*3.1. Microstructures and Microhardness Variation of FSW Al–Cu Dissimilar Joints*

Figure 2 represents the backscattering electron images (SEM/BEI) of cross-sectional micrographs within the WZ of various FSW Al–Cu dissimilar joints for illustration. The cross-sectional images are taken at the stable part of the seam. The light gray area is the Cu matrix, and the dark gray area is the Al matrix. It can be seen that sound A1/C1, A6/C1, and A1/C2 dissimilar joints are successfully achieved by the FSW process, while the pure Cu (C1100) and brass (C2600) plates are always positioned at the advancing side (AS). The good welding quality of all of the Al–Cu joints is obtained in the present study without the typical cavity defects that tend to occur within the WZ region.

**Figure 2.** Scanning electron microscopy (SEM)/backscattering electron image (BEI) cross-sectional micrographs (on the welded direction (WD) plane) of the FSW (**a**) A1/C1, (**b**) A6/C1, and (**c**) A1/C2 joints (both of the C1 and C2 base metals were positioned at the advancing side).

Comparing Figure 2a with Figure 2b, it is apparent that the plastic flow of the Al and Cu base metals within the WZ region is quite different between the A1/C1 and A6/C1 joints. Figure 2a shows the cross-sectional microstructure of the A1/C1 joint for illustration. A large amount of pure Cu matrix and several dispersed large irregular Cu debris (as those encircled in Figure 2a) are mixed with the Al matrix in the WZ of the A1/C1 joint. A clear A1/C1 joint boundary represents that just a mechanical kissing bond is formed between AA1050H-aluminum and C1100-pure Cu at the welding interface of A1/C1 specimens after the FSW process. In addition, a small amount of Al–Cu reacting mixture, which is the result of the intense material plastic flow during FSW, is observed at the top and near the bottom of the WZ (indicated by the triangular marks). The mixture can be considered the metallurgical bonded zone of AA1050H-aluminum and C1100-pure Cu base metals. Unlike the A1/C1 joint, no typical dispersed Cu debris or small Cu particles are observed within the WZ of the A6/C1 joint, as shown in Figure 2b. It is worth noting that a significant wide range of material plastic flow is observed between the Al and Cu matrixes with an obvious onion rings microstructure (indicated by the triangular marks) in the WZ of the A6/C1 joint. It can be recognized that the A6/C1 joint displays much better welding quality, with a larger area fraction of its metallurgical bonded zone and no mechanical kissing bond between the AA6061-T6-aluminum and C1100-pure Cu base metals, compared with the A1/C1 joint. Figure 2c shows the cross-sectional microstructure of the A1/C2 joint for illustration. Comparing Figure 2c with Figure 2b, an Al–Cu metallurgical bonded zone, which is the result of the material plastic flow of AA1050H-aluminum and C2600-brass during FSW, is formed within the WZ of the A1/C2 joint (Figure 2c). However, the Al–Cu metallurgical bonding area fraction of the A1/C2 joint is apparently less than that of the A6/C1 joint. Moreover, a clear joint boundary between AA1050H-aluminum and C2600-brass base metals is observed in the WZ region. Unlike the metallurgical bonding effect within the Al–Cu reacting mixture during FSW, the clear joint boundary observed in the A1/C2 jointis thought of as a mechanical kissing bonded interface, as indicated by the triangular marks in Figure 2c.

Figure 3a,b show the SEM/BEI images and EPMA analysis results for the element distribution maps of Al and Cu within the WZ of the A1/C1 and A6/C1 joints, respectively. As shown in Figure 3a, an apparent joint boundary exists at the welding interface of the AA1050H-aluminum (the dark gray region) and C1100-pure Cu (the light gray region) base metals. In addition, some bright small particles, which are mainly composed of the Cu element, are detected within the AA1050H-aluminum matrix (see Figure 3a). These regions can be recognized as the reacting mixtures resulted from the intense material plastic flow of Al and Cu base metals during FSW, as indicated by the triangular marks in Figure 2a. It can be reasonably deduced that some Al–Cu intermetallic compounds (IMCs) are formed within the Al–Cu reacting mixture region, and the phase composition of Al–Cu IMCs will be identified by following X-ray diffraction analysis. Figure 3b shows the SEM/BEI images and EPMA analysis results in the WZ of the A6/C1 joint. The dark gray and light gray regions correspond to the AA6061-T6-aluminum and C1100-pure Cu base metals, respectively. As shown in Figure 3b, the FSW A6/C1 joint displays a significant material plastic flow and reacting mixtures within the WZ compared with the A1/C1 joint. Therefore, comparing Figure 3b with Figure 2b demonstrates that the metallurgical bonded zone consists of severe stirring plastic flow between AA6061-T6-aluminum and C1100-pure Cu base metals with a large amount of reacting mixtures of dispersed Al–Cu IMCs particles. The distribution of Al–Cu IMCs particles and microstructural morphologies of the A6/C1 joint are quite different from that of the A1/C1 joint. Figure 3c shows the SEM/BEI image and a line scanning analysis result with the distribution of Al and Cu elements in the WZ of the FSW A1/C2 joint. The dwell time via the SEM/BEI EDS mapping to collect the data for the Al and Cu elements is about 10 minutes. The dark gray and light gray regions correspond to the AA1050H-aluminum and C2600-brass base metals, respectively. However, unlike the reacting mixtures represented in Figure 3a,b, an obvious interfacial layer with mixing Al and Cu elements is observed between the base metals, and it can be recognized as the Al–Cu IMCs thick layer that was formed on the interface of AA1050H-aluminum and C2600-brass during the FSW process. Generally, the FSW dissimilar

joints are failed at the WZ or along the interface between the base metals after the mechanical tests. Therefore, the tensile strength, measuring data fluctuation, failure behaviors, and welding reliability of the FSW A1/C1, A6/C1, and A1/C2 joints will be strongly affected by the metallurgical bonding effect, the particle distribution, and the morphologies of reacted Al–Cu IMCs within the WZ. As a result of the above-mentioned observation and analysis of microstructures, the correlation between the microstructural features, tensile strength, and reliability of these FSW Al–Cu dissimilar joints will be discussed in the following sections.

**Figure 3.** The electron probe micro-analyzer (EPMA) analysis for Al and Cu elements distribution within the WZ of (**a**) the FSW A1/C1 joint, and (**b**) the FSW A6/C1 joint. (**c**) The SEM/BEI line-scanning and Al–Cu elements mapping at the Al/Cu interface of the FSW A1/C2 joint.

The X-ray diffraction patterns obtained from the WZ region of various FSW dissimilar joints are given in Figure 4. Figure 4a,b show the XRD patterns of the A1/C1 and A6/C1 joints, respectively. In addition to the strong diffraction peaks of major α-Al (main peaks detected at 2θ = 38.47◦, 44.74◦, 65.13◦, and 78.23◦, JCPDS 04-0787) and Cu (main peaks detected at 2θ = 43.30◦, 50.43◦, and 74.13◦, JCPDS 04-0836) base metals, some relatively weak peaks are also detected within the WZ region of these Al–Cu dissimilar joints. According to the Al–Cu equilibrium phase diagram, several Al–Cu IMCs, such as AlCu (η2), Al2Cu (θ), Al2Cu3 (ε), Al3Cu4 (ζ2), and Al4Cu9 (γ) can be found in the Al–Cu binary alloy system. However, it is recognized that the Al-rich Al2Cu phase and Cu-rich Al4Cu9 phase are the two major IMCs formed during the Al–Cu metallurgical reaction [23,26,41]. Therefore, these weak diffraction peaks obtained in the WZ region of A1/C1 joints are then identified as the IMCs of Al2Cu (main peaks detected at 2θ = 37.87◦, 42.59◦, 47.33◦, and 47.81◦, JCPDS 25-0012) and Al4Cu9 (the

diffraction peak detected at 2θ = 44.12◦, JCPDS 24-0003). Comparing Figure 4b with Figure 4a, we can see that the diffraction peaks of the Al2Cu and Al4Cu9 IMCs for the A6/C1 joint are significantly sharper and much more obvious than those of the A1/C1 joint. The differencein peak intensity means that more Al2Cu and Al4Cu9 IMCs are obtained within the WZ region of the A6/C1 joint. Referring to the SEM/BEI images and the element distribution maps obtained by EPMA, the XRD analysis result of the FSW A6/C1 joint (see Figure 4b cf. Figures 2b and 3b) corresponds to the microstructural feature, which displays a more widely Al–Cu reacting mixture distribution with a larger area fraction of the metallurgical bonded zone than the FSW A1/C1 joint (see Figure 4a cf. Figures 2a and 3a). Figure 4c is the XRD pattern of the FSW A1/C2 dissimilar joint, which included strong diffraction peaks of α-Al and Cu0.64Zn0.36 (α-brass, main peaks detected at 2θ = 42.33◦, 49.28◦, and 72.25◦, JCPDS 50-1333) base metals. The result illustrates that the WZ region of the A1/C2 joint also consists mainly of Al and Cu (i.e., α-brass) base metals with a fair amount of Al2Cu and Al4Cu9 IMCs. In addition, a γ-Cu5Zn8 compound (diffraction peaks at 2θ = 43.30◦ and 62.88◦, JCPDS25-1228) is also observed in the A1/C2 joint, as indicated by the triangular marks in Figure 4c. As a result, it can be recognized that the γ-Cu5Zn8 compound is another minor reaction product accompanied by the formation of Al2Cu and Al4Cu9 IMCs for the A1/C2 joint during FSW. Since less Al–Cu reacting mixture and a clear mechanical kissing bonded boundary of the A1/C2 joint are observed (see Figure 2c), it implies that the weldability of A1/C2 joint is worse than that of the A1/C1 and A6/C1 joints.

**Figure 4.** X-ray diffraction patterns of the friction stir welded (**a**) A1/C1, (**b**) A6/C1, and (**c**) A1/C2 dissimilar joints within the welding zone (WZ).

Figure 5 displays the distribution of micro-Vickers hardness (HV) profiles measured within the WZ regions of FSW A1/C1, A6/C1, and A1/C2 dissimilar joints. The hardness data recorded on the transverse cross-sections of FSW-joined specimens, and the Vickers indenter testing positions are located at 1.5 mm depth from the surface. The micro-Vickers hardness test reveals the average values of AA1050H, AA6061-T6 aluminum alloys, commercial pure copper (C1100), and brass (C2600) base metals to be about HV32.6 ± 4.7, HV61.2 ± 5.3, HV62.1 ± 3.0, and HV103.2 ± 4.5, respectively. According to the profiles represented in Figure 5, the hardness of FSW Al–Cu dissimilar joints significantly increases in the WZ region relative to both Al and Cu base metals. The data fluctuation at the transverse direction is the resulted of the heterogeneous microstructure of the WZ. Referring to the cross-sectional microstructures as mentioned in Figure 2, the distribution range of the increased

hardness profiles (in the position of about ± 4 mm from the center, as shown in Figure 5) is almost the same as that of the Al–Cu metallurgical bonded region with obvious metallic plastic flow. Therefore, it can be recognized that the increased hardness profiles of the A6/C1 and A1/C2 joints can be attributed to the apparent Al–Cu reacting mixture within the WZ region (see Figure 5 cf. Figure 2b,c). In addition, the high hardness values of the A6/C1 and A1/C2 joints are also the result of the significant strain-hardening effect of intense materials plastic deformation as well as the particle strengthening effect with the formation and uniform redistribution of Al2Cu, Al4Cu9, and a fair amount of γ-Cu5Zn8 IMCs (such as those observed by the XRD analysis in Figure 4) during the FSW process. However, compared with the A6/C1 and A1/C2 joints, the hardness profile for the A1/C1 joint is lower as a result of an insufficient metallurgical bonding reaction and a lesser amount of Al–Cu reacting IMCs mixture (see Figure 2a cf. Figure 4).

**Figure 5.** The variation of microhardness of various FSW Al–Cu dissimilar joints. The indentations are made with a spacing of 0.2 mm on the WD plane (AS: advancing side; RS: retreating side).
