Grant: Alliance Grants

The additive manufacturing (AM) industry is tackling challenges in the serial production of metallic components by implementing multi-laser powder bed fusion (M-LPBF) systems with in-situ monitoring for quality assurance (QA). As with conventional manufacturing methods, reliable QA methods and platforms are essential for effectively deploying this technology in serial production. However, the use of multiple laser beams introduces additional processes and environmental disturbances, reducing reliability and repeatability. Furthermore, in-situ monitoring generates extensive data mixed with noise, necessitating advanced solutions for defect detection. This proposal aims to develop machine learning-driven QA algorithms for quad-LPBF systems, integrating optical tomography (OT) and photodiode melt pool monitoring (MPM) data to produce defect-free Cobalt Chromium Molybdenum (CoCrMo) components for the energy sector. This innovative approach seeks to enhance QA and defect detection, surpassing current methods, through four key objectives: 1) Integration and dimensionality reduction of OT and photodiode-based MPM datasets collected during the M-LPBF process; 2) Development of knowledge-based ML algorithms for defect detection during M-LPBF of CoCrMo alloy using integrated OT/MPM datasets; 3) Defect minimization in CoCrMo test components using an integrated quad-laser PBF and OT/MPM system; and, 4) Optimum process parameters development for fabrication of complex-shaped CoCrMo components using the integrated system developed in objective 3. The outcome of this technology will be pervasive, with a deep and broad impact on many industries (e.g., aerospace, automotive, and energy) that will benefit from reliable M-LPBF.