About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
Presentation Title |
Additive Manufacturing Digital Twin (AMDT): Part Level Process Map Characterization Using Physics Based Simulation and Machine Learning |
Author(s) |
Peter Myung-Won Pak, Francis Ogoke, Amir Barati Farimani |
On-Site Speaker (Planned) |
Peter Myung-Won Pak |
Abstract Scope |
We explore a novel approach to Digital Twin (DT) modeling in Additive Manufacturing (AM), focusing on geometry dependent defects within the fabricated part to create a part level process map. Established analytical models such as those from Eagar-Tsai and Rosenthal are used to generate physics based simulations of melt pool heat flow throughout the part. More specifically, we investigate approaches for solving boundary interactions to address the inherent complexities of part geometries. Machine Learning models are further applied to expedite and enhance the output, enabling the prediction and identification of potential defects with increased efficiency. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Machine Learning, Modeling and Simulation |