About this Abstract |
| Meeting |
2025 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Predicting Chemistry-Dependent Mechanical Behavior in High-Entropy Alloys: Iterative Design Insights from the BIRDSHOT Center Using Data-Driven and Generative Models |
| Author(s) |
Nicolas Flores, Trevor Hastings, Mrinalini Mulukutla, Wenle Xu, Daniel Lewis, Bibhu Sahu, Daniel Salas Mula, Danial Khatamsaz, Jacob Hempel, Douglass Allaire, Ibrahim Karaman, James Paramore, Brady Butler, George Pharr, Vahid Attari, Raymundo Arroyave |
| On-Site Speaker (Planned) |
Nicolas Flores |
| Abstract Scope |
Structural High Entropy Alloys (HEAs) are crucial in advancing technology across various sectors, including aerospace, automotive, and defense industries. These alloys encompass a vast design space, making the discovery of new alloys a challenging task that necessitates implementing methods to discern the underlying patterns within the existing chemistry, process, structure, and property data. This work presents a series of Ni-Co-Fe-Cr-V-Mn-Cu-Al alloys that were selected using multi-objective Bayesian optimization, batch synthesized, characterized, and analyzed using machine learning for trends in composition-property relationships. We show various compositional sensitivity analyses including observations on causes of samples that exhibited brittle and fractured nano-indentations. Finally, an autoencoder model is trained and used to generatively predict target properties from alloy chemistry. These results demonstrate an effective computational approach towards rapid material design and analysis in the new age of materials discovery. |
| Proceedings Inclusion? |
Planned: |