About this Abstract |
Meeting |
2020 TMS Annual Meeting & Exhibition
|
Symposium
|
Materials Design Approaches and Experiences V
|
Presentation Title |
Materials Discovery and Design using Heritage Data |
Author(s) |
Amit K. Verma, Jeffrey A Hawk, Vyacheslav Romanov, Jennifer L W Carter |
On-Site Speaker (Planned) |
Amit K. Verma |
Abstract Scope |
Heritage data for different classes of high-temperature alloys was studied using machine learning to accelerate the materials discovery for the next generation of materials with better creep strength and toughness. Visualization techniques such as t-distributed stochastic neighbor embedding were utilized to explore the information gaps that exist within the data and regression methods such as linear and lasso regression were utilized to identify the alloying elements that contribute to creep strength, tensile strength, and ductility. The talk will focus on limitations of data for data-driven studies, highlight the importance of data collection for future studies, and will go through contemporary power-law creep methodologies that were used to reduce the timeline for materials selection. Overall, this work contributes to the field of materials design by (1) employing machine learning to highlight the missing opportunities, and (2) emphasizing that well constructed design space can accelerate the discovery of new materials. |
Proceedings Inclusion? |
Planned: Supplemental Proceedings volume |