| About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Data Driven Alloy Design of High Entropy Alloys for Temperature Dependent Mechanical Properties |
| Author(s) |
Sayak Bal, Pikee Priya |
| On-Site Speaker (Planned) |
Pikee Priya |
| Abstract Scope |
Alloy design through experiments requires high-throughput experimentations which requires a lot of time and energy and is difficult or even impossible considering the multi-dimensional compositional possibilities in High Entropy Alloy (HEA) space. Data driven methodologies like Machine Learning (ML) techniques can easily be used to extrapolate the computational and/or experimental datasets to unknown compositions reducing the number of experiments and theoretical calculations by many folds. Machine Learning in this case, has been used as a tool for regression correlating a set of physics-informed variables or features to the target property of temperature dependent yield strength and elongation to failure from the literature (~735 data points). Genetic Algorithm is used to optimize the compositions from a 24 elements space to design structural alloys with yield strength as high as ~1800 MPa both at elevated temperatures of ~1000 K and cryogenic temperatures of ~273 K. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Entropy Alloys, High-Temperature Materials, Machine Learning |