About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
CALPHAD Based Screening for Rapid Training of Alloy Design Models |
| Author(s) |
Bernard Gaskey, Mikayla Obrist, Arindam Debnath, Janith Wanni, Avanish Mishra, Nithin Mathew, Saryu Fensin |
| On-Site Speaker (Planned) |
Bernard Gaskey |
| Abstract Scope |
Machine learning models promise to revolutionize alloy design by rapidly sampling composition space to optimize alloys for a specific application. However, training such models requires data that is not easily obtained experimentally. To bridge this capability gap, the thermodynamics-based calculation of phase diagrams (CALPHAD) approach can be used to rapidly generate large property datasets. This data can be combined with experimental results to provide a more robust and diverse dataset for model training. Here, we demonstrate this approach to design multi-component alloys with constraints on both service properties such as strength and density as well as manufacturability criterion including crack susceptibility and phase stability. This hybrid computational workflow allows rapid, automated design of candidate alloy compositions, greatly reducing the required experimental input without a significant loss in fidelity. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Entropy Alloys, Computational Materials Science & Engineering, |