In-Depth Characterization of Laser-Welded Aluminum-and-Copper Dissimilar Joint for Electric Vehicle Battery Connections

With advancements in the automotive industry, the demand for electric vehicles (EVs) has remarkably increased in recent years. However, the EV battery, which is a vital part of the EV, poses certain challenges that limit the performance of the EVs. The joining of dissimilar materials for different components affects the electrical and mechanical performances of EV batteries. Laser beam welding is a promising technique for joining Al and Cu for application in secondary battery fabrication because of the precise control over heat input and high process speed. However, the production of Al–Cu joints remains challenging because of the differences between their thermal and metallurgical properties and the resulting formation of brittle and hard intermetallic compounds, which reduce mechanical and electric properties. Thus, it is vital to characterize the weld to improve joint performance and enhance the laser welding process. This study investigates the joining of an Al alloy (AA1050) with Ni-coated Cu using a continuous-wave Yb fiber laser. The evaluation of the weld morphology showed a correlation between the weld characteristics and process parameters (laser power and welding speed). The weld interface width and penetration depth into the lower sheet (Cu) both increased with increasing heat input. Optical microscopy of the weld cross-section revealed many defects, such as voids and cracks. Scanning electron microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDS) was employed to examine the weld microstructure. The composition analysis revealed the presence of mixed morphology of Al–Cu eutectic alloy (α-Al+Θ-Al2Cu) phase in the form of dendrites in the weld fusion zone with traces of the highly brittle Al4Cu9 phase at a high heat input condition. Furthermore, the electrical contact resistance of the weld seam was measured to determine the correlation between heat input and resistance. In addition, Vickers microhardness measurements were performed on the weld cross-section to validate the SEM/EDS results.


Introduction
Recently, the automotive sector has witnessed a remarkable shift from conventional internal combustion engines (ICEs) to electric vehicles (EVs)/hybrid-electric vehicles [1,2]. EVs have almost net-zero emissions, making them a better alternative to traditional ICEbased vehicles [3,4]. Electric batteries, which are a vital part of EVs, currently have low efficiency and low energy density, which limits the driving range compared with ICE-based vehicles [5]. Hence, battery efficiency and energy density must be improved to improve the performance of EVs [6,7].
The EV battery pack consists of several thousand battery cells interconnected using tabs and busbars of dissimilar materials; typically aluminum and copper, in the case of Li-ion batteries. Thus, the reliable joining of different connections and the materials used is of significantly enhanced the ductility of the joints. Similarly, Chen et al. [27] studied Cu-Al welding using Ni-P-coated Al to obtain Al-Cu joints without brittle IMCs. The obtained welds were donut-shaped and free from detrimental Al-Cu IMCs; weld defects were also inhibited. Huang et al. [28] investigated the metal mixing phenomenon in the laser keyhole welding of dissimilar metals using experimental and numerical modeling approaches. They found that recoil pressure contributes to the upward flow of the melt pool, resulting in the migration of Cu from the bottom to the top. Marangoni waves generate a backward flow and two side vortices that promote the intermixing of the base metals. Furthermore, it was reported that laser power is directly related to both recoil pressure and Marangoni stress, while increasing the welding speed inhibited the mixing of base metals in the fusion zone.
The above literature survey reveals that many studies have been conducted on laserwelded joints of dissimilar metals. However, further studies are needed to evaluate the effects of laser process parameters on quality of joints in terms of weld morphology and weld defects. In this study, laser lap welding of Al (AA1050) and thin Ni-coated Cu was conducted. The effects of laser power and welding speed on the weld morphology and weld defects were investigated using optical microscopy. Subsequently, the microstructure and composition of the weld zone were characterized using scanning electron microscopy (SEM) with energy-dispersive X-ray spectroscopy (EDS), and the electrical properties of the joint were evaluated by measuring the electrical contact resistance of the weld joint. Additionally, the mechanical properties of the weld were determined using the Vickers microhardness test.

Materials and Methods
In this experimental study, 70 mm × 40 mm plates of Cu (>99.5%, 1.5 mm thick) coated with a thin layer of Ni (3 µm) and Al alloy (AA1050, 0.75 mm thick) were welded in a lap configuration. Table 1 lists the chemical composition of the AA1050 alloy.  Figure 1 illustrates the experimental setup and clamping device (jig). The samples were arranged in a lap-joint configuration by positioning aluminum on top of copper with a 10 mm wide overlap, as shown in Figure 2. The samples were cleaned with ethanol to remove impurities from the surfaces of the plates. A single-mode continuous-wave Yb: fiber laser (Raycus RFL-C2000) with a maximum power output of 2000 W and wavelength of 1080 nm was employed for welding. The laser beam was delivered by an optical fiber and focused on the specimen's surface using a focusing lens with a focal length of 182 mm. The beam size focused on the workpiece surface was 100 µm. In laser lap welding, full penetration of the laser beam into the top sheet is required to achieve complete fusion of the base metals. For this, a preliminary bead-on-plate test was conducted by gradually varying the speed and laser power to determine the process conditions necessary for achieving full penetration of the laser beam onto the Al plate.  To obtain the full penetration conditions for the Al plate, three experimental groups named A, B, and C were designed with laser powers of 1500, 1600, and 1700 W, respectively, and the scan speed was varied from 15 to 30 mm/s in 5 mm/s increments for each power level. Laser welding was performed with the beam focused on the top surface of the Al plate using the experimental process parameters listed in Table 2.    To obtain the full penetration conditions for the Al plate, three experimental groups named A, B, and C were designed with laser powers of 1500, 1600, and 1700 W, respectively, and the scan speed was varied from 15 to 30 mm/s in 5 mm/s increments for each power level. Laser welding was performed with the beam focused on the top surface of the Al plate using the experimental process parameters listed in Table 2.  To obtain the full penetration conditions for the Al plate, three experimental groups named A, B, and C were designed with laser powers of 1500, 1600, and 1700 W, respectively, and the scan speed was varied from 15 to 30 mm/s in 5 mm/s increments for each power level. Laser welding was performed with the beam focused on the top surface of the Al plate using the experimental process parameters listed in Table 2. After laser welding, the samples were preliminarily observed using optical microscopy (AM4515 Series, Dino Lite) to examine the weld surface. To characterize the weld microstructure and weld geometry, specimens (15 mm × 10 mm) were prepared by diamondcutting the welded samples at the center point. Standard metallographic samples were prepared by cold mounting the specimens in resin and subsequently grinding them using SiC papers, followed by polishing in diamond suspensions of 3 and 1 µm.
Optical microscopy was performed to investigate the cross-sectional morphology of laser-welded joints. The weld geometry of the cross-section (see Figure 3) was characterized by measuring the top bead width (l), penetration depth (p), and weld interface width (w). Microstructure analysis of the weld cross-sections was performed using SEM (TESCAN VEGA3) equipped with EDS. The weld seam microstructure was examined using SEM, and EDS was employed to perform elemental composition analysis of the IMCs in the weld fusion zone (FZ). After laser welding, the samples were preliminarily observed using optical microscopy (AM4515 Series, Dino Lite) to examine the weld surface. To characterize the weld microstructure and weld geometry, specimens (15 mm × 10 mm) were prepared by diamond-cutting the welded samples at the center point. Standard metallographic samples were prepared by cold mounting the specimens in resin and subsequently grinding them using SiC papers, followed by polishing in diamond suspensions of 3 and 1 μm.
Optical microscopy was performed to investigate the cross-sectional morphology of laser-welded joints. The weld geometry of the cross-section (see Figure 3) was characterized by measuring the top bead width (l), penetration depth (p), and weld interface width (w). Microstructure analysis of the weld cross-sections was performed using SEM (TESCAN VEGA3) equipped with EDS. The weld seam microstructure was examined using SEM, and EDS was employed to perform elemental composition analysis of the IMCs in the weld fusion zone (FZ). The electrical contact resistance of the weld joint was measured using the four-wire method [15]. The measurements were performed using a Keithley Integra Series Model 2460 ohmmeter with a measurement current of 100 mA. Each measurement was performed by applying a current for 60 s and then taking the average value of the measured resistance. Measurements were performed at a testing length of 30 mm across the weld joint. A schematic of the resistance measurement setup is shown in Figure 4a, and the current flow direction is illustrated in Figure 4b. The electrical contact resistance of the weld joint was measured using the four-wire method [15]. The measurements were performed using a Keithley Integra Series Model 2460 ohmmeter with a measurement current of 100 mA. Each measurement was performed by applying a current for 60 s and then taking the average value of the measured resistance. Measurements were performed at a testing length of 30 mm across the weld joint. A schematic of the resistance measurement setup is shown in Figure 4a, and the current flow direction is illustrated in Figure 4b.  Microhardness was measured using a Vickers microhardness tester (Mitutoyo HM-100 series) to investigate the mechanical properties of the weld. Measurements were performed at a 50-gf load and 10 s dwell time. Indentations were carried out horizontally along the weld cross-section at three different positions, as shown in Figure 5.

Analysis of the Weld Morphology and Defects
Optical micrographs of the top beads of the weld joints are shown in Figure 6. Under varying parameters, laser welding yielded weld joints of uniform bead width. The bead width tended to increase with laser power and decrease with increasing welding speed. The correlation between the bead width, laser power, and welding speed can be explained by the change in the heat input because of the changes in parameters. Optical microscopy of the weld joint revealed the presence of surface defects, such as spatter, while cracks were formed at a relatively high heat input as a result of hot cracking inside the melt pool. Surface cracks were clearly visible in groups B and C, while in group A no cracks were observed; however, spatters were observed around the weld joint. Overall, the severity of cracking tended to decrease with decreasing heat input. As well as the heat input, cracks can also be affected by power density (applied laser energy per unit area and time). Group C including C4 used the higher power density condition (see Table 2) compared with other groups (A and B). This high power density condition may be the reason for the crack generation in C4, though the heat input is relatively low. Microhardness was measured using a Vickers microhardness tester (Mitutoyo HM-100 series) to investigate the mechanical properties of the weld. Measurements were performed at a 50-gf load and 10 s dwell time. Indentations were carried out horizontally along the weld cross-section at three different positions, as shown in Figure 5.

Analysis of the Weld Morphology and Defects
Optical micrographs of the top beads of the weld joints are shown in Figure 6 varying parameters, laser welding yielded weld joints of uniform bead width. T width tended to increase with laser power and decrease with increasing weldin The correlation between the bead width, laser power, and welding speed can be ex by the change in the heat input because of the changes in parameters. Optical mic of the weld joint revealed the presence of surface defects, such as spatter, whil were formed at a relatively high heat input as a result of hot cracking inside the m Surface cracks were clearly visible in groups B and C, while in group A no crac observed; however, spatters were observed around the weld joint. Overall, the se cracking tended to decrease with decreasing heat input. As well as the heat inpu can also be affected by power density (applied laser energy per unit area and time C including C4 used the higher power density condition (see Table 2) compar other groups (A and B). This high power density condition may be the reason for t generation in C4, though the heat input is relatively low.

Analysis of the Weld Morphology and Defects
Optical micrographs of the top beads of the weld joints are shown in Figure 6. Under varying parameters, laser welding yielded weld joints of uniform bead width. The bead width tended to increase with laser power and decrease with increasing welding speed. The correlation between the bead width, laser power, and welding speed can be explained by the change in the heat input because of the changes in parameters. Optical microscopy of the weld joint revealed the presence of surface defects, such as spatter, while cracks were formed at a relatively high heat input as a result of hot cracking inside the melt pool. Surface cracks were clearly visible in groups B and C, while in group A no cracks were observed; however, spatters were observed around the weld joint. Overall, the severity of cracking tended to decrease with decreasing heat input. As well as the heat input, cracks can also be affected by power density (applied laser energy per unit area and time). Group C including C4 used the higher power density condition (see Table 2) compared with other groups (A and B). This high power density condition may be the reason for the crack generation in C4, though the heat input is relatively low. Cross-sectional images of the welds obtained using optical microscopy are shown in Figure 7. All samples show a fully developed weld in the Cu sheet. The laser beam irradiation melted Al, and the molten weld metal penetrated the Cu sheet through the keyhole mode. The top part of the weld FZ exhibits a convex profile with a high width-to-depth ratio, whereas the lower part of the FZ exhibits a low width-to-depth ratio. This is attributable to the low melting temperature of Al, which resulted in a wide FZ in the top part, owing to the high energy density of the laser beam. Further observations revealed a large number of weld defects, such as cracks and voids of different sizes. At a high heat input, the temperature inside the melt pool increased above the vaporization point of Al, which generated a recoil pressure [21]. The recoil pressure inside the melt pool tends to render the melt pool flow more turbulent. This pressure and atmospheric gases trapped inside the FZ during welding lead to the formation of voids after solidification. The formation of porosity/void in the laser welding is often caused by metallurgical factors, such as dissolved hydrogen in the melt pool. Other than the trapping of gas, environmental factors also influence porosity formation, such as air temperature, air humidity, and dissolved oxygen. Increasing both air temperature and humidity results in increased porosity [29]. In addition to other factors, heat input in laser welding also affects porosity formation. As can be seen in Figure 7, compared to high heat input conditions, the porosity level in samples such as A2, A4, and B4 is low. It is pertinent to mention here that heat input is the dominant factor for a high degree of porosity in this type of welding. Because of the high heat input, more energy is accumulated in the weld metal, resulting in an unstable and turbulent melt pool flow. The recoil pressure generated in the melt pool coupled with the turbulent melt pool flow leads to porosity in the weld metal due to the trapping of dissolved gases that form bubbles in the melt pool. Therefore, the high porosity in group C and some of the group B samples can be attributed to the high heat input. In the case of large heat input, longitudinal cracks are observed, as can be seen in sample B1, owing to Cross-sectional images of the welds obtained using optical microscopy are shown in Figure 7. All samples show a fully developed weld in the Cu sheet. The laser beam irradiation melted Al, and the molten weld metal penetrated the Cu sheet through the keyhole mode. The top part of the weld FZ exhibits a convex profile with a high width-todepth ratio, whereas the lower part of the FZ exhibits a low width-to-depth ratio. This is attributable to the low melting temperature of Al, which resulted in a wide FZ in the top part, owing to the high energy density of the laser beam. Further observations revealed a large number of weld defects, such as cracks and voids of different sizes. At a high heat input, the temperature inside the melt pool increased above the vaporization point of Al, which generated a recoil pressure [21]. The recoil pressure inside the melt pool tends to render the melt pool flow more turbulent. This pressure and atmospheric gases trapped inside the FZ during welding lead to the formation of voids after solidification. The formation of porosity/void in the laser welding is often caused by metallurgical factors, such as dissolved hydrogen in the melt pool. Other than the trapping of gas, environmental factors also influence porosity formation, such as air temperature, air humidity, and dissolved oxygen. Increasing both air temperature and humidity results in increased porosity [29]. In addition to other factors, heat input in laser welding also affects porosity formation. As can be seen in Figure 7, compared to high heat input conditions, the porosity level in samples such as A2, A4, and B4 is low. It is pertinent to mention here that heat input is the dominant factor for a high degree of porosity in this type of welding. Because of the high heat input, more energy is accumulated in the weld metal, resulting in an unstable and turbulent melt pool flow. The recoil pressure generated in the melt pool coupled with the turbulent melt pool flow leads to porosity in the weld metal due to the trapping of dissolved gases that form bubbles in the melt pool. Therefore, the high porosity in group C and some of the group B samples can be attributed to the high heat input. In the case of large heat input, longitudinal cracks are observed, as can be seen in sample B1, owing to the possible formation of brittle IMCs and the thermal cycling effect that made the weld prone  The weld joint was characterized by measuring the weld geometry. The penetrat depth (p) into the lower Cu sheet was measured from the interface between the two m terials, whereas the weld interface width (w) was measured at the interface between two sheets, as depicted in Figure 3. The measured penetration depth and weld interf width were plotted against the power (Figure 8). With increasing power, both the pe tration depth and the weld interface width tended to increase. According to Figure 8a 20 mm/s, the weld width increases from 1525.57 μm at 1500 W to 2355.46 μm at 1700 whereas, at other speeds, a slight change in the weld width with an increase in powe observed. In the case of 15 mm/s scan speed, a relatively large increase in the interf width was found. The scan speed of 15 mm/s resulted in the highest heat input, as sho in Table 2. The rate of increase in the heat input was 33.3% when the scan speed chang from 20 mm/s to 15 mm/s. However, when the scan speed changed from 25 mm/s to mm/s and from 30 mm/s to 25 mm/s, the rates of increase in the heat input were only 2 and 20%, respectively. It is considered that the relatively large increase in the heat in induced the large increase in the interface width. As shown in Figure 8b, the penetrat depth correlates with the laser power and welding speed. As is evident from Figure 8, penetration depth tends to increase with laser power and decrease with increasing we ing speed. The conditions used in this study yielded a weld seam with partial penetrati which eventually reduced the intermixing of the base metals and the corresponding f mation of IMCs, thus improving the joint strength. As reported in [20,27], the partial p etration of the weld in the lower base metal resulted in less mixing of Cu and the w metal, which reduced the formation of brittle and hard IMCs. It can be concluded that l heat input is favorable for minimizing the generation of detrimental IMCs, which even ally enhances joint strength. The weld joint was characterized by measuring the weld geometry. The penetration depth (p) into the lower Cu sheet was measured from the interface between the two materials, whereas the weld interface width (w) was measured at the interface between the two sheets, as depicted in Figure 3. The measured penetration depth and weld interface width were plotted against the power (Figure 8). With increasing power, both the penetration depth and the weld interface width tended to increase. According to Figure 8a, at 20 mm/s, the weld width increases from 1525.57 µm at 1500 W to 2355.46 µm at 1700 W, whereas, at other speeds, a slight change in the weld width with an increase in power is observed. In the case of 15 mm/s scan speed, a relatively large increase in the interface width was found. The scan speed of 15 mm/s resulted in the highest heat input, as shown in Table 2. The rate of increase in the heat input was 33.3% when the scan speed changed from 20 mm/s to 15 mm/s. However, when the scan speed changed from 25 mm/s to 20 mm/s and from 30 mm/s to 25 mm/s, the rates of increase in the heat input were only 25% and 20%, respectively. It is considered that the relatively large increase in the heat input induced the large increase in the interface width. As shown in Figure 8b, the penetration depth correlates with the laser power and welding speed. As is evident from Figure 8, the penetration depth tends to increase with laser power and decrease with increasing welding speed. The conditions used in this study yielded a weld seam with partial penetration, which eventually reduced the intermixing of the base metals and the corresponding formation of IMCs, thus improving the joint strength. As reported in [20,27], the partial penetration of the weld in the lower base metal resulted in less mixing of Cu and the weld metal, which reduced the formation of brittle and hard IMCs. It can be concluded that low heat input is favorable for minimizing the generation of detrimental IMCs, which eventually enhances joint strength.

Analysis of the Weld Microstructure and Composition
The microstructure of the selected low-and high-heat-input welds was analyzed. Figure 9 shows the SEM images of sample A4 corresponding to the low heat input condition, and the results of the EDS point and line scanning analysis are shown in Table 3 and Figure 10, respectively. Different zones in the weld FZ are indicated by letters A, B, and C, as shown in Figure 9. According to the Al-Cu binary phase diagram, the composition analysis given in Table 3 shows that the weld metal is entirely an Al-Cu eutectic alloy (α-Al+Θ-Al2Cu) with 6.38-20.18 at.% Cu. In the upper part of the FZ, as indicated by letter A, a mixed morphology of dark dendrite (α-Al eutectic) and gray structure (Θ-Al2Cu eutectic) phases is observed (see Figure 9). The EDS point analysis results show a high content of Al in the FZ, which is an indication of the Al-Cu eutectic phase. Near the Al-Cu interface, indicated by letter B, the microstructure is a mixture of gray and dark structures (a dark dendritic structure is more dominant than the gray structure). Region B is close to the Al-Cu interface, and the presence of a darker dendrite structure (α-Al eutectic) is likely due to the Ni layer that inhibited the diffusion of Cu into the weld metal in this region. Point analysis of spectra 4, 5, and 6 shows similar chemical compositions, suggesting a similar type of phase formation as that of region A. The composition corresponding to spectrum 7 indicates a solid solution of Cu with 89.85 at.% Cu with traces of Al and Ni. Near the FZ-Cu interface, indicated by letter C, the composition is the same as those of A and B; however, morphological changes in the shape of columnar structures are observed near the interface. It is expected that, at the FZ-Cu interface, the diffusion of Cu into the weld metal intermixing led to the growth of the columnar morphology. The evolution of different microstructures is controlled by different parameters, including cooling rate and alloy composition of the weld pool. In the weld pool, these parameters significantly vary from region to region and process to process, which makes it quite challenging to predict the microstructures of the weld zone. However, columnar grains usually grow in a direction that is opposite to the cooling direction, and they can grow further when the cooling rate is relatively low.
To further analyze the IMC formation, EDS line scanning was performed at the FZ-Cu interface, and the results are shown in Figure 10. The line scanning indicates an almost uniform distribution of Al and Cu in the FZ; however, the concentration gradient at approximately 24-32 μm along the scanning path shows a decrease in Al and an increase in Cu content, which hints at the IMC phase change. A similar type of line scanning was also performed vertically along the FZ from top to bottom, as shown in Figure 11. Compared to the line scanning at the interface, the vertical line scanning results show non-uniform distributions of Al and Cu, with the Al content dropping to almost zero at the interface.

Analysis of the Weld Microstructure and Composition
The microstructure of the selected low-and high-heat-input welds was analyzed. Figure 9 shows the SEM images of sample A4 corresponding to the low heat input condition, and the results of the EDS point and line scanning analysis are shown in Table 3 and Figure 10, respectively. Different zones in the weld FZ are indicated by letters A, B, and C, as shown in Figure 9. According to the Al-Cu binary phase diagram, the composition analysis given in Table 3 shows that the weld metal is entirely an Al-Cu eutectic alloy (α-Al+Θ-Al 2 Cu) with 6.38-20.18 at.% Cu. In the upper part of the FZ, as indicated by letter A, a mixed morphology of dark dendrite (α-Al eutectic) and gray structure (Θ-Al 2 Cu eutectic) phases is observed (see Figure 9). The EDS point analysis results show a high content of Al in the FZ, which is an indication of the Al-Cu eutectic phase. Near the Al-Cu interface, indicated by letter B, the microstructure is a mixture of gray and dark structures (a dark dendritic structure is more dominant than the gray structure). Region B is close to the Al-Cu interface, and the presence of a darker dendrite structure (α-Al eutectic) is likely due to the Ni layer that inhibited the diffusion of Cu into the weld metal in this region. Point analysis of spectra 4, 5, and 6 shows similar chemical compositions, suggesting a similar type of phase formation as that of region A. The composition corresponding to spectrum 7 indicates a solid solution of Cu with 89.85 at.% Cu with traces of Al and Ni. Near the FZ-Cu interface, indicated by letter C, the composition is the same as those of A and B; however, morphological changes in the shape of columnar structures are observed near the interface. It is expected that, at the FZ-Cu interface, the diffusion of Cu into the weld metal intermixing led to the growth of the columnar morphology. The evolution of different microstructures is controlled by different parameters, including cooling rate and alloy composition of the weld pool. In the weld pool, these parameters significantly vary from region to region and process to process, which makes it quite challenging to predict the microstructures of the weld zone. However, columnar grains usually grow in a direction that is opposite to the cooling direction, and they can grow further when the cooling rate is relatively low. In summary, the analysis results provide an overall distribution of constituent eleme in the weld and the probability of IMC formation inside the weld and at the interface.    To further analyze the IMC formation, EDS line scanning was performed at the FZ-Cu interface, and the results are shown in Figure 10. The line scanning indicates an almost uniform distribution of Al and Cu in the FZ; however, the concentration gradient at approximately 24-32 µm along the scanning path shows a decrease in Al and an increase in Cu content, which hints at the IMC phase change. A similar type of line scanning was also performed vertically along the FZ from top to bottom, as shown in Figure 11. Compared to the line scanning at the interface, the vertical line scanning results show non-uniform distributions of Al and Cu, with the Al content dropping to almost zero at the interface. In summary, the analysis results provide an overall distribution of constituent elements in the weld and the probability of IMC formation inside the weld and at the interface.    Table 3. Chemical composition of zones denoted in Figure 9. SEM images of sample C2 (high heat input) are shown in Figure 12, and the chemical compositions of the marked points in different zones are listed in Table 4. It is evident from the appearance of the weld that the penetration depth of Al is greater than in sample A4. The number and size of the defects are also notably higher in sample C2. A magnified view of region A in Figure 12 clearly shows voids or pores of varying sizes. Weld cracks are also observed in the same area. These cracks can possibly be hot cracks, which mostly formed through the generation of unbearable shrinkage stress during the solidification of the molten metal. This type of crack is very common during the fusion welding of nonferrous materials. However, this type of cracking can also occur because of the formation of highly brittle IMCs after the solidification of the melt pool. Because dissimilar welding of Al and Cu involves the generation of brittle IMC phases, crack generation is likely caused by these IMCs. It can be observed that the average Cu content in the weld metal is slightly higher than under low heat input. The composition of the marked spectrum 2 in region A shows an increased Cu content (76.84 at.% Cu), which clearly confirms the formation of a brittle Al 4 Cu 9 phase near the crack. The results of spectra 1 and 3 suggest the possible presence of an Al-Cu eutectic alloy (α-Al+ Θ-Al 2 Cu), similar to previous results. Region B, near the Al-Cu interface, also shows the type of morphological appearance and composition very similar to the previous sample. The enlarged SEM image of region B also exhibits defects in this region. Using the Al-Ni binary phase diagram, the presence of Ni 3 Al can be detected in spectrum 7, owing to the Ni interlayer. In region C, at the FZ-Cu interface, a thin continuous layer of IMC is observed in the form of columnar dendrites, as indicated in Figure 12. The results of spectrum 8 show a similar composition to the previous cases, whereas spectrum 9 shows the possible presence of the Θ-Al 2 Cu eutectic phase. At the same time, at the position of spectrum 10, the composition is 100% Cu. formed through the generation of unbearable shrinkage stress during the solidification of the molten metal. This type of crack is very common during the fusion welding of nonferrous materials. However, this type of cracking can also occur because of the formation of highly brittle IMCs after the solidification of the melt pool. Because dissimilar welding of Al and Cu involves the generation of brittle IMC phases, crack generation is likely caused by these IMCs. It can be observed that the average Cu content in the weld metal is slightly higher than under low heat input. The composition of the marked spectrum 2 in region A shows an increased Cu content (76.84 at.% Cu), which clearly confirms the formation of a brittle Al4Cu9 phase near the crack. The results of spectra 1 and 3 suggest the possible presence of an Al-Cu eutectic alloy (α-Al+ Θ-Al2Cu), similar to previous results. Region B, near the Al-Cu interface, also shows the type of morphological appearance and composition very similar to the previous sample. The enlarged SEM image of region B also exhibits defects in this region. Using the Al-Ni binary phase diagram, the presence of Ni3Al can be detected in spectrum 7, owing to the Ni interlayer. In region C, at the FZ-Cu interface, a thin continuous layer of IMC is observed in the form of columnar dendrites, as indicated in Figure 12. The results of spectrum 8 show a similar composition to the previous cases, whereas spectrum 9 shows the possible presence of the Θ-Al2Cu eutectic phase. At the same time, at the position of spectrum 10, the composition is 100% Cu.        Based on the SEM and EDS results, sample C2 took the highest heat input, while sample A4 took the lowest heat input. With the increase in heat input by altering the welding parameters, the diffusion kinetics increased, enhancing the amount of Cu inside the weld zone. Furthermore, the penetration depth of the melted Al also increased with heat input. Although the high heat input increased the penetration depth of the weld metal into Cu, it also introduced a higher amount of weld defects, such as porosity and weld Based on the SEM and EDS results, sample C2 took the highest heat input, while sample A4 took the lowest heat input. With the increase in heat input by altering the welding parameters, the diffusion kinetics increased, enhancing the amount of Cu inside the weld zone. Furthermore, the penetration depth of the melted Al also increased with heat input. Although the high heat input increased the penetration depth of the weld metal into Cu, it also introduced a higher amount of weld defects, such as porosity and weld cracks, leading to the formation of brittle IMCs. When joining dissimilar metals, the formation of IMCs is inevitable to form a metallurgical bond. The IMC amount and phases directly affect the mechanical and electrical properties of the joint; hence, the formation of IMCs must be controlled by minimizing interdiffusion. A high heat input increases the weld melt pool temperature, further enhancing diffusion, resulting in a higher amount of detrimental IMCs, solute segregation, and brittleness in the weld metal. Thus, it is advisable to perform laser welding at a low heat input.

Vicker's Microhardness Test
To characterize the mechanical properties of the weld joint, microhardness was measured using the Vickers hardness method. Figure 14 shows the Vickers microhardness test profiles as a function of distance from the weld center. To prevent deformation by adjacent indentations, the indent spacing was at least three times larger than the measured diagonal length of the first indentation. In Figure 14a, the microhardness profile corresponding to specimen A4 (low-heat input condition) is presented. Evidently, the weld FZ has higher hardness than the base metals, which can be attributed to the local microstructure and formation of brittle and hard IMC phases. The increased hardness can also be explained by the enhanced ionic and covalent bonds, and reduced metallic bonding in the weld metal zone, as reported in [23]. The microhardness values of the Al and Cu base metals were 24 and 54 HV, respectively, whereas the hardness of the weld was in the range of 140-310 HV. This hardness value corresponds to the less brittle Al 2 Cu IMC phase (Table 5), which is vital for the metallurgical bonding of the base metals in dissimilar welding. Local variations in hardness inside the FZ are observed owing to the presence of pores and voids.     Figure 14. Hardness profile as a function of the distance from the weld center: (a) A4; (b)B4; (c) C2; (d) C4.

Electrical Contact Resistance
The results of the electrical contact resistance measurements of the weld joint are shown in Figure 15. The resistance was measured following the four-point method: the principle of the resistance measurement is explained in [15]. The resistance correlated with the sample number and legends 1, 2, 3, and 4 show an increase in the welding speed from 15 to 30 mm/s for each experimental group, as shown in Figure 15. The measured resistance is in the range of 335.60-416.28 μΩ and the lowest resistance is observed for sample C4 (laser power of 1700 W and welding speed of 30 mm/s). Samples C1 and C2 exhibit the highest values, which is likely due to weld defects and weld morphology. No  Figure 14b shows the microhardness profile corresponding to sample B4. As can be seen from the figure, the hardness profile is almost the same as that of sample A4. However, a slight fluctuation in the local hardness distribution can be observed possibly due to the evolution of different microstructures and porosity formation in the weld metal. Figure 14c depicts the microhardness profile of sample C2 (high heat input condition), where a significant increase in hardness value (up to 530 HV) can be observed. This increase in microhardness is in line with the results obtained from the SEM/EDS analysis. The high heat input led to the increased mixing of Cu into the FZ; thus, possibly leading to the formation of the Cu-rich Al 4 Cu 9 phase (see Table 4), based on the EDS composition results in the preceding section (see Table 4). The presence of the brittle phase caused the formation and propagation of cracks in the FZ, as can be observed in the SEM image of the sample (see Figure 12). Similar to the previous sample, the microhardness of the FZ is higher than that of the base metals. Furthermore, the hardness profile shows that microhardness values at the interface and lower part of the FZ are higher than that in the upper part of the FZ. Figure 14d presents the hardness profile corresponding to sample C4. The maximum hardness value observed was 441 HV, while the local minimum hardness value was 105.1 HV in the upper region, possibly due to the presence of voids. In summary, the microhardness in the FZ is higher than those of the base metals because of the formation of hard and brittle IMC phases and microstructural evolution. More brittle and hard phases were detected at high heat inputs because of the increased diffusion of Cu inside the FZ and subsequent intermixing in the weld metal.

Electrical Contact Resistance
The results of the electrical contact resistance measurements of the weld joint are shown in Figure 15. The resistance was measured following the four-point method: the principle of the resistance measurement is explained in [15]. The resistance correlated with the sample number and legends 1, 2, 3, and 4 show an increase in the welding speed from 15 to 30 mm/s for each experimental group, as shown in Figure 15. The measured resistance is in the range of 335.60-416.28 µΩ and the lowest resistance is observed for sample C4 (laser power of 1700 W and welding speed of 30 mm/s). Samples C1 and C2 exhibit the highest values, which is likely due to weld defects and weld morphology. No significant differences in the measured resistance between the groups have been detected; however, the resistance tended to decrease with increasing welding speed within the same group. The high resistance in the case of high heat input is also characterized by greater mixing of Cu in the weld, resulting in the formation of more brittle IMCs with high resistivity. With increasing penetration depth into Cu, the amount of Cu mixed in the weld is expected to increase, and thus, more Cu-rich brittle IMCs can be formed in the FZ [30]. It is vital to note that electrical contact resistance has a great impact on the overall performance of the battery. In summary, the formation of IMCs significantly enhances the electric resistance owing to the increased mixing of the base metals. Moreover, weld defects and weld morphology also affect the electrical contact resistance of the joint. tivity. With increasing penetration depth into Cu, the amount of Cu mixed in the weld is expected to increase, and thus, more Cu-rich brittle IMCs can be formed in the FZ [30]. It is vital to note that electrical contact resistance has a great impact on the overall performance of the battery. In summary, the formation of IMCs significantly enhances the electric resistance owing to the increased mixing of the base metals. Moreover, weld defects and weld morphology also affect the electrical contact resistance of the joint.

Conclusions
In this study, thin Al sheets (AA1050) were welded with thick Cu sheets using CW fiber laser welding with varying laser power and welding speed. The morphology, microstructure, electrical contact resistance, and mechanical properties of the welds were evaluated. The experimental results are summarized as follows:

•
The laser power and welding speed affected the weld characteristics. Weld defects, such as voids and cracks, appeared on the weld cross-sections.
• SEM/EDS of the weld microstructure revealed that the weld metal consisted of mixed-morphology dark (α-Al eutectic) and gray (Θ-Al2Cu eutectic) phases.
• At a high heat input, a more brittle Al4Cu9 IMC phase was formed, which likely caused the formation of cracks in the weld metal.
• The electrical contact resistance tended to decrease with increasing welding speed; however, the difference in resistance was negligible with varying laser power.

Conclusions
In this study, thin Al sheets (AA1050) were welded with thick Cu sheets using CW fiber laser welding with varying laser power and welding speed. The morphology, microstructure, electrical contact resistance, and mechanical properties of the welds were evaluated. The experimental results are summarized as follows:

•
The laser power and welding speed affected the weld characteristics. Weld defects, such as voids and cracks, appeared on the weld cross-sections. • SEM/EDS of the weld microstructure revealed that the weld metal consisted of mixedmorphology dark (α-Al eutectic) and gray (Θ-Al 2 Cu eutectic) phases. • At a high heat input, a more brittle Al 4 Cu 9 IMC phase was formed, which likely caused the formation of cracks in the weld metal.

•
The electrical contact resistance tended to decrease with increasing welding speed; however, the difference in resistance was negligible with varying laser power.

•
The Vickers microhardness test validated the IMC phase identification in the weld metal. Moreover, the hardness of the weld metal was higher than that of the base metals, indicating the possibility of IMC formation.