About this Abstract |
Meeting |
2022 TMS Annual Meeting & Exhibition
|
Symposium
|
Materials and Chemistry for Molten Salt Systems
|
Presentation Title |
Data-driven Models for Corrosion of Structural Alloys in Molten Chloride Salts |
Author(s) |
Christopher D. Taylor, Brett Tossey |
On-Site Speaker (Planned) |
Christopher D. Taylor |
Abstract Scope |
Increasing effort is being put into testing materials performance in high temperature, molten salt environments. To facilitate future data collection efforts and materials selection decision-making, a schema has been developed for corrosion data collected on alloys in molten salts. The schema provides the basis for a data-driven effort to characterize corrosion of alloys in molten salts using features such as composition, microstructure, and environmental parameters. In this talk, we will present the structure of the database, initial data visualization efforts and the results of predictor evaluation. Pathways towards combining science-based models for materials degradation with data-driven methods (such as machine learning algorithms like decision trees, random forest, and linear/non-linear regression methods) will be presented and discussed. Knowledge gaps identified as a result of the data collection and visualization efforts will also be presented, along with original data collected for alloy corrosion metrics in our laboratory. |
Proceedings Inclusion? |
Planned: |
Keywords |
ICME, Machine Learning, Environmental Effects |