Influence of evaluation technique on the fatigue assessment of Nitinol in medical applications

The fatigue behavior of superelastic nitinol in medical devices is strongly dependent on the stress and strain states experienced over the device’s full life cycle. Finite element analysis (FEA) enables quantification of fatigue-relevant strains in components such as heart-valve frames under in-vivo or in-vitro loading and allows evaluation of strain–displacement relationships in surrogate benchtop specimens. Surrogate data are commonly used to define a fatigue limit, which serves as a benchmark for assessing device fatigue performance.

Accurate prediction of fatigue life requires careful interpretation of simulated loads. In-vivo, superelastic nitinol devices can experience multiaxial and nonproportional loading, including strain-direction rotation due large deformations, as well as load reversals. Preloads further affect the mean strain state and fatigue response. As a result, predicted fatigue life can be sensitive to the postprocessing method used to calculate alternating and mean strains from FEA results.

This study systematically evaluates how different strain evaluation techniques influence predicted fatigue life in surrogate specimens and representative device geometries, highlighting that postprocessing methodology alone can substantially alter fatigue predictions, even when using identical FEA results. All simulations were conducted using the commercial FE solver Abaqus with the in-built superelastic material model. These findings provide critical guidance for robust FEA-based fatigue assessment in nitinol-based medical devices, supporting improved design reliability and long-term implant safety.