| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Additive Manufacturing Modeling, Simulation and Machine Learning
|
| Presentation Title |
AM Microstructure Image Prediction Using Dimension Reduction |
| Author(s) |
Arulmurugan Senthilnathan, Paromita nath, Sankaran Mahadevan |
| On-Site Speaker (Planned) |
Arulmurugan Senthilnathan |
| Abstract Scope |
Variability in the additive manufacturing (AM) process and material properties affect the microstructure which influences the macro-scale mechanical response. Systematic quantification and propagation of this uncertainty and process design under uncertainty require numerous process-structure-property (P-S-P) simulations. However, P-S simulations (thermal models), which relates the microstructure to the process parameters, are computationally expensive. Therefore, an inexpensive surrogate model is necessary. Moreover, training the surrogate model to predict the high-dimensional microstructure image generated by P-S simulation is challenging. Hence, a novel dimension reduction strategy is first introduced to reduce the dimensions of microstructure image. A surrogate model is then constructed to predict the reduced dimensional features given AM process parameters. The microstructure image is then reconstructed from the predicted low-dimensional features and verified against the phase-field predicted microstructure, using moment invariants. Developing this surrogate modeling approach paves the way for solving computationally expensive tasks such as uncertainty quantification and optimal process design. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, Computational Materials Science & Engineering, Machine Learning |