*3.2. Tensile Failure Strength of FSW Al–Cu Dissimilar Joints*

The tensile strength of the FSW dissimilar Al–Cu joint is evaluated as a criterion to the joint performance. Table 2 lists a comparison of the tensile strength and elongation of base metals and the FSW A1/C1, A6/C1, and A1/C2 dissimilar joints. Since a sound FSW A6/C2 joint without welding cavity defects in the WZ region is not successfully obtained, therefore, the tensile properties of A6/C2 joints are too low to be used in this study. In addition, the data fluctuation of A6/C2 joints is dispersed to show a valid statistical significance for the following reliability analysis in this study. It can be seen that a maximum tensile strength of about 212.7 MPa and the highest elongation of about 22.3% are obtained for the FSW A6/C1 joint. Figure 6 shows the representative fractured specimens of FSW A1/C1, A6/C1, and A1/C2 dissimilar joints after tensile tests. The Al–Cu dissimilar joints failed with distinguished fracture characteristics, and the fracture location can be clearly observed in these photos. The fracture of FSW A1/C1 joints is usually located at the WZ region, as shown in Figure 6a. Since the average tensile strength of A1/C1 joints is about 85% of the AA1050H aluminum alloy (121.8 ± 2.3 MPa) according to earlier measurements for base metals, the fracture characteristics that are displayed in Figure 6a and the decrease in the tensile strength of the A1/C1 joints should be the result of the inhomogeneous welding microstructure, which features a mechanical kissing bond at the interface between AA1050H-aluminum and C1100-pure Cu (see Figure 2a cf. Figure 6a).

Comparing Figure 6b with Figure 6a, it is worth noting that the fracture region of the FSW A6/C1 joints is distant from the WZ and nearly located at the heat-affected zone (HAZ) with an obvious necking deformation of the AA6061-T6 Al base metal. As a result, both the tensile strength and the elongation of the A6/C1 joints are significantly higher than that of the A1/C1 joints (see Table 2), and

the fracture characteristic without failure at the WZ of the FSW A6/C1 joints, which is represented in Figure 6b, can be regarded as an illustration of better joint quality. Referring to the SEM/BEI cross-sectional images and EPMA analysis results shown in Figures 2b and 3b, it is noted that the much better tensile strength of the A6/C1 joints is achieved from their good joint quality with a larger area fraction of the metallurgical bonded zone and a firmly material metallurgical joining between the AA6061-T6 Al alloy and the C1100 commercial pure copper through the FSW process. The intense material plastic flow with obvious onion rings (see Figure 2b) and uniformly dispersive refined hard IMCs particles in the Al–Cu reacting mixture enhance the metallurgical bonding effect within the WZ. Meanwhile, the movement of dislocations will be effectively impeded by the tangle of high dislocation density and the refined hard IMCs particles during plastic deformation. Therefore, it can be recognized that the FSW A6/C1 dissimilar joint with increased hardness profiles and much higher tensile strength is obtained due to the strong metallurgical bonding effect.


**Table 2.** Tensile properties of base metals and various FSW Al–Cu dissimilar joints.

\* Each value was the average of a least three tests; † Each value was the average of 20 tests (*n* = 20).

**Figure 6.** Macrographs of fractured FSW (**a**) A1/C1, (**b**) A6/C1, and (**c**) A1/C2 dissimilar joints after tensile tests.

Among these Al–Cu dissimilar joints, the FSW A1/C2 joint apparently exhibits the worst tensile strength and displays much lower elongation (see Table 2), with an obvious brittle fracture at the WZ region along the joining interface of AA1050H-aluminum and C2600-brass base metals, as illustrated in Figure 6c. The notable degradation in the tensile strength of the A1/C2 joints should be closely correlated with the structural inhomogeneity. Referring to Figures 2c and 3c cf. Figure 6c, a large area fraction of the mechanical kissing bond with a continuous thick IMCs layer was significantly formed at the Al/brass interface within the WZ of the A1/C2 joints. It demonstrates that the reduction of effective metallurgical bonding and the formation of a continuous thick IMCs layer yielded a weak interfacial A1/C2 joint. As a result, the detrimental decohesive ruptures notably occurred at the weak mechanical kissing bonded boundary of AA1050H-aluminum and C2600-brass before other structural defects caused the failures of the A1/C2 joint throughout the WZ.

According to the above-mentioned examinations, FSW A1/C1, A6/C1, and A1/C2 dissimilar joints display quite different welding microstructures, interfacial joining characteristics, hardness profiles, and phase compositions within the WZ regions. Since the tensile strength of FSW Al–Cu dissimilar joints should be closely affected by the variation of welding qualities, the statistical analysis by the Weibull model on the data variance of tensile tests made it possible to assess the microstructural features in determining the failure behaviors and reliability of FSW Al–Cu dissimilar joints. The Weibull statistical analysis detail on the tensile mechanical properties and failures related to the welding qualities of FSW Al-Cu joints will be discussed in the following section.
