Nonlocal modeling and analysis of spatiotemporal patterns in non-isothermal phase transformation of NiTi strips

The spatiotemporal evolution of macro-domain patterns in a NiTi thin strip during stress-induced nonisothermal phase transformation (PT) are investigated through 2D FEM simulations using nonconvex nonlocal continuum model and 1D theoretical analysis. The FEM simulations shows that, with the increase of applied stretching rate (nominal strain rate), the way of PT in a thin strip (30 mm x 2.5 m m x 0.5 mm) changes from the sequential nucleation and growth mode in the low rates range (<= 0.03/s) to the emergence of periodic patterns in the medium rates range (0.03/s similar to 3.0/s) and eventually to stable homogeneous deformation mode in the high rates range (>= 3.0/s). The domain spacing is governed by a power-law scaling to the strain rate, where the exponent changes from -1/2 (low rates range) to -1/6 (medium rates range). The 1D theoretical analysis further demonstrates that the rate-dependent patterns are essentially governed by a nondimensional internal material parameter (A) over bar (0) and a nondimensional external driving parameter (k) over bar. (A) over bar (0) represents the ratio of adiabatic hardening to isothermal softening in PT of the material, while (k) over bar controls the time scale competition of heat release and heat conduction, through which the thermodynamic condition of PT changes from near isothermal ((k) over bar -> infinity) to near adiabatic ((k) over bar -> 0) with the increase of the strain rate. It is shown that, as long as the material has the properties of isothermal softening and adiabatic hardening (i.e., (A) over bar (0) > 1), the conventional nucleation-growth paradigm of PT will eventually break down for small (k) over bar (or large stain rate). This is due to the fact that the increase of the energy barrier caused by the thermal and the interfacial effects can convexify the energy landscape of the system near the adiabatic condition and suppress the nucleation events at the macroscopic level. (C) 2020 Elsevier Ltd. All rights reserved.