About this Abstract |
Meeting |
2025 TMS Annual Meeting & Exhibition
|
Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
Presentation Title |
Microstructure Prediction in Laser Powder Bed Fusion via Physics-Based Modeling and In-situ Sensor Data Fusion |
Author(s) |
Kaustubh Deshmukh, Mihir Darji, Antonio Carrington, Alex Riensche, Christopher B. Williams, Prahalada Rao |
On-Site Speaker (Planned) |
Kaustubh Deshmukh |
Abstract Scope |
This work is aimed at predicting microstructure evolution in parts made using the laser powder bed fusion (LPBF) additive manufacturing process. Microstructure features, such as grain size, morphology, and orientation, are key determinants of functional mechanical properties. A prediction framework rooted in physics and sensor-based data-driven integration is created here. For this purpose, Inconel 718 parts were fabricated under nine different processing conditions. An experimentally validated thermal model is used to simulate the part-scale thermal history of the parts and generate end-of-cycle temperatures and cooling times. An In-situ monitoring dataset concerning melt pool-scale sensor signatures is generated via continuous thermal and optical tomography imaging. A supervised learning model is proposed for this study. Experimentally characterized microstructure features are fused with thermal model and in-situ sensor data to train this framework and predict microstructure features. Thus, this work takes a critical step towards multi-scale control of microstructure to predetermine the part quality in LPBF. |
Proceedings Inclusion? |
Planned: |
Keywords |
Additive Manufacturing, Characterization, Machine Learning |