Steven Kiyabu is pursuing his Ph.D. in Mechanical Engineering and Scientific Computing. He began working in the Energy Storage and Materials Simulation Lab under the direction of Prof. Donald Siegel in 2015 as an undergraduate and continued working there for the duration of his Ph.D. His research project has focused on using both computation as well as data science to discover superior materials for next generation thermal energy storage systems. Salt hydrates in particular demonstrate promise for high-capacity heat storage systems, but only a limited number have been characterized experimentally. As such, Steven has used high throughput density functional theory calculations to compute basic thermodynamic quantities and heat storage metrics from first principles for thousands of salt hydrates. Furthermore, he developed a feature engineering scheme to train high accuracy machine learning models that could predict the thermodynamics of salt hydrates, something that had previously been shown to be a difficult challenge. From these machine learning models, he identified several useful design rules for heat-storing salt hydrates.
Beyond his research, Steven is involved in GradCru, a Christian graduate student ministry at the University of Michigan. He has been an active participant from the beginning of graduate school and has served as the student president from August 2019 - August 2021, performing many administrative tasks for this student-led organization.