This National Science Foundation (NSF) Engineering Program (CFDA 47.041) Project Grant award of $249,999 to the University of Wisconsin System is to establish a physics-informed machine learning (PIML) framework to predict and mitigate fatigue life variability in metal additive manufacturing (AM) components. The project aims to: 1) characterize the fatigue behavior of 316L stainless steel and Ti-6Al-4V alloy samples fabricated via laser powder bed fusion AM and post-processing, 2) develop the PIML framework to integrate physical fatigue models with data-driven deep learning to accurately predict fatigue life and scattering, and 3) optimize AM processes to reduce fatigue variability. The new knowledge and modeling methods from this 3-year project are expected to have a disruptive impact on the AM industry by providing predictive tools and scattering mitigation strategies to enable consistent high-quality printed components. The award also supports workforce training through interdisciplinary research and learning opportunities for students.
Generated 1/28/25, 9:44 AM