U.S. University Establishes Fusion Energy Seed Fund to Support Four Interdisciplinary Research Projects

The Institute for Fusion Studies at The University of Texas at Austin recently launched a fusion energy seed fund to support four interdisciplinary research projects. These projects will bring together teams from plasma physics, engineering, and computational science to address challenges in control, materials, particle confinement, and edge plasma simulation faced by fusion devices such as tokamaks.

The fund is jointly financed by the Institute for Fusion Studies (IFS), the Oden Institute for Computational Engineering and Sciences, and the Cockrell School of Engineering at The University of Texas at Austin. Each project is led by a principal investigator from the Oden Institute or the Cockrell School of Engineering, paired with a co-principal investigator from the Institute for Fusion Studies, and will support one graduate student for a two-year research period.

In recent years, fusion energy research and development has continued to gain momentum. The U.S. Department of Energy (DOE) supports related research through mechanisms such as the Fusion Innovation Research Engine (FIRE), in which researchers from the Institute for Fusion Studies at The University of Texas at Austin are already involved in two collaborative projects. The DOE has also launched the "Genesis" initiative to support universities, national laboratories, and industry in leveraging artificial intelligence to accelerate fusion commercialization.

Among the four funded projects, three will focus on exploring artificial intelligence surrogate models that use training data to replace portions of time-consuming, high-fidelity large-scale computations, reducing computational cost and time while maintaining result accuracy.

In the plasma control area, the research team plans to develop a digital twin model of tokamak plasma based on indirect sparse measurements, enabling real-time acquisition of the plasma's three-dimensional shape, position, and other key parameters, which will be used to optimize magnetic coil control. The project involves Diego del-Castillo-Negrete from the Institute for Fusion Studies and Omar Ghattas from the Oden Institute and the Cockrell School of Engineering, with the goal of developing faster and more reliable real-time control optimization tools. Researchers hope to test the relevant models in the future on devices such as the TCV tokamak in Lausanne, Switzerland, or the DIII-D tokamak at General Atomics.

In the plasma-facing component materials area, the research team will use large-scale computer simulations to screen candidate liquid metal alloy materials. The plasma inside a fusion reactor generates intense heat fluxes, radiation, and high-energy particles that can damage wall surfaces and other plasma-facing components. Since no single liquid metal can simultaneously meet all performance requirements, researchers will evaluate the behavior of liquid metals such as lithium, indium, gallium, and tin when alloyed with other materials, and use simulation data to train machine learning models for predicting material behavior over longer timescales. The project involves David Hatch, Narayana Aluru, and Yuanyue Liu, among others.

In the alpha particle confinement area, researchers will develop artificial intelligence surrogate models for tracking the trajectories of high-energy alpha particles. Alpha particle losses can affect the temperature and density required for the plasma to sustain fusion, and traditional particle orbit simulations are computationally intensive, making it difficult to iterate repeatedly during the early design phase. The project team hopes the new models will improve computational efficiency by more than an order of magnitude, enabling alpha particle confinement assessments to be incorporated earlier into the fusion device design optimization process.

The fourth project focuses on the tokamak plasma edge region, specifically scrape-off layer simulation. The divertor is responsible for exhausting excess heat and high-energy particles, and its design requires an accurate understanding of plasma edge behavior. Due to the complex physics in this region, existing high-precision simulations often incur significant computational costs. The research team plans to introduce new numerical methods to develop faster and more economical high-precision simulation tools, providing support for the design of fusion reactor edge plasmas and divertors.