Nearly 60 years ago, NASA engineers tested an experimental nuclear thermal rocket engine to determine if nuclear-powered propulsion might be a viable option for deep-space exploration such as a manned mission to Mars.
The test was successful, and even though the rocket ran for only three-and-a-half minutes, engineers continue to study the data it generated. Robert Demuth, a Ph.D. candidate in nuclear engineering in USC’s Molinaroli College of Engineering and Computing, recently completed a NASA-funded project to develop a high-fidelity multiphysics model of that original XE-Prime rocket engine.
“Changes in regulation and nuclear safety culture mean that the open-air testing conducted with the XE-Prime engine would not be permissible today,” Demuth says. “Future full-scale testing would require a dedicated ground-test facility to capture all radionuclides or an in-space demonstration, both of which are prohibitively expensive. That’s why high-fidelity modeling and simulation play an important role in reducing development costs and timelines.”
The model Demuth developed was benchmarked against experimental data from the XE-Prime testing and now provides a foundation for developing and evaluating new modeling methods for nuclear thermal propulsion, he says.
“That’s why high-fidelity modeling and simulation play an important role in reducing development costs and timelines.”
“It takes a lot of computational time and resources to run a high-fidelity multiphysics model,” says Demuth, noting that those constraints make it difficult to investigate, even virtually, the operational complexities of a nuclear thermal rocket engine. “To address this, we’re developing a reduced-order model designed to reproduce the behavior of the high-fidelity model at or near real-time speeds,” Demuth says. This will allow researchers to quickly study how different control and monitoring systems affect a nuclear thermal rocket engine.
The reduced-order model also provides the foundation for a future university-funded project to develop a real-time nuclear thermal rocket engine simulator, Demuth says. The simulator — similar to an aircraft flight simulator — would allow students to operate and interact with the rocket engine while learning about space mission design, orbital mechanics, thermal-hydraulics, reactor physics and systems analysis. The goal is to provide an interactive learning experience that fosters interest in nuclear engineering.
Demuth, a native of Gray Court, South Carolina, began doing undergraduate research in his first semester in the lab of Travis Knight, chair of the mechanical engineering department and graduate program director of nuclear engineering. He went on to earn a master’s in nuclear engineering in 2025 before continuing his doctoral research.
In addition to multiphysics modeling and simulation of nuclear thermal rocket engines, Demuth has been involved in a range of experimental and computational projects such as material testing for aluminum-clad spent nuclear fuel; material characterization for Westinghouse ADOPT fuel; spent nuclear fuel drying; wireless sensor development and evaluation for spent nuclear fuel storage; and radiation shielding and systems analysis of transportable nuclear reactors.
During his time as a graduate student, Demuth has worked in the Idaho National Laboratory’s reactor physics department and currently works part time as a space nuclear systems engineer for Analytical Mechanics Associates, supporting NASA nuclear modeling and simulation efforts. With completion of his doctoral degree just a few months away, he’s considering career opportunities to continue working in nuclear modeling and simulation after graduation.