Objective An intelligent material, NiTi shape memory alloy is widely used in mechatronics, aerospace, medical devices, and other fields due to its excellent properties, e.g., biocompatibility, corrosion resistance, shape memory effect, and super-elasticity. Successful adoption of NiTi depends on its intrinsic characteristics and applications bring about by connection with other materials. Copper has high thermal and electrical conductivity, ductility, and corrosion resistance, which plays an important role in electrical, pipeline engineering, aerospace, and other fields. Recently, the dissimilar joining of NiTi/Cu to electrothermal actuator has become a concern in this field because dissimilar joint of NiTi/Cu cannot satisfy shape memory effect requirements and ensure the high electrical conductivity of components. Laser welding is particularly suited to dissimilar NiTi joining compared with other connection modes; however, the weld mechanical properties decrease significantly compared to those of the base metal due to the brittle Ni-Ti intermetallics in the weld; thus, requirements can not be satisfied. To solve this problem, studies have investigated dissimilar welding of NiTi alloys by adjusting the welding parameters. In this study, we take NiTi alloy and copper wire as research targets and study the microstructure variation rules and the properties of NiTi/Cu laser-welded joints by changing laser offsets, which provides potential guidance for the application of the dissimilar welding of NiTi alloys and copper.

Methods Laser welding of NiTi (Ni with an atomic number fraction of 50.2%) wire with 400-μm diameter and copper (Cu with an atomic number fraction of 99.9%) wire is performed using a pulsed laser. First, the wires are cleaned using acetone, ethanol, and deionized water to remove oil stains and contaminations prior to laser welding. Laser welding is conducted using a Miyachi Unitek LW50A pulsed Nd∶YAG laser (peak power is 0.9 kW, laser wavelength is 1.064 μm). During welding, pure argon is used as a shielding gas. Various laser offsets are obtained using different laser beam positions: 100 μm on NiTi side, 50 μm on NiTi side, centerline, 50 μm on copper side, and 100 μm on copper side (Fig.1). Cross sections of the welded joints are mounted in epoxy and grinded with sandpaper (up to number 1200) and then polished successively to 2.5, 1.0, and 0.5 μm using diamond sprays. This is followed by etching with Kroll reagent for 1 min. The microstructures are observed using an Olympus BX51M optical microscope and a Zeiss Ultra Plus field emission scanning electron microscope equipped with EDX to analyze the compositions. A Clemex CMT automated micro-Vickers hardness tester is used to make a series of 50-g indents across the fusion zone, 50 μm apart with a dwelling time of 10 s. The joints are using an Instron 5548 micro tester at a strain rate of 3×10 ^-4 s ^-1.

Results and Discussions Laser offset is found to play a significant role in the microstructure due to the difference in mixing patterns and composition distributions. The results demonstrate that weld width decreased when moving the laser position from NiTi to Cu (Fig. 5), and the uniform distribution of the mixing pattern inside the weld zone changes to the local segregation [Figs. 7(a)--7(d)]. Welds with offsetting of 50 μm on the NiTi and centerline exhibited dendritic solidification microstructures, and welds with offsetting of 50 and 100 μm on Cu comprise a mixture of dendritic, cellular, and lamellar microstructures (Fig. 6). The hardness of the weld seam is reduced by with shifting the laser position from the NiTi side to the Cu side. When the laser offset is on the Cu side, local high hardness values appeare in the NiTi-rich region [Figs. 7(e)--7(h)]. The 100-μm Cu offset joint fracture in the weld zone during tensile loading due to the cracks insight, and the strength decreased significantly compared to the Cu base metal (Fig.8).

Conclusions The results demonstrate that the proportion of NiTi alloys in the molten pool decreases gradually, and the decrement of NiTi alloys is greater than the increment of Cu when moving the laser position from NiTi to Cu, which results in reduced weld width. When the laser offset is on the Cu side, the increase of copper makes the weld zone have a very fast cooling rate and solidify quickly, which leads to the liquid copper and liquid NiTi in the molten pool not being fully mixed and forming element segregation. The welds with the offsetting of 50 μm on the NiTi and centerline exhibit homogeneous dendritic solidification microstructures that are also Ni_xTi_yCu_z intermetallics. Welds with offsetting of 50 and 100 μm on Cu comprise a mixture of dendritic, cellular, and lamellar microstructures composed of NiTi intermetallics, CuTi intermetallics, Ni_xTi_yCu_z intermetallics, and a copper solid solution. The hardness of the weld seam decreases by shifting the laser position from the NiTi side to the Cu side. When the laser offset is 50 μm on the NiTi side and the centerline, the hardness distribution in the weld zone is uniform, and average hardness is approximately 520 and 340 HV, respectively. When the laser offset is on Cu side, the hardness in the weld is very uneven, and local high hardness values appear in the NiTi-rich region. By changing the laser offset from 50 μm on the NiTi side to 50 μm on the Cu side, the NiTi/Cu dissimilar welded joint strength is close to that of the copper base metal, which is primarily due to preferential failure of the softer copper base metal in tension. The 100-μm Cu offset joint fractures in the weld zone during tensile loading due to the cracks insight, and the strength decreases significantly compared to Cu base metal.