Abstract NiTi shape memory alloys (SMAs) are widely used in applications ranging from medical devices to aerospace and automotive structural actuators, yet their performance is often limited by large hysteresis and highly nonlinear pseudoelasticity, which can reduce efficiency and hinder precise control. This study presents a computational investigation of two microstructural engineering strategies aimed at achieving a more linearized and predictable stress–strain response. The first strategy, termed soft confinement , introduces concentration modulations (CMs) produced by dissolving Ni 4 Ti 3 nanoprecipitates, generating smooth spatial gradients in the martensitic start temperature (Ms) through controlled variations in the Ni concentration field. The second strategy, hard confinement , employs amorphous–crystalline composite microstructures in which a non-transforming amorphous phase serves as a robust physical barrier to martensitic progression. Phase-field simulations are used to systematically evaluate the effectiveness of each approach. The results demonstrate that both soft and hard nanoconfinement strategies successfully regulate the otherwise avalanche-like martensitic transformation (MT), enabling controlled strain release and yielding a more linear, stable, and tunable superelastic response.
Sriram et al. (2026) studied this question.