Nuclear reactor containment flows - Modelling of stably stratified layer erosion by a turbulent jet

Abstract : A number of international benchmarks were devoted to revealing the capability of CFD codes to predict the temporal evolution of the concentration and velocity fields of the nuclear reactor containment atmosphere in the course of severe accidents. In the most recent OECD/NEA international benchmark exercise on containment flows, a stably-stratified helium-air layer was eroded by a free turbulent jet coming from below. Velocity and helium concentration fields were measured in the course of the experiment. The results of the benchmark have shown that a correct prediction of the temporal development of the concentration field does not necessarily mean that the velocity field was resolved accurately as well. This can suggest that a wrongly predicted velocity field can compensate an erroneously modeled mass transport, still leading to a relatively correct concentration field.This work examines numerically the temporal evolution of the velocity and concentration fields for the conditions of an international benchmark exercise on containment flows performed in PANDA facility at PSI, Switzerland. A number of preliminary separate effect studies are performed on the way to choosing the final modeling scheme. It is shown that $k$-$\omega$ SST models significantly overestimate the mixing rates, whereas the standard $k$-$\epsilon$ model overestimates the spreading of the jet and its center-line velocity decay rate. A good compromise seems to be found in modification of the $C_{1 \epsilon}$ constant of the $k$-$\epsilon$ model allowing to simulate erosion of the stratified layer by round jets more reliably.