A professor at the Cullen College of Engineering will be tapping into the potential of AI systems in order to explore high heat applications that wouldn’t be practical with conventional experimentation.
Hadi Ghasemi, J. Willard Gibbs Distinguished Professor of Mechanical & Aerospace Engineering, is the single PI for the research, “AI-Driven Topological Optimization and Interfacial Physics for Extreme Thermal Management.” The U.S. Navy’s Office of Naval Research (ONR) will fund the proposal for $500,000 over three years.
According to the project’s abstract, efficient thermal management is essential for advancing AI systems, electronics and photonics.
“Although extensive experimental efforts have been devoted to the fabrication and testing of cooling structures over the past several decades, the vast design space, spanning materials, length scales, and morphological complexity, makes exhaustive experimental exploration impractical,” the authors wrote. “Identifying optimal architectures for high-heat-flux applications through experimentation alone is therefore infeasible.”
The grant continues Ghasemi’s earlier work in two different ways — research into AI, and partnering with the ONR. Both avenues have been fruitful and self-sustaining.
Earlier this year, Ghasemi and doctoral students Amirmohammad Jahanbakhsh and Saber Badkoobeh Hezaveh shared research about how thin films could reduce the heat impact of data centers. “Physics-informed neural network based topology optimization for thin-film evaporation in hierarchical structures” was published in International Journal of Heat and Mass Transfer. That research was the product of a 2023 ONR grant for $500,000.
Ghasemi joined the faculty of Cullen in 2014. His NanoTherm Research Group has excelled both in terms of research produced, with more than 100 publication credits in the past five years, and students mentored. More than 30 students have earned an advanced degree with Ghasemi as their advisor.