Thermal processing of polycrystalline NiTi shape memory alloys

The objective of this study is to examine the effect of heat treatment on polycrystalline Ti-50.9 at.% Ni in hot-rolled and cold-drawn states. In particular, we examine microstructure, transformation temperatures as well as mechanical behavior in terms of both uniaxial monotonic testing and instrumented Vickers micro-indentation. The results constitute a fundamental understanding of the effect of heat treatment on thermal/stress-induced martensite and resistance to plastic flow in NiTi, all of which are critical for optimizing the mechanical properties. The high temperature of the hot-rolling process caused recrystallization, recovery, and hindered precipitate formation, essentially solutionizing the NiTi. The subsequent cold-drawing-induced a high density of dislocations and martensite. Heat treatments were carried out on hot-rolled, as well as, hot-rolled then cold-drawn materials at various temperatures for 1.5 h. Transmission Electron Microscopy observations revealed that Ti3Ni4 precipitates progressively increased in size and changed their interface with the matrix from being coherent to incoherent with increasing heat treatment temperature. Accompanying the changes in precipitate size and interface coherency, transformation temperatures were observed to systematically shift, leading to the occurrence of the R-phase and multiple-stage transformations. Room temperature stress-strain tests illustrated a variety of mechanical responses for the various heat treatments, from pseudoelasticity to shape memory. The changes in stress-strain behavior are interpreted in terms of shifts in the primary martensite transformation temperatures, rather then the occurrence of the R-phase transformation. The results confirm that Ti3Ni4 precipitates can be used to elicit a desired isothermal stress-strain behavior in polycrystalline NiTi. Instrumented micro-indention tests revealed that Martens (Universal) Hardness values are more dependent on the resistance to dislocation motion than measured uniaxial pseudoelastic or shape memory response. Based on comparison of hardness and the stress required to induce martensite, it is shown that the resistance to dislocation motion and the ease of the stress-induced martensite transformation cannot be simultaneously maximized, although an optimal combination should exist. Measuring indentation depth before and after heating more distinctly confirmed shape memory or pseudoelastic behavior.