On the Modeling of the Thermo-Mechanical Responses of Four Different Classes of NiTi-Based Shape Memory Materials Using a General Multi-Mechanism Framework

© 2014 Elsevier Ltd. The properties of a shape memory alloy (SMA) have been shown to be highly dependent on the chemical composition and thermo-mechanical processing applied to the material. These differences dictate the degree of superelasticity, pseudoplasticity, shape memory effect, and evolution under mechanical/thermal loading cycles, that is observed in the material. Understanding and utilizing these unique phenomena has become essential in many engineering applications. It is, therefore, important to provide two key ingredients in any SMA constitutive model; (i) a sufficiently comprehensive scope in the mathematical formulation to handle different classes of SMA materials; and (ii) a general model parameterization derived from fundamental tests that can be used for a specific SMA as intended for use in a given application. The present work is aimed at a detailed investigation of the interaction aspects between the above items (i) and (ii) in the context of using a recent three-dimensional, multimechanism-based SMA framework to model the experimentally measured responses of four different classes of SMA materials: (a) a commercial superelastic NiTi, (b) a powder metallurgically-processed NiTi-based SMA material, (c) a commercial Ni49.9Ti50.1 actuation material, and (d) a high-temperature Ni50.3Ti29.7Hf20 alloy. To facilitate the parameterization task, the model parameters are classified into two groups, i.e., (1) fixed parameters that are designed to capture the non-linear, hysteretic response under any thermo-mechanical loading condition, and (2) a set of functionally dependent material parameters which account for a number of refinements including asymmetry in tension and compression responses, temperature- and stress-state dependencies, etc. The results of the work showed that the complexity of the characterization is dependent on the SMA feature exploited by the specific application intended, which in turn dictates the amount and type of test data required to accurately predict a given application response.