| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Accelerating Materials Discovery of HEA’s through Constraint Based High Throughput Design, Synthesis and Batch Bayesian Optimization Framework |
| Author(s) |
Mrinalini Mulukutla, Raymundo Arroyave, Danial Khatamsaz, James Paramore, Brady Butler, Trevor Hastings, Daniel Lewis, Daniel Salas, Nicole Person, Wenle Xu, Douglas Allaire, George Pharr, Ibrahim Karaman |
| On-Site Speaker (Planned) |
Mrinalini Mulukutla |
| Abstract Scope |
High entropy alloys have been of great interest to the materials research community for the development of advanced materials with exceptional properties. Efficient and accelerated exploration of these vast compositional spaces has been an ongoing challenge with conventional high throughput experimentation/computational methods. We address this challenge by the implementation of a framework that employs a composition agnostic, multi-objective, multi-constraint co-design for performance, and manufacturability. Using 6-element phase space (Co, Cr, Fe, Ni, V, and Al), we defined the space through intelligent constraint-based filtering, produced candidate alloys by vacuum arc melting followed by characterization for objectives relevant to structural materials for extreme conditions. They are iterated in a closed loop by Batch Bayesian Optimization to identify Pareto set for the subsequent iterations. Optimal exploration involving five successful iterations showcases the superiority of the framework’s powerful machine learning algorithms suggesting scope for higher fidelity systems in future works. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Entropy Alloys, Computational Materials Science & Engineering, Other |