Exploring the stress concentration factor in additively manufactured materials: A machine learning perspective on surface notches and subsurface defects

This investigation aims to establish a comprehensive computational methodology for analyzing the effects of notch geometry, subsurface porosity, and their interaction on the structural integrity of metallic components. Utilizing a machine learning algorithm, the study examines the stress concentrators and their statistical significance in determining structural integrity. Previous studies have examined various facets of this methodology, yet this investigation employs a systematic approach to examine all pertinent factors. Results reveal that the notch opening angle does not significantly influence stress concentration, and crack mouth opening displacement is heavily affected by the notch depth, which can limit its usage for studying the growing cracks, especially in the presence of subsurface pores. This study contributes to the understanding of acceptance thresholds and the economic viability of maintenance operations suggested in various governing standards.