| About this Abstract |
| Meeting |
TMS Specialty Congress 2025
|
| Symposium
|
The 7th International Congress on 3D Materials Science (3DMS 2025)
|
| Presentation Title |
Extracting Grain Boundary Mobilities From 3D Data Using Convolutional Neural Networks |
| Author(s) |
Jules M. Dake, Leonard Lauber, Thomas Wilhelm, Lukas Petrich, Orkun Furat, Volker Schmidt, Carl E Krill |
| On-Site Speaker (Planned) |
Jules M. Dake |
| Abstract Scope |
With the development and advancement of diffraction-based imaging techniques, an abundance of time-resolved 3D datasets of grain growth and other coarsening phenomena have recently become available. Large-scale 3D simulations of microstructural evolution are also now quite common thanks to progress in massively parallel computing architectures. These simulations rely on accurate values of grain boundary parameters (e.g., grain boundary energy and mobility) to return useful predictions; however, extracting these parameters from 3D experimental data, which is discretized in time and space, remains challenging. Here, we introduce a machine learning approach utilizing a convolutional neural network to directly extract grain boundary mobilities from sequential 3D mappings of grain growth in a single-phase aluminum alloy. |
| Proceedings Inclusion? |
Undecided |