copper substrate, consisted of the Al₄Cu₉ phase as well. In other words, both the blocky reactants and the U₃ layer formed on the nugget side near the welding interface, as illustrated in Fig. 3(e), were composed of the Al₄Cu₉ phase. Figure 6(c) presents the grain orientation distribution within this region. In the U₁ layer, a fine-grained microstructure approximately 7 µm in thickness was observed adjacent to the copper base material (U₃ layer). The Al₂Cu grains within the layer exhibited a relatively refined morphology, with an average diameter of approximately 17 µm. The formation of the fine-grained layer in this region can be attributed to the relatively higher cooling rate at this location, which resulted from its proximity to the copper base material. The formation of the U₃ layer (Al₄Cu₉) located between the U₁ layer (Al₂Cu) and the copper substrate is primarily due to the higher copper concentration in that region. According to the effective heat of formation model, the formation of IMC phases at the welding interface not only satisfies the thermodynamic requirements but is also influenced by atomic concentration [28]. In the interfacial region between the U₁ layer (Al₂Cu) and the copper substrate, the effective concentration of Cu was relatively high, while Al was the limiting element. Consequently, Al₄Cu₉ was formed at this interface [28]. Similarly, within the U₁ layer, the blocky reaction products composed of Al₄Cu₉ were also formed. During the RSW process, spatter may occur as a result of excessive local current density or overheating. In the RSW of aluminum and copper, when the aluminum side melted while the copper remained in a solid state, spatters generated on the copper side were introduced into the molten aluminum. Due to the brief heating duration, the copper particles incorporated into the liquid aluminum did not have sufficient time to diffuse over a significant distance, leading to a relatively high local concentration of Cu atoms. Upon solidification, the blocky reaction products of Al₄Cu₉ were formed in situ.

Figure 7(a) shows the effect of welding current and welding time on the electrical resistance of the joint. The measured electrical resistances of the T2 copper test sheet (170 mm × 30 mm × 1 mm) and the A1060 aluminum test sheet (170 mm × 30 mm × 2 mm) were approximately 105 µΩ and 146 µΩ, respectively. Within the selected range of welding parameters, the resistance value of the RSWed Al/Cu joints remained relatively stable with variations in welding current and welding time, maintaining an approximate value of 112 µΩ. The resistance value of the joint is significantly lower than the average resistance (125.5 µΩ) of the copper and aluminum test sheets, indicating a higher current-carrying capacity at the weld spot. This also suggests that utilizing zinc as an interlayer to enhance the nugget size of the RSWed Al/Cu joint can effectively improve the current-carrying capacity of the joint.

Figure 7(b) shows the influence of welding current and welding time on the tensile shear load resistance of the RSWed Al/Cu joint. As the welding current increased or the welding time was extended, the tensile shear load of the RSWed Al/Cu joint exhibited a trend of initial increase followed by a subsequent decrease. The maximum tensile shear load, approximately 4.23 kN, was achieved when the welding current was set to 28 kA and the welding time was maintained at 400 ms. In the tensile test, the failure modes observed in the obtained joints were primarily categorized as interfacial tearing and button-type fracture. Here, interface tearing refers to the failure mode observed during tensile testing, wherein the joint separated along the interface region. In contrast, the button-type failure is characterized by the retention of the nugget on the copper side after fracture, while a corresponding hole was left on the aluminum side. All RSWed Al/Cu joints welded under conditions involving a welding current higher than 28 kA and a welding time longer than 400 ms exhibited button-type failure, whereas joints produced under other welding conditions experienced interfacial tearing failure.

When the RSWed joint failed in an interfacial tearing mode, the tensile shear load of the joint was primarily influenced by the nugget diameter and the interfacial reaction layer. The tensile shear load of the joint increases with the enlargement of the nugget diameter; however, the formation of thick reaction layer at the interface exerts an adverse effect on the joint's tensile shear strength. Within the welding current range of 25 kA to 28 kA and the welding time range of 325 ms to 400 ms, the nugget diameter increased with both increasing welding current and extended welding time (as illustrated in Fig. 1(c) and 1(d)). This trend directly contributes to the observed increase in tensile shear load of the RSWed Al/Cu joint under these welding conditions as shown in Fig. 7(b). In this context, the nugget diameter served as the primary factor influencing the tensile shear load of the RSWed Al/Cu joint. When the RSWed joint failed in a button-type mode, the nugget diameter and electrode indentation depth become the primary factors influencing the tensile shear load of the joint. The nugget diameter has a positive influence on the tensile shear load capacity of the joint. However, when the electrode indentation becomes excessive, the remaining thickness of the aluminum plate in the weld zone decreases. This can not only result in button-type failure of the joint but also reduce its tensile shear load capacity. When the welding current was larger 28 kA and the welding was longer 400 ms, the RSWed Al/Cu joint fabricated under these conditions exhibited a button-type failure under external loading, primarily due to increased electrode indentation (as illustrated in Fig. 1(c) and 1(d)). Within this parameter range, both increasing the welding current and extending the welding time resulted in greater electrode indentation depth. This phenomenon explains the observed reduction in the tensile shear load of the joint with higher welding currents or longer welding times, as depicted in Fig. 7(b). Compared to the tensile shear load of the directly RSWed joint reported in reference [24], the tensile shear load of the RSWed joint incorporating a zinc interlayer was significantly higher. For instance, the tensile shear load of the joint with a zinc interlayer reached approximately 4.23 kN (as shown in Fig. 7(b)), which is about 1.31 times greater than that of the direct RSWed joint, where the tensile shear load was measured at 3.22 kN under the same welding current of 28 kA [24]. This phenomenon can be attributed to the fact that the application of a zinc interlayer in the RSW of aluminum and copper facilitated an increase in the nugget diameter, thereby improving the tensile shear load capacity of the joint.

Figure 8(a) illustrates the fracture crack propagation path of the joint. This phenomenon was identified on the cross-section after the two fractured surfaces of the joint were reassembled. As illustrated, the fracture crack primarily propagated along the welding interface in the peripheral region of the weld, whereas the crack extended within the nugget in the central region of the weld. Figure 8(b) and 8(c) present the magnified views of locations F and G in Fig. 8(a), respectively. Residual reaction layers were observed on both sides of the fracture crack, indicating that the failure of the RSWed Al/Cu joint initiated within the reaction layer located in the peripheral region of the weld under the applied external load. Based on the structure of the interfacial reaction layer shown in Fig. 2, it can be inferred that the fracture crack propagated within the U₁ layer. The relatively thick U₁ layer, composed of Al₂Cu, formed at the interface and acted as the weak link in the RSWed Al/Cu joint, leading to fracture under external force. As illustrated in Fig. 8(c), the fracture crack deviated toward the nugget zone during the tensile shear testing of the joint. This phenomenon is attributed to the effect of additional torque generated during the tensile shear testing of the joint. During the tensile shear test of the joint, the local region of the joint experienced torsion due to the additional torque generated. As a result, the direction of the externally applied load gradually deviated from being parallel to the welding interface and shifted toward the nugget zone [29]. This load redistribution caused the crack to propagate toward the nugget. Even so, the reaction layer at the interface cracked (at location K in Fig. 8(c)), but the expansion distance along this direction was relatively short. This also indicates that the reaction layer composed of IMCs was the weak link of the joint. The relatively thick IMCs layer formed at the welding interface is the primary cause of joint cracking along the interface, whereas an increase in nugget diameter can enhance the load-bearing area of the joint. In other words, the interfacial IMC layer determined the failure mode of the joint, while the nugget diameter influenced the magnitude of the tensile shear load that the joint can withstand. Therefore, while increasing the nugget size, effectively controlling the thickness of the IMCs layer formed at the welding interface is crucial for further enhancing joint performance, and remains an important area for future research.

In this study, T2 copper plate and A1060 aluminum plate were welded by using RSW with an interlayer of zinc. The salient conclusions obtained from this study are as follows:

(1) RSW of aluminum and copper using a zinc interlayer can promote the formation of larger nuggets through low-temperature eutectic liquefaction of the zinc and aluminum system.

(2) At the welding interface zone of the joint, an Al₂Cu layer was formed adjacent to the copper side, while a eutectic structure layer composed of Al₂Cu and α-Al was observed near the nugget on the aluminum side.

(3) As the welding current increased or the welding time was extended, the tensile shear load of the joint exhibited a trend of initial increase followed by a subsequent decrease. The maximum tensile shear load, approximately 4.23 kN, was achieved when the welding current was set to 28 kA and the welding time was maintained at 400 ms.

(4) The failure modes of the obtained joints were primarily manifested in two forms: interfacial tearing and button-type fracture. The IMC layer formed at the welding interface was the main cause of interface tearing in the joint.