Finite Element Modeling of Springback in NiTi Medical Devices Using Phase-Specific Stress Relaxation

Thursday, May 7, 2026: 10:50 AM
Ms. Francesca Rossi , Politecnico di Torino, Turin, Piedmont, Italy
Dr. Dario Carbonaro , Politecnico di Torino, Turin, Piedmont, Italy
Mr. Carlo Guala , Corcym, Saluggia (VC), Piedmont, Italy
Mr. Giovanni Giordano , Corcym, Saluggia (VC), Piedmont, Italy
Mr. Francesco Valle , Corcym, Saluggia (VC), Piedmont, Italy
Mr. Marco Bussone , Corcym, Saluggia (VC), Piedmont, Italy
Prof. Claudio Chiastra , Politecnico di Torino, Turin, Piedmont, Italy
Mr. Cristian Faggian , Corcym, Saluggia (VC), Piedmont, Italy
Prof. Alberto Audenino , Politecnico di Torino, Turin, Piedmont, Italy
Nickel–titanium (NiTi) is widely used in the manufacturing of medical devices due to its superelastic and shape-memory properties. These devices are subject to springback, a phenomenon caused by residual stresses introduced during the shape-setting process, where NiTi tubes are heat-treated within molds. Upon removal, these stresses cause deviations from the intended geometry, thereby reducing manufacturing accuracy. To address this issue, a finite element (FE)-based computational approach was developed to predict and mitigate springback in NiTi devices, offering a valuable tool to enhance design precision and minimize trial-and-error adjustments during the development process. The approach was demonstrated on annuloplasty ring prototypes, using four NiTi tubes (laser-cut and non-laser-cut) in two sizes, shape-set in circular molds, heat-treated, and with their geometries measured post-process. A FE model was implemented in the implicit finite element code Abaqus Standard (Dassault Systemes Simulia Corp.), incorporating super-elastic NiTi material properties and a phase-dependent stress relaxation coefficient. Calibration was performed using experimental data from a full NiTi tube (size 1), resulting in a relaxation coefficient of 25% (ra) for austenitic-phase stresses, while all the other stresses were set to zero. The model achieved a maximum prediction error of 0.62%. Validation against experimental results for the remaining tube types and sizes confirmed the model’s predictive accuracy across different NiTi tube designs (maximum prediction error of 1.57%). Overall, the proposed approach provides an effective tool to predict springback, supporting more precise and efficient development of NiTi annuloplasty rings.
See more of: Modeling and simulation II
See more of: Technical Program