| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Structure-Property Relationships of Bulk Metallic Glasses
|
| Presentation Title |
Machine Learning Predictions of Glass Forming Ability |
| Author(s) |
Dane Morgan, Benjamin Afflerbach, Lane Schultz, Maciej Polak, Paul Voyles, Izabela Szlufarska |
| On-Site Speaker (Planned) |
Dane Morgan |
| Abstract Scope |
In this talk we discuss our recent work on predicting measures of glass forming ability, such as critical casting thickness and critical cooling rate, using machine learning methods. We demonstrate that some limitations on existing data can be overcome through using text mining with large language models and integration of multimodal data from casting, critical cooling rate, and melt spinning studies. We explore the impact of different types of features, including characteristic temperatures, molecular simulation data, and elemental properties, which require orders of magnitude different efforts to obtain but can produce different qualities of results. We finish with a discussion of some of the challenges of using such machine learning models to discover new bulk metallic glasses. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, |