Abstract
The increasing adoption of additive manufacturing (AM) for metallic components in critical applications necessitates robust fatigue life prediction methodologies. Unlike conventionally manufactured materials, AM alloys exhibit unique microstructural features, including characteristic grain structures, porosity, and residual stresses, which significantly influence their fatigue performance. This study presents a comprehensive investigation into microstructure-sensitive fatigue life prediction models tailored for additively manufactured metallic alloys. We review existing models and propose an integrated approach that accounts for the key microstructural constituents and defect characteristics inherent to AM processes like Selective Laser Melting (SLM) and Electron Beam Melting (EBM). The methodology incorporates microstructural parameters such as grain size, phase distribution, and defect morphology (porosity, lack of fusion) into established fatigue frameworks, such as continuum damage mechanics and fracture mechanics. Machine learning techniques are explored to identify sensitive microstructural features and accelerate the prediction process. Experimental data from literature, alongside novel simulation results, are used to validate the proposed models. Our findings indicate that accounting for microstructural anisotropy and the synergistic effects of internal defects is crucial for accurate fatigue life prediction. The developed models demonstrate improved correlation with experimental fatigue data compared to traditional approaches, highlighting the importance of a microstructure-informed design philosophy for AM components. This research provides a foundation for developing more reliable and predictive fatigue assessment tools, essential for ensuring the safety and durability of AM metallic structures in demanding engineering fields.