A probabilistic approach to modeling microstructural variability and fatigue behavior in ASTM F562 medical grade wire

Several previous studies have demonstrated the importance of constituent inclusion particle distributions on variability in observed fatigue lifetimes. Such models have often concentrated on crack growth behavior in standard test specimens such as single edge notch tension (SENT) geometries. The methodology used here is similar to others in that constituent particles sizes were modelled as initial cracks that were grown to failure using a fracture mechanics approach. Lognormal distributions were fit to inclusion size parameters of 0.178 mm [0.007 in] ASTM F562 alloy medical grade wire. A critical volume approach was used to generate random subsets of inclusion sizes from which particles were chosen with a maximum likelihood of micro-crack generation. These sites were then grown at several stress levels to failure using a plasticity-induced crackclosure model to predict total fatigue life. The resulting lives were found to be in good general agreement with experimentally observed variability in rotary beam fatigue testing.