| About this Abstract |
| Meeting |
2021 TMS Annual Meeting & Exhibition
|
| Symposium
|
Bulk Metallic Glasses XVIII
|
| Presentation Title |
Machine Learning from Elemental and Simulation Features for Predicting Glass Forming Ability |
| Author(s) |
Dane Morgan, Benjamin Afflerbach, Lane Schultz, Janine Erickson, Dan Thoma, John Perepezko, Carter Francis, Paul Voyles, George Bokas, Jianqi Xi, Izabela Szlufarska |
| On-Site Speaker (Planned) |
Dane Morgan |
| Abstract Scope |
Robust accessible features that can accurately predict glass forming ability remain a grand challenge for the field of bulk metallic glasses. Here we present two approaches to this problem. First, we use machine learning to correlate a large database of measured and estimated critical cooling rates with elemental features and then predict critical cooling trends in a number of ternary metal systems. Then we demonstrate that molecular dynamics modeling can potentially be combined with readily accessible experimental data to obtain characteristic temperatures that can be correlated with critical casting diameters. Potentially useful correlations are obtained but significant uncertainties still plague these models and the open challenges will be discussed. |
| Proceedings Inclusion? |
Planned: |