Improvements and algorithmical considerations on a recent three-dimensional model describing stress-induced solid phase transformations

During mechanical loading-unloading cycles shape-memory alloys (SMA) are able to undergo large deformations without showing residual strains (pseudoelasticity) or recovering them through thermal cycles (shape memory effect). Motivated by stress-induced solid phase transformations, these unique behaviours induce the SMA exploitation in innovative and commercially valuable applications, stimulating, consequently, the interest in the development of constitutive models. Also if many models are now available in the literature, effective three-dimensional proposals are still few and limited in several aspects. In this paper, a three-dimensional thermomechanical model recently proposed by Souza et al. (European Journal of Mechanics-A/Solids, 1998; 17:789-806.) is taken into consideration; such a model is of particular interest for its effectiveness and flexibility, but it also shows some limitations and missing links in the algorithmical counterparts. This work discusses some improvements to the original model as well as the development and the implementation of a robust integration algorithm to be adopted in a numerical scheme, such as a finite-element framework. Copyright (C) 2002 John Wiley Sons, Ltd.