*3.2. The Element Diffusion at the NiTi/CuSn10 Interface*

The samples observed with SEM were shown in Figure 6 (500×). The widths of interfacial regions of five samples were measured and the results were indicated in Figure 6. The maximum was 255.32 ± 27.32 µm while the minimum was 120.13 ± 18.64 µm.

**Figure 6.** Interfacial SEM images of sample (**a**) No.1; (**b**) No.2; (**c**) No.3; (**d**) No.4; (**e**) No.5.

It can be seen that the width of the interfacial region tended to increase with the increase of energy density. This was mainly because as the energy density increased, the molten pool movement became more violent and thus promoting the mixing of these two alloys. However, this promoting effect had a certain range. When the energy density was higher than a certain value, it can be seen that interfacial region thickness did not increase and tended to be stable. When it comes to a single parameter, it can be seen when comparing samples No.1 and No.3 that when the laser power was constant, the higher the scanning speed, the lower interfacial region thickness was. This was because the diffusion time was shorter when the scanning speed was higher [5]. It can also be seen when comparing samples No.2 and No.4 that when the scanning speed was constant, the larger the laser power and the higher the interfacial region thickness was. This was because the larger the laser power was, the larger the energy input was; thus the molten pool was deeper. Furthermore, the element distribution at the interface was analyzed. The main elements Ni, Ti, Cu, and Sn of the two alloys were taken as the analysis objects. The chemical composition was measured at a point every 2.88 µm along the building direction and 50 points were selected for each sample. The distribution results of interface elements were obtained as shown in Figure 6. Due to the small content of Sn, the change trend was not obvious, but it can still be seen that the content decreased when crossing the interface. The content of Ni and Ti decreased continuously while the content of Cu increased continuously when crossing the interface, indicating that the element mixing at the interfacial region was relatively sufficient.

## *3.3. The Microhardness at the NiTi/CuSn10 Interface*

The microhardness of each sample from CuSn10 alloy to NiTi alloy was measured at 50 µm intervals, and the results were shown in Figure 7. It can be seen that the microhardness was almost constant at first. Then it suddenly became larger at the interface, next it suddenly became smaller, and finally it became constant again. According to the change trend of microhardness, the area near the interface can be divided into CuSn10 zone, interfacial region, and NiTi zone. It is worth noting that the microhardness of the interfacial region was much higher than that of both sides. Sing et al. [27] found intermetallic compounds at the interface, which can make microhardness become higher. Zhou et al. [22] found that less brittle and hard Cu and Ni rich intermetallics (i.e., Cu3Ti, Ni3Ti) and more brittle Ti-rich intermetallics (i.e., Ti2Cu, Ti2Ni) could be formed in the process of dissimilar laser welding of NiTi and copper. This indicated that intermetallic compounds were possible to form between Cu alloy and NiTi alloy. Therefore, it can be considered that some kind of intermetallic compound was formed at the interface. In order to test the existence of interfacial intermetallic compound, XRD analysis was conducted on the interfaces of the five samples, and the experimental results were shown in Figure 8.

**Figure 7.** Microhardness change at the interface.

**Figure 8.** XRD pattern of the NiTi/CuSn10 interfaces.

First, the related phases of NiTi alloy and CuSn10 alloy were determined by phase identifying. Cu(Sn) and Cu41Sn11 were phases that often appeared in SLM CuSn10 alloy while B2 and B190 were two phases of NiTi alloy. After this, it was found that a new phase appeared, which was found to be Ni4Ti<sup>3</sup> after searching and comparing. This was because the scanning speed and laser power used at the interface were not high so the heating and cooling rate of the molten pool was low, which was conducive to the formation of Ni4Ti<sup>3</sup> [28]. Ni4Ti<sup>3</sup> is a very fine phase, which can produce dispersion strengthening and increase the strength of the sample interface [29]. Therefore, the formation of Ni4Ti<sup>3</sup> intermetallic compound played a key role in the increase of microhardness at the interface. At the same time, it can be inferred that the interfacial bonding performance had been improved.

#### *3.4. Bonding Strength Test of NiTi/CuSn10 Interface*

The interfacial bonding strength between NiTi and CuSn10 has a great influence on the properties of the whole part. In order to study the bonding strength of NiTi/CuSn10, tensile tests of dissimilar materials samples were conducted, and the results were shown in Figure 9a.

The strengths of NiTi and CuSn10 single alloy formed vertically by SLM had been optimized and were used as the control groups with values of 508.6 ± 42.2 MPa and 422.9 ± 10.8 MPa, respectively. All the tensile parts broke at the interface of the two materials. It can be seen that sample No.1 had the worst interfacial bonding strength (156.0 ± 7.0 MPa), equivalent to 37% of that of CuSn10 single alloy. Sample No.3 had the best interfacial bonding strength (309.7 ± 34.4 MPa), equivalent to 75% of that of CuSn10 single alloy. Based on previous results, although there were vertical microcracks at the interface of process parameters corresponding to sample No.3, the interfacial bonding properties of sample No.3 was still better than that of other samples with no observed defects. This indicated that the vertical microcracks did not have a significant impact on interfacial bonding properties. The relationship between interfacial bonding strength and interfacial region thickness was shown in Figure 9b. It can be obtained from Figure 9b that the interfacial bonding strength was positively correlated with interfacial region thickness. This was because the higher the interfacial region thickness was, the more solid metallurgical bonding could be formed at the interface so that the interfacial bonding strength would be higher. Columnar grains at the interface and grain refinement in the interfacial region played a strengthening role. Based on the phase analysis, the improvement of interfacial bonding strength was also related to the formation of fine Ni4Ti<sup>3</sup> strengthening phase.

**Figure 9.** (**a**) The interfacial bonding strengths of different samples compared with the tensile strength of CuSn10 and NiTi single alloy (**b**) relationship between interfacial bonding strength and interfacial region thickness.

Low magnification and high magnification SEM images were taken for each fracture, as shown in Figure 10.

At low magnification (200×), it can be seen that the fracture was very flat, and some unmelted metal powder can be seen occasionally, which indicated a brittle fracture as shown in Figure 10a. The whole fracture was divided into many small planes. At high magnification (2000×), some holes due to the gas not escaping and gullies between planes due to the existence of vertical cracks in these samples can be seen. On account of brittle fracture, these samples cannot achieve strength comparable to that of a single metal material. However, it is worth noting that a few dimples can be found in the fracture of sample No.3 at high magnification (5000×); that is, the sample had certain ductility. This indicated that a sample could have good interfacial bonding properties under the strengthening of columnar grain at the interface, grain refinement in the interfacial region and Ni4Ti<sup>3</sup> with appropriate content.

**Figure 10.** SEM images of fracture of sample (**a**) No.1 (200×); (**b**) No.1 (2000×); (**c**) No.2 (200×); (**d**) No.2 (2000×); (**e**) No.3 (200×); (**f**) No.3 (5000×); (**g**) No.4 (200×); (**h**) No.4 (2000×); (**i**) No.5 (200×); (**j**) No.5 (2000×).
