| Abstract Scope |
Laser powder bed fusion (LPBF) adoption faces hurdles from repeatability issues and slow, costly traditional qualification methods. In-situ process monitoring (ISPM) using multiple high-resolution sensors like Optical Tomography (OT), Melt Pool Monitoring (MPM), photodiodes, and IR cameras offer a path to faster, model-based qualification. This research utilizes multi-modal ISPM data to detect process anomalies and variations and predict final part quality. Strong correlations were found between OT signals and defect formation, which enables a data-driven framework to detect anomalies which lie outside the defined thresholds of Grey Values (GV). Furthermore, these real-time signals facilitate the closed-loop control to actively alter process parameters during the build, such as adjusting laser power. This integrated monitoring, prediction, and control approach significantly accelerates qualification, reduces testing burdens, and enhances LPBF reliability for industrial applications. |