| The motivation for the present study is to develop a series of interconnected virtual simulations of simple, operating nuclear reactors. Once complete, the virtual reactor should have the ability to computationally simulate neutron physics, the physics of fluids, the properties of materials, and reactor control systems. The existence of a virtual reactor will allow for the development and improvement of terrestrial research reactors and space-based reactors. In the present study, a series of modules and methodologies were developed which focus on neutron physics and the physics of fluids. Furthermore, these modules were benchmarked against experimental data and theoretical calculations for the PUR-1 research reactor and a representative space reactor, respectively. A novel technique using neural network interpolation was developed to find reactivity change as a result of control rod movements. Point kinetics was then used to find the time dependent behavior of the reactor following the reactivity change. Additionally, 2D CFD models of the PUR-1 research reactor and a representative space reactor were developed, coupled to the neutron physics codes AGENT and MCNP5, and tested against existing data. The overall agreement between the experimental data, theoretical calculations, and computational models was good; however, it was noted that available data for two-phase fluid behavior is limited. |
