Dr. Jean Ragusa to develop predictive software for nuclear reactors
With dwindling coal and oil reserves and increased awareness about climate change, zero-carbon energy sources-such as nuclear energy-are in the spotlight. In fact, as part of its Nuclear Energy University Program, the Department of Energy is funding Texas A&M University’s Dr. Jean Ragusa’s research that will improve the working of nuclear reactors. Ragusa is an assistant professor in the Nuclear Engineering Department and associate director of the Institute for Scientific Computation, a part of the Texas Engineering Experiment Station (TEES). TEES is the engineering research agency for the State of Texas and is a member of The Texas A&M University System. For this project, Ragusa will be collaborating with Dr. Pavel Solin, an associate professor of mathematics at the University of Nevada, Reno. Ragusa and Solin aim to provide engineers with significantly improved simulations of the various physical phenomena taking place in nuclear reactors. They will use their $587,000 grant to develop new and highly sophisticated computational methods to predict the behavior of complex coupled processes occurring in nuclear reactors, such as neutron flux, thermal-hydraulics and structural mechanics. Current models consider or "solve" one physical process at a time. For example, these models that can tell you where neutrons are in a reactor. How accurate this prediction is depends on how closely the model mimics the conditions in a reactor. Scientists know that a physical process is simultaneously influenced by different processes, but since current models consider only one physical phenomenon at a time, their predictions are not sufficiently accurate. Ragusa and Solin will develop open-source software based on advanced numerical analysis to simultaneously solve multiple physical processes and provide accurate predictions of the reactions occurring in reactors. Their multiphysics simulations will rely on the high-level software platforms the two researchers have developed independently in the past. Ragusa and Solin’s research could help improve the efficiency of existing nuclear reactors. Since this research will allow scientists to rely more on predictive simulations rather than expensive experimental mock-ups, it will also help scientists design reactors at a lesser cost. "Since experiments in nuclear science are expensive," Ragusa said, "before you build a reactor, it is important to verify how it will function and ensure that safety margins are respected; this can be done at a lower cost if computer models are more accurate. Predictive models are important because they can answer these questions using high-performance computing." Ragusa and Solin are working with other national laboratories, including Idaho National Laboratory, to develop their software. "Popularizing this relatively new technique will help nuclear engineers solve problems faster and more accurately," Ragusa said.