Predicting Localization in Superelastic Nitinol Implants During Monotonic Loading

Computational modeling is commonly used to predict nitinol implant mechanics and the critical influence of phase transformation throughout the device loading history. Conventional superelastic models implement phase transformation using hardening transformation surfaces with strictly positive slopes. However, literature studies have provided evidence of softening during phase transformation when localization phenomena are stabilized, for example using laminated specimens (e.g., Hallai and Kyriakides 2013, Greenly Kyriakides and Tsimpoukis 2024). Here, we investigate the use of Drucker-Prager plasticity with softening yield surfaces to simulate localization during monotonic loading. In brief, finite element simulations of round wire Z-specimens under monotonic extension are performed in ABAQUS using 1) a standard superelastic model and 2) Drucker-Prager plasticity with matched elastic moduli, mean plateau stresses, and tension-compression asymmetry to (1) but a negative plateau slope. Simulation predictions are then compared with force-displacement measurements and digital image correlation (DIC) strain data. Results show better qualitative agreement between simulations and experiments using the Drucker-Prager model, particularly in predicting the force-displacement response and the shape of phase transformation boundaries. The softening simulations cannot realistically predict the mechanics during unloading or cyclic conditions due to the reliance on plasticity modeling. However, the approach can predict peak and full-field strains during monotonic stages of device history, including shape setting and crimping. Softening simulations can also provide insight into the location and spatial distribution of localization bands at critical regions where fractures are anticipated to initiate. Calculations for converting conventional superelastic properties to representative Drucker-Prager parameters are provided to support future investigations.