Unique functional properties such as the low stiffness, superelasticity, and biocompatibility of nickel-titanium shape-memory alloys provide many applications for such materials. Selective laser melting of NiTi enables low-cost customization of devices and the manufacturing of highly complex geometries without subsequent machining. However, the technology requires optimization of process parameters in order to guarantee high mass density and to avoid deterioration of functional properties. In this work, the melt pool geometry, surface morphology, formation mode, and thermal behavior were studied. Multiple combinations of laser power and scanning speed were used for single-track preparation from pre-alloyed NiTi powder on a nitinol substrate. The experimental results show the influence of laser power and scanning speed on the depth, width, and depth-to-width aspect ratio. Additionally, a transient 3D FE model was employed to predict thermal behavior in the melt pool for different regimes. In this paper, the coefficients for a volumetric double-ellipsoid heat source were calibrated with bound optimization by a quadratic approximation algorithm, the design of experiments technique, and experimentally obtained data. The results of the simulation reveal the necessary conditions of transition from conduction to keyhole mode welding. Finally, by combining experimental and FE modeling results, the optimal SLM process parameters were evaluated as P = 77 W, V = 400 mm/s, h = 70 μm, and t = 50 μm, without printing of 3D samples.
Keywords: SLM; nickel–titanium; process parameter optimization; shape-memory alloys (SMAs); single track.