The cohesive fatigue model of A. Ural and K. D. Papoulia (Modeling of fatigue crack growth with a damage-based cohesive zone model, European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2004)) is modified and implemented within the finite element code Abaqus. The model follows a bi-linear damage-dependent traction-displacement relation coupled with a damage evolution equation characterized by three material parameters corresponding to damage accumulation, crack closure and stress threshold.

High cycle fatigue is computationally intractable with cycle-by-cycle calculations. To make high cycle fatigue simulations possible, different extrapolation schemes have been proposed in the literature, with varying degrees of complexity, to account for the nonlinearity of the equations. Based on simple observations, two such schemes are proposed and tested in this work. A logarithmic scheme is found easy to implement, as well as capable of extrapolating the accumulation of material damage due non-constant amplitude fatigue loads. Finite element results are compared with high cycle fatigue test results for an aluminum alloy. Close matches between the test data and finite element simulations are obtained for different loading conditions.

The cohesive model is also used to capture the effect of a single peak overload, viz. crack retardation, in a ductile 316L steel alloy under plane stress conditions. The results indicate that a higher peak load results in higher fatigue crack retardation. The results also agree with experiments that suggest that strain hardening, not crack closure, is the leading mechanism for the overload effect.