About this Abstract |
Meeting |
2024 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
Presentation Title |
Multiscale and Machine Learning Modeling for Texture Prediction during Additive Manufacturing |
Author(s) |
Sudipta Biswas, Som Dhulipala, Peter German, Alexander Lindsay, Matthew Eklund, Andrea Jokisaari |
On-Site Speaker (Planned) |
Sudipta Biswas |
Abstract Scope |
Additive manufacturing (AM) techniques provide the opportunity to simultaneously design new materials and components with complex structures in less time, enabling faster material developments. However, the texture of the materials produced by such techniques is significantly different from conventionally manufactured materials. Such variations in microstructure make qualifying AM products challenging for extreme environment applications. Here, we integrate multiscale mechanistic models with machine learning approaches to establish the process-structure-property (PSP) correlation for additively manufactured 316L stainless steel. A multiphysics model is developed to demonstrate grain formation and grain growth during the laser powder-bed fusion process. The morphology of the AM product is analyzed to correlate how different process parameters impact the final microstructure. Additionally, the machine learning model is utilized to map the PSP linkage in a computationally efficient, reliable, and cost-effective way. Such maps pave the path forward to optimize the AM process based on desired microstructural and property requirements. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Computational Materials Science & Engineering, Machine Learning |