A multiscale method for mixed convective systems. Coupled calculations with ATHLET and OpenFOAM of the PHENIX NCT

The Generation IV International Forum proposed six innovative reactor concepts as most promising. One of those concepts is the sodium cooled fast reactor (SFR). Its research and development has long history and shows high potential to meet GEN-IV criteria. One of those is the PHENIX reactor build in France. Construction began in 1968 and connected to the grid in 1973. The small-scale prototype reactor was in full operation until 2004 and then mainly used for research on transmutation and accident scenarios. The final shutdown of the PHENIX reactor was in 2009. Before it was finally shut down, several final tests were planned and performed, including the natural convection test (NCT), which is used in this work for calculations of the primary circuit behavior. The NCT is used as a benchmark excercise for his work, more specifically as a blind test for system code qualification and its validation. These codes use a lumped parameter approach and are used to calculate transient behavior of system thermal-hydraulics (STH) of complete and complex systems like nuclear power plants are. In Germany, the Analysis of thermal-hydraulics of leaks and transients (ATHLET) code is being developed by the Gesellschaft fuer Anlagen- und Reaktorsicherheit (GRS) gGmbH. As all operating German power plants are cooled by water, ATHLET is used only with material properties of water. One part of the THINS (Thermal-Hydraulics of Innovative Nuclear Systems) project was to qualify light water reactor (LWR) STH codes for GEN-IV purposes. This work shows the extension of the ATHLET code to sodium through a multi-fluid approach. Computational fluid dynamics (CFD) is used in many domains of fluid dynamics' calculations. It provides high quality and high resolution results of the desired domain. Hence the computational effort - and therefore costs - is high and time consuming, a more optimized method is in focus of another part of the THINS project. Combining the effectiveness of a STH code with the high quality of CFD (where needed) is leading to coupled approaches. After an introduction, this work presents the modification of the ATHLET STH code for sodium applications. After that, the feasibility of the implementation is shown with a generated STH model of the PHENIX primary circuit. As far as possible by the NCT the implementations are assessed and discussed. Afterwards, the hot pool of the PHENIX primary circuit is modeled in CFD and calculated with the open source CFD toolbox OpenFOAM. The hot pool has been chosen, as it is one of three large volumes that is considered to have high three-dimensional effects that cannot be represented with STH only. The Reynolds Averaged Navier-Stokes (RANS) method is applied with a k-ε turbulence model. To merge the two different scales of system thermal-hydraulics and computational fluid dynamics, a coupling methodology is being developed and verified. Its implementation is shown and discussed with the PHENIX NCT. During the transient scenario, the coupled solution shows different behavior in comparison to the STH standalone calculation. This is due to strong three-dimensional effects taking place in the hot pool of the PHENIX primary circuit that cannot be captured with STH standalone. Differences between STH standalone and STH/CFD coupled calculations are discussed in that chapter. Another theoretical scenario is presented, which applies the verified and assessed coupling model and scheme to show the extrapolation capability of the coupling methodology. It can also be observed, when CFD shows uniform flow field behavior at small magnitude in vertical direction in the hot pool, that results between STH/CFD coupled calculation and STH standalone calculation are very similar.