| About this Abstract |
| Meeting |
Materials Science & Technology 2020
|
| Symposium
|
High Entropy Materials: Concentrated Solid Solution, Intermetallics, Ceramics, Functional Materials and Beyond
|
| Presentation Title |
Machine Learning and Data Analytics for Identification of HEA Compositions and Processing Conditions Resulting in Enhanced Fatigue Resistance |
| Author(s) |
Xuesong Fan, Baldur Steingrimsson, Orlando Rios, Anand Kulkarni, Duckbong Kim, Peter K. Liaw |
| On-Site Speaker (Planned) |
Peter K. Liaw |
| Abstract Scope |
This presentation outlines an innovative approach to application of machine learning and data analytics, aimed at accelerating the identification of high-entropy alloy (HEA) compositions and process conditions resulting in attractive fatigue resistance. We present general methodology for predicting the fatigue resistance of HEAs, one capable of accounting for physics-based dependencies. We show that HEAs generally exhibit fatigue resistance superior to that of conventional alloys. For a given composition, we indicate, through application of data analytics, that the fatigue resistance of HEAs seems primarily correlated with the ultimate tensile strength (UTS), followed by the defect properties, grain size, and process parameters. Hence, given the multiple sources that impact the fatigue resistance, we note that accurate prediction of the fatigue resistance requires knowledge not only of the UTS, but also of defect properties, grain size, and process conditions. We demonstrate consistency of our predictions with empirical rules and experimental findings. |