Societal Impact
The NanoTherm Research Group studies thermal transport, phase change, surface physics, and physics-informed AI with the goal of connecting fundamental science to energy, water, computing, and infrastructure challenges.
Future technologies will require thermal systems that are smaller, more powerful, more reliable, and more efficient. Our research aims to provide the scientific foundation for those systems by combining heat transfer, interfacial physics, materials design, and physics-informed computation. The long-term impact is to enable advanced computing, sustainable energy conversion, resilient infrastructure, and engineered surfaces that can operate under demanding real-world conditions.
Thermal management for AI data centers
High-power AI data centers generate substantial heat because dense GPU and accelerator deployments operate at high power densities. Effective thermal management is essential for reliable operation. Cooling energy consumption can account for nearly half of a data center's total energy use, making thermal design a central issue for sustainable computing.
Our research in physics-informed AI and topology optimization aims to accelerate the design of high-performance cooling architectures. We will invent and develop a prototype of a novel cooling system for AI data centers with the potential to reduce cooling power consumption by up to 60%, while providing stable cooling performance independent of ambient temperature.
The envisioned system will be deployed as an add-on to existing infrastructure, requiring no interruption to data-center operations. This approach is designed to make advanced cooling compatible with current facilities while reducing energy demand and improving operational reliability.
Lower cooling power
AI-guided cooling design targets reduced energy use in high-density computing facilities.
Stable operation
Thermal systems are designed for robust performance across ambient conditions.
Retrofit deployment
Add-on architecture is intended to work with existing data-center infrastructure.
Energy storage and hydrogen systems
Energy storage is essential for reliable use of renewable electricity, electrified transportation, and resilient power systems. Because solar and wind generation vary with weather and time, storage systems must move energy across hours, days, and seasons while maintaining high safety and efficiency.
Hydrogen is an important energy carrier for long-duration and large-scale storage. Compressed hydrogen, liquid hydrogen, and material-based storage systems each involve coupled heat transfer, phase change, interfacial transport, and safety constraints. Our research on thermal transport, hydrates, and nanoscale confinement helps clarify how hydrogen can be stored, released, and managed in practical energy systems.
Long-duration storage
Hydrogen can store energy at scales and durations that are difficult for short-term batteries alone.
Thermal control
Heat transfer governs compression, liquefaction, boil-off, hydrate growth, and release processes.
Clean energy systems
Improved storage can help connect renewable generation with industry, transportation, and grid reliability.
Ice and frost physics
Ice and frost formation affect transportation, aircraft, wind turbines, power transmission, refrigeration, heat exchangers, and building systems. Ice accumulation can reduce efficiency, increase maintenance cost, and create safety risks. Understanding the physics of nucleation, frost growth, ice adhesion, and ice shedding is therefore important for resilient infrastructure.
Our work connects surface physics, heat transfer, and mechanics to develop durable icephobic and ice-shedding materials. By studying how ice forms, adheres, and fractures on surfaces, we aim to guide coatings and structures that reduce ice accumulation and lower the energy and operational cost of de-icing.