Sensitivity study regarding the grid refinement and the choice of SGS model coefficient is reported. Discussion is based on time-averaged simulation results; therefore, the sampling error is evaluated as well. This study shows that variation of grid resolution or the SGS model coefficient both lead to a rather small difference in the predicted mean wall temperature ((3%), but on the other side to a considerable difference (in the range of 10 %) in the calculated mean stagnation Nusselt number. The performed analysis of the LES multiple jet impingement case showed that presented LES results are sufficiently accurate to be used for benchmarking of statistical RANS models.
COBISS.SI-ID: 31988007
A significant improvement of numerical prediction with the SST-SAS model was observed compared to the standard SST-URANS model. The most important difference between these two models is the unsteady behaviour of jets predicted by the SST-SAS model, while the standard SST-URANS model completely suppresses the flow unsteadiness due to a too high eddy viscosity. Statistically averaged results obtained with the SST-SAS and SST-URANS models are very similar. However, a significant difference between these two models can be observed in the prediction of the resolved turbulence kinetic energy that has a key role at the heat transfer between the fluid and the wall. Both simulations were verified against the well-resolved LES results.
COBISS.SI-ID: 31968807
The accident scenario considered in this study considers that the cryostat in DEMO fusion reactor is filled with helium gas. This enables the heat transfer with natural convection between the cryostat thermal shield, which remains actively cooled, and cryostat, which is on the outside in contact with the atmosphere at room temperature. This can cause a significant cool-down of the cryostat walls. The open source computational fluid dynamics (CFD) code OpenFOAM is used to simulate natural convection of helium gas and to assess the temperature distribution inside the compartment between the cryostat thermal shield and the cryostat, and the temperature of the inner cryostat wall. The results are also compared to previous results obtained with the CFD code ANSYS CFX. Very high temperature differences between the lower and upper part of the cryostat wall were predicted. It was found that they can be significantly reduced by establishing a forced air convection on the outside surface of the wall.
COBISS.SI-ID: 33181223