Conventionally, engineers have to perform fatigue testing, either in stress or in strain controlled mode, to determine the fatigue properties of a material, which costs a great deal of time and money. Therefore, the concept of computational fatigue design has been proposed and received an increasing interest in recent years. In this research, a microstructure-based computational fatigue design model, named TMW model, is further studied by first using it to predict the fatigue crack nucleation lives of eight different alloys and steels, and comparing the predicted lives with that calculated values from the Coffin-Manson-Basquin relations which are obtained from experimental data fitting, second by developing the mathematical expressions of the surface roughness factor in the TMW model in terms of the arithmetical mean deviation of the assessed profile which can be determined experimentally, thus making the TMW model more applicable. In addition, a microstructure-based finite element analysis (FEA) model is created to investigate the effect of microstructural inhomogeneity (grain orientation) on the fatigue crack nucleation life of nickel-based alloy Haynes 282 in different strain ranges from Low cycle fatigue (LCF) to High cycle fatigue (HCF) at different stress amplitudes. Grain orientations are randomly assigned to a material representative volume element (RVE) with 20 random functions created for both HCF and LCF simulations. The TMW model shows effective for predicting the fatigue crack nucleation life of structural materials. The FEA simulation reveals that potential fatigue crack nucleation sites are likely to occur at grain boundaries right on or close to the free surface of the RVE. For both HCF and LCF cases, the simulated fatigue crack nucleation in Haynes 282 is more likely to occur in the grains of [1- 1 1] family, while the grains with the least probability of fatigue crack nucleation tend to be oriented towards [0 0 1] family.
Share this page on your timeline