| About this Abstract |
| Meeting |
MS&T22: Materials Science & Technology
|
| Symposium
|
High Entropy Materials: Concentrated Solid Solutions, Intermetallics, Ceramics, Functional Materials and Beyond III
|
| Presentation Title |
A Bayesian Approach to Efficiently Discover Refractory High Entropy Alloys |
| Author(s) |
Raymundo Arroyave |
| On-Site Speaker (Planned) |
Raymundo Arroyave |
| Abstract Scope |
We present some recent advances in the development of frameworks to efficiently explore and exploit vast chemical-property spaces. Our framework relies on a Bayesian Optimization formalism to discover optimal materials under resource constraints. The framework employs multi-fidelity approaches, is capable of handling multiple objectives and constraints simultaneously, and can recommend multiple parallel queries at once. The framework is superior to traditional ICME approaches due to the capability of integrating experiments and simulations within a unified setting. It is also more efficient than combinatorial materials science approaches due to its iterative nature. The framework is deployed in the discovery of novel refractory high entropy alloys (RHEAs). While the optimization targets focus on mechanical performance, we also discuss some initial results, in which we evaluate the oxidation behavior as well as the performance of candidate alloys under 3D printing conditions. The framework is highly transferrable to other High Entropy Materials discovery problems. |