Micro-mechanical investigation of fatigue behavior of Al alloys containing surface/superficial defects

A crystal plasticity finite element (CPFE) simulation framework has been proposed in this study for the prediction of combined detrimental effect of mean stress and defects on the fatigue behavior of aluminum alloy. Experimental data for mean-stress effect on fatigue life and crack growth behavior was obtained on metal inter gas (MIG) welded joints of Al-5083/Al-5.8%Mg alloy plates and has been detailed in authors' previous work, referred later. The present study focuses on its prediction using computational framework only. A 2D representative model for material's microstructure was used for the simulations, generated using an anisotropic tessellation algorithm using the EBSD measurements data. A total of 10 different microstructure models were generated for each loading condition using six-node plane strain type quadratic triangular (CPE6) elements of mesh size 6 mu m. Two different types of cases were investigated: one without defect and other with a semi-circular surface defect. The simulated loadings at different stress ranges and stress ratios (R-ratio) were similar to the experimental conditions for the better comparison of the results. Significant heterogeneity in the distribution of R-ratios and the far-field applied R-ratio was observed. When defect was not considered, a clear deviation in the predicted fatigue lives from the experimental data was observed at different R-ratios: the predicted fatigue lives were higher than the experimentally observed fatigue lives. This was probably because of not considering the detrimental effect of defects on fatigue lives. But, when the defects were considered, the predicted results for different R-ratios were consistent with the experimental fatigue lives. The proposed CPFE simulation framework not only predicted well the effect of defects and mean stress on the fatigue lives, but also the scatter induced in them due to the defects.