|About this Abstract
||2022 TMS Annual Meeting & Exhibition
||Additive Manufacturing: Advanced Characterization with Synchrotron, Neutron, and In Situ Laboratory-scale Techniques II
||In-situ Process Monitoring for Laser Powder Bed Fusion: A Data-driven Approach
||Anant Raj, Dongli Huang, Benjamin Thomas Stegman, Hany Abdel-Khalik, Xinghang Zhang, John W Sutherland
|On-Site Speaker (Planned)
The critical issue of part-to-part repeatability continues to impede large-scale adoption of additive manufacturing beyond rapid prototyping. A wide variety of in-situ monitoring techniques have been developed to address this issue. For laser powder bed fusion, co-axial melt-pool monitoring is one of the most widely used techniques that provides direct access to the quality of the melt-pool. However, an extensive quantitative analysis of the variation of the in-situ melt-pool signatures based on print parameters and process fluctuations and their impact on the final part quality has yet to be undertaken. In this study, we employ supervised and unsupervised machine learning to demonstrate that the in-situ melt-pool signatures can be distinctly mapped onto corresponding print parameters along with process fluctuations like variation across the build-plate. Further, our analysis suggests that in-situ signatures can be leveraged to predict the mechanical properties of the part, enabling development of control algorithms for better repeatability.
||Additive Manufacturing, Machine Learning, Other