Abstract
Fatigue strength is vital for engineering applications of aluminum alloys. Accurate models incorporating mesoscopic defect-representing features are one of the issues for accurate fatigue strength prediction. A fatigue life prediction method based on meso-defect-representing features is proposed in this study. Based on staged fatigue damage, meso-defect data was obtained by X-ray CT. After 3D reconstruction and simplification, porosity, shape, and location were selected as the meso-defect-representing features using correlation coefficient analysis. Weights of meso-defect features were determined through FEM simulation. A mesoscopic damage variable incorporating the weights of porosity, shape, and location for meso-defect was defined. Correlation between fatigue life and meso-defect features was established through the mesoscopic damage variable. Experimental verification results showed that the prediction method is an effective method for fatigue life assessment.