|About this Abstract
||1st World Congress on High Entropy Alloys (HEA 2019)
||High Entropy Alloys 2019
||Using Machine Learning to Guide Multiple Principle Element Alloy Design
||Jose Loli, Yining He, Amish Chovatiya, Bryan Webler, Zachary Ulissi, Jack Beuth, Maarten De Boer
|On-Site Speaker (Planned)
Multiple principle element alloys (MPEAs) can exhibit exceptional resistance to wear, creep, corrosion and high temperature oxidation. The optimization of MPEAs can be daunting due to the vast number of possible compositions. If properties can be calculated as a function of composition, numerous optimization strategies to guide alloy design exist. For complex phenomena, in this work: high temperature oxidation, it is not feasible to make calculations of all parts of the problem. However, there are calculatable quantities that should correlate with oxidation resistance. Our approach here is to investigate if quantities calculated via CALPHAD software can be used to develop optimized compositions. A database of thermodynamic quantities was generated for 4 and 5 element equimolar combinations from a 12 element palette. An objective function and black-box optimization software were utilized to select 5 most promising compositions. Arc-melted buttons of these compositions are being made and characterized to test the predictions.