CORA (COmputational Methods for Reactor Analysis) centers on developing methods and computational codes for analyzing neutron behavior within nuclear reactor cores, supporting both reactor design and performance evaluation. This includes simulating neutron transport and interactions, optimizing core configurations (fuel assembly arrangement, control rod placement, etc.), and assessing key performance parameters like reactivity, power distribution, temperature profiles, transient behavior, and safety metrics. We focus specifically on code development, verification, and validation using experimental data and benchmarks. Our research also explores advanced numerical methods and models to improve the accuracy and efficiency of reactor simulations, covering both steady-state and transient analyses and potentially incorporating coupled multi-physics simulations.
CORA is dedicated to advancing computational methods for reactor physics and enhancing the analysis and simulation of nuclear systems. Our work involves developing innovative methodologies and computational tools to model and understand neutronic behavior within nuclear reactor cores, supporting accurate and efficient reactor analysis and design.
The broader research streams are:
Computational Reactor Physics
Methods for Neutron Transport and Diffusion Analysis
Acceleration of Neutronic Codes
Computational Intelligence in Nuclear Engineering
Hornet focuses on developing GPU-oriented methods for efficient nuclear data processing, enabling continuous-energy Monte Carlo reactor simulations on GPUs. It provides fast handling of cross-sections and ACE sampling for incident neutrons and thermal scattering, while remaining Monte Carlo–agnostic for broad applicability. Implemented in CUDA C++ with an object-oriented design, Hornet achieves significant speedups without compromising accuracy. Verification against MCS and integration with GREAPMC across standard benchmarks confirmed sub-0.1% deviations in power distributions.