Toward a Digital Twin: Micro‐CT‐Resolved Fatigue‐Life Prediction for SLM AlSi10Mg

Résumé fourni par la source

ABSTRACT A microcomputed tomography (micro‐CT)‐based framework was developed to predict the fatigue life of selective laser melting (SLM) AlSi10Mg specimens with different porosity levels and defect morphologies. Because full‐specimen pore‐resolved finite element analysis (FEA) is computationally impractical, specimen‐specific regions of interest (ROIs) were extracted from experimentally identified fracture‐critical locations for fatigue simulations. Predicted fatigue lives were consistently conservative, with an average log‐scale deviation of 13.64%, while preserving the experimental fatigue‐life trend. Additional simulations between 140 and 9 MPa enabled the extraction of ROI‐specific FEA‐derived fatigue strength. Void volume fraction (VVF) exhibited inverse relationships with both predicted fatigue life and fatigue strength. Comparisons of ROIs with similar VVF showed that volume‐weighted average sphericity also affected fatigue performance, indicating that defect morphology contributes beyond porosity alone. The proposed framework provides a defect‐resolved and morphology‐sensitive basis for fatigue assessment and supports future digital‐twin–oriented prediction of additively manufactured AlSi10Mg components.