GSolid Project is a computational materials research initiative focused on understanding the solid solubility of oxygen in metal alloys, with an emphasis on high-entropy alloys (HEAs). The project integrates experimental data, thermodynamic modeling, and first-principles calculations to address a major gap in alloy design: the lack of reliable thermodynamic data for oxygen dissolution in complex metallic systems.
This work combines data from the Materials Platform for Data Science (MPDS) and The Materials Project, enabling the extraction and analysis of phase diagrams, liquidus curves, and solid-state energies for metal–oxygen systems. Using these datasets, the project computes regular solution mixing enthalpies and investigates how factors such as electronegativity influence oxygen solubility in multi-component alloys.
To overcome the limitations of traditional CALPHAD and DFT approaches when used in isolation, this project develops a hybrid computational strategy that merges experimental thermodynamic boundaries with DFT-based energy predictions—supporting more accurate characterization of oxidation resistance and alloy design for HEAs.
This work represents a focused subcomponent of the broader GLiquid Project led by the Sun Research Group at the University of Michigan. To visually explore the full scope of the project's findings, you can visit the Binary Phase Diagram Map or the Ternary Interpolation App
Check out the project on GitHub
The repository includes:
> Data pipelines for extracting and preprocessing materials data from MPDS and Materials Project
> Jupyter notebooks documenting the analysis workflow
> Utility scripts for thermodynamic calculations
> Machine learning workflows built in Orange Data Mining to assess and predict mixing enthalpies
Project collaboration was made possible through the Michigan Institute for Data & AI in Society (MIDAS) Faculty Student Research Connections Program.