About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
Presentation Title |
Hyperspectral In-Situ Process Monitoring with High-Speed Infrared Pyrometry, Eddy Current Testing, and Machine Learning, for Predictive Analysis of AM Part Properties |
Author(s) |
Medad C.C. Monu, Dermot Brabazon |
On-Site Speaker (Planned) |
Medad C.C. Monu |
Abstract Scope |
This study investigates the integration of hyperspectral in-situ process monitoring techniques by high-speed infrared pyrometry and eddy current monitoring (ECM), enhanced by machine learning (ML), to predict and analyze the properties of 316L stainless steel (SS) parts produced through Powder Bed Fusion-Laser Beam (PBF-LB). Key material properties evaluated are density (measured by Archimedes and micro-CT methods), Vickers microhardness, and surface roughness. High-speed infrared pyrometry provides real-time thermal data of the melt pool, ensuring accurate temperature monitoring. ECM offer insights into electrical conductivity and material structural integrity. ML algorithms are employed to analyze and correlate the extensive datasets from these monitoring techniques, aiming to predict the target material properties and optimize process parameters. This hyperspectral approach with ML promises significant advancements in predictive analysis and process optimization for PBF-LB manufacturing of 316L SS. Thereby, enhance the consistency and quality of the produced components, paving the way for PBF-LB digital twins. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Machine Learning, Characterization |