Abstract: This project employs a combination of CALPHAD calculation and machine learning method to provide alloy design guidance for two cases: 1) determining composition-process-cracking relation for laser power bed fusion (LPBF) with tool steel and 2) discovering HEA alloy space with promising high-temperature oxidation resistance.

Melt pool heat transfer modeling was constructed to provide a thermal profile for a DICTRA simulation. A CALPHAD calculation was conducted to estimate bulk thermodynamic properties and microsegregation for large amounts of candidate compositions. Different machine learning methods were deployed to select optimized ones for different scenarios. The benefits and limitations of CALPHAD-aided alloy design will be discussed by comparing the computational and experimental results for the two cases.

Bio: Yining He is a Ph.D. student in  the Materials Science and Engineering department of Carnegie Mellon University. She received a B.S. from Shandong University and M.S. from Carnegie Mellon University, both in materials science and engineering. Her research work is focused on alloy development for LPBF additive manufacturing, covering titanium, tool steel, and high-entropy alloys.