| Popis: |
PhD (Nuclear Engineering), North-West University, Potchefstroom Campus The fourth-generation nuclear reactors have many safety features, including the reactor cavity cooling system (RCCS). This system is intended to remove decay heat from the reactor cavity. Designs exist for both air-cooled and water-cooled configurations. It operates completely passively and is designed to operate without external inputs. The accurate simulation of the phenomena in a reactor cavity cooling system is important for design calculations, to ensure reactor designs that are passively safe. The codes that are used to simulate the phenomena must therefore be verified and validated. One method to validate both a particular design and simulation code is to build an experimental facility based on a prototype. In many cases scale models are constructed of prototype designs. The facility constructed at the University of Wisconsin (Madison) is an example of a scale model of an RCCS. The facility represents a 5° radial slice of the RCCS of the modular high temperature gas-cooled reactor (MHTGR) design. The prime focus of the study is the heated section/heater box of the facility. A study was done by Lisowski (2013) which encompassed single and two-phase experiments as well as a simplified version that was numerically simulated. This study aims to expand on the single-phase simulations done by Lisowski, by making fewer geometry simplifications. The study focussed on simulating the heated section of the UW RCCS facility by using 3D computational fluid dynamics (CFD) and simulating the full loop by using 1D system CFD. The 3D simulations explored different geometric configurations, namely circular riser tubes with and without fins, square tubes with and without fins and rectangular tubes with and without fins. The heat fluxes and temperatures on the surfaces were examined and compared between the different geometry configurations. The 3D CFD code was also coupled with a 1D systems CFD code. The systems were coupled such that the 1D and 3D code would transfer boundary conditions. The codes were linked such that each respective code’s strength could be exploited for the best combination of accuracy and effective use of the computational resources. Two variations were simulated and similarly to the 3D cases, the heat fluxes and temperatures on the various surfaces were examined for comparison. The limitations of the study mean that the results can’t be used as a direct comparison with the experiments but can nonetheless be used to philosophically understand the phenomena that occur within the facility. The finned configurations shielded the back wall from direct radiation heat transfer from the heater wall, which is preferable in terms of preserving the structural integrity of the reactor cavity wall. The cases without fins resulted in higher temperatures on the back walls, but more uniform temperature distributions around the circumference of the riser tubes. This in turn resulted in lower stresses due to uniform thermal expansion. Radiation heat transfer was the dominating mechanism, but the convection heat transfer nonetheless contributed a non-negligible fraction of the heat transfer. The convection heat transfer coefficients were calculated on the surfaces in the air-filled cavity and for the inner surfaces of the water-cooled riser tubes. Two variations of the coupled methodology were tested, and the wall temperatures compared well between the coupled methods as well as with the 3D method. The convection heat transfer coefficients were also compared for the inner surfaces of the middle riser tube for the three approaches. The coefficients compared reasonably well in terms of the trends that were exhibited by the different approaches on the four inside surfaces. Doctoral |
