Computational reactor physics
Reactor physics is the study of how neutrons move and distribute themselves in a nuclear reactor, with the goal of understanding, predicting, and controlling the fission chain reaction. It is essential because reactor efficiency and safety both depend on accurate models of neutron behavior and feedback mechanisms. Research in this field spans topics such as neutron transport and diffusion, reactor dynamics, criticality safety, fuel cycle analysis and reactor design.
Scientific machine learning/artificial intelligence
In the context of nuclear engineering research, scientific machine learning and artificial intelligence can aid in the incorporation of physics-based simulation and experiments into predictive modeling for various aspects of reactor operation. These techniques are particularly useful when applied to complex systems with real-time computation constraints. They also tend to provide a particular emphasis on the use of existing physical principles to enhance the accuracy and reduce the training burden of data-driven models. Research in this field spans topics such as surrogate modeling, active learning, time-series modeling and feature extraction.
Multiphysics simulation
Multiphysics simulation focuses on the coupled modeling of neutronics, thermal hydraulics, fuel performance and structural mechanics to capture the complex interactions between these physical phenomena that ultimately govern reactor behavior. As each physical phenomenon may occur on different spatial domains or timescales, the development of coupling methods for capturing the true behavior of the reactor system are nontrivial. Research in this field spans topics such as stability analysis, data transfer and acceleration methods.
Reactor dynamics and control
Reactor dynamics and control is a specific field of nuclear engineering research that focuses on the characterization and manipulation of the temporal behavior of nuclear reactors. Although certainly linked to multiphysics simulation, in this context, work in this area will include more applied aspects of nuclear engineering. Design basis transients such as startup, shutdown and load following as well as accident transients form a critical route of analysis for advanced reactors. Control algorithms that enhance the safe operational flexibility of reactors are also of interest. Research in this field spans topics such as model predictive control, instability identification, digital twinning, accident analysis and control element characterization.
Verification and validation
Verification and validation are the systematic processes of ensuring that computational models accurately capture both the intended mathematical modeling approach as well as the real-world behavior of the subject of the simulation. Given the complexity and safety importance of the reactor system, a significant importance has been placed on this area guarantee the relevance of advanced simulation frameworks to physical reactor systems. Historically, this area has been primarily focused on the evaluation of physics-based modeling and experimentally-derived data but can be expanded to ensure the dependability of innovations associated with data-driven modeling approaches. Research in this area spans topics such as uncertainty quantification, sensitivity analysis and benchmarking.
