Stress and Fracture Mechanics Approach to Fatigue Life Prediction of Nitinol Medical Devices: An Inclusion Analysis

Fatigue life prediction for Nitinol cardiovascular devices relies traditionally on strain-life (ε-N) curves and modified Goodman diagrams representing statistical probabilities of fracture under specific mean strain and strain amplitude conditions. Because the fatigue life of commercial Nitinol depends on flaw-tolerant empirical methods due to inclusion-dominated failure mechanisms, prediction reliability is based on the probability of finding inclusions of critical size and orientation at critical locations. This necessitates testing campaigns of tenths to hundreds of specimens for millions to billions of cycles. Traditional stress-based damage tolerance approaches relying on long cracks fail to provide reliable predictions as Nitinol devices often operate in the short crack regime.

Quantitative metallographic characterization of Nitinol from different melt sources and melt methods was conducted to evaluate inclusion size, shape, and distribution in both transverse and longitudinal orientations following ASTM guidelines. Analytical results from SEM/BSE imaging and inclusion analysis showed VAR Nitinol exhibited extreme inclusion dimensions of 20μm and >270μm, while VAR/EBR Nitinol achieved inclusion sizes <10μm.

Experimental studies of surrogate specimens manufactured from these different Nitinol tubing variants demonstrate a transition to stress-based fracture mechanics approaches through strategic control of inclusion sizes. Such approaches allow to define safe device operation and differentiate between two fatigue regimes: i) a>a_c where fatigue threshold is dependent on crack length/inclusion size, and ii) a<a_c where threshold corresponds to the intrinsic fatigue life of flaw-free material. These investigations allow determination of the critical crack propagation thresholds for short cracks, enabling deterministic fatigue life prediction for complex devices.