About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Theory and Design of Metallic Glasses
|
| Presentation Title |
Exploring Metallic Glasses Orders of Magnitude Faster |
| Author(s) |
Sebastian A. Kube, Maxwell Shinn, Vignesh Selvaraj, Apurva Mehta, Dane Morgan, Jan Schroers |
| On-Site Speaker (Planned) |
Sebastian A. Kube |
| Abstract Scope |
The physical understanding and predictability of metallic glass forming ability and properties remain limited. Experimental data are scarce and biased compared to the vast potential composition space. Motivated by this, we use robotics, high-throughput synthesis, and automatic characterization to accelerate experimental testing by orders of magnitude. Integrated with modeling and machine learning, we can reveal complex physical behavior in the big picture: Using high-throughput synthesis via co-sputtering, we have generated a diversely sampled data set on glass formation, comprising ~13,000 ternary and quinary alloy compositions across 20 different elements. We test simple, interpretable models, which turn out surprisingly accurate, but also point towards the inherent limitations of film-based synthesis. Accelerating synthesis and testing with a focus on bulk-alloys is urgently needed: We provide a preview of our autonomous, self-driving laboratory platform “AlloyBot” developed at UW-Madison, highlighting its automatic arc melting capability that can synthesize 100+ alloy compositions per week. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Thin Films and Interfaces |