Accelerated Discovery of Multi-Phase Alloys Through Machine Learning Surrogate Models

Team: #13 Samuel Price

Program: Materials Science and Engineering

In this senior design project, a comprehensive, element-agnostic approach for identifying
promising new alloys is presented. I demonstrate that machine learning algorithms can
be used to create accurate surrogate models of CALPHAD and can rapidly assess the phase stability of
millions of candidate alloys. This approach can greatly expedite the alloy design process.

Johns Hopkins University Whiting School of Engineering
Johns Hopkins University Applied Physics Laboratory
Jonah Erlebacher
Ian McCue

Team Members

  • Samuel Price

Project Links