Predicting the fatigue life of an AlSi10Mg alloy manufactured via laser powder bed fusion by using data from computed tomography

Abstract

A modelling strategy is proposed to evaluate the influence of defect morphology on the fatigue limit of additively manufactured Al alloys by: (i) obtaining an x-ray micro-Computed Tomography (μ-CT) 3D image of the material, (ii) computing the Equivalent Inertia Ellipsoid of each individual pore, (iii) modelling the influence of the defect on the fatigue limit through the Defect Stress Gradient (DSG) approach coupled to the Eshelby theory and, (iv) 3D mapping the criticality of each individual defect. For this fatigue study, an AlSi10Mg alloy was manufactured by laser powder bed fusion using sub-optimal deposition parameters in order to produce large lack-of-fusion defects. After a T6 heat treatment, tension-compression fatigue tests, with R = −1, were conducted on specimens oriented with their loading axis either parallel or normal to the Z-axis of the additive manufacturing equipment. Two samples were characterised before μ-CT testing in order to characterise the initial 3D defect population. Each sample was fatigued step by step in order to determine the fatigue limit. The fracture surface was observed in order to identify the critical defect in the initial μ-CT image. A comparison with the fatigue results led to the following conclusions: (i) when the longest axis of the defect is perpendicular to the loading axis, modelling the defect as an equivalent inertia prolate ellipsoid gives better results (5 % error on the fatigue limit) than modelling it as a simple equivalent sphere (22 % error on the fatigue limit), (ii) the prolate ellipsoid is not relevant when the longest axis of the defect is oriented along the loading axis; in this case an oblate equivalent ellipsoid should be used, (iii) the concept of ‘size’ for a complex 3D shaped defect should be linked to the inertia and the loading, (iv) with this approach, surface defects are shown to be more critical than internal ones for fatigue life and, (v) a 3D defect criticality map of the entire sample can be plotted to provide visual feedback on which defects are the most critical for fatigue life.