Statistical Evaluation of Nitinol Fatigue Using for Both Single and Multi-Amplitude Conditions

Nitinol is used extensively for cardiovascular devices given its super-elastic behavior and favorable fatigue performance. Although nitinol fatigue is often evaluated by calculating a factor of safety based on strain amplitude and a selected fatigue strength percentile, such an approach is limited. These limitations include not directly incorporating the complete statistical distribution of the fatigue strength and having limited capability of combining multiple loading modes and magnitudes, which almost always occur in implanted medical devices. Further, there is a growing trend in the industry toward evaluating a probability of fracture, rather than a fatigue safety factor. The purposes of this study were to (A) further evaluate the cumulative fatigue performance of nitinol through incorporation of both tubing and wire specimens and (B) to evaluate the ability of statistical methods to predict long term fatigue performance relative to fatigue safety factor approaches.

Diamond shaped tubing specimens were used for the fatigue testing. The specimens were first crimped, after which they were cycled at a range of strain amplitudes. Both single amplitude and multi-amplitude tests were performed. The test results demonstrated a cumulative effect of multiple cyclic strains on nitinol fatigue. Predictive capability of statistical fits that incorporate strain, stress, and strain energy were evaluated. The results demonstrated that incorporation of both stress and strain through strain energy resulted in the most consistent predictive ability compared to strain alone.