Modeling and Prediction of Training Effects in Elastocaloric Materials

Training phenomena play a crucial role in the performance of elastocaloric materials, as repeated cycling alters both their mechanical response and functional efficiency. The evolution of transformation stresses, hysteresis, and remanent strains is not random but results from a complex interaction between external loading conditions and intrinsic material characteristics. While parameters such as strain rate or temperature determine the dynamics of individual cycles, alloy composition, processing route, and heat treatment shape the underlying microstructure, thereby influencing how the material adapts during training.

In this work, we focus on a representative NiTi-based alloy to establish a modeling framework for describing and predicting training behavior. Experimental data is used to extract the progressive changes in key transformation properties, which then serve as input for a thermodynamics-based model. By reproducing the cyclic evolution of stresses, strains, and hysteresis, the model provides a pathway to connect microscopic mechanisms with macroscopic material response.

Beyond reproducing observed behavior, the approach also targets the prediction of the material response. With alloy-specific parameters and defined training conditions as input, the framework aims to determine the mechanical response after tens or even hundreds of cycles. In doing so, it contributes to a deeper understanding of stability and degradation phenomena and supports the design of elastocaloric materials with improved long-term reliability.