Reduced-resolution RANS approach to grid spacer fuel assemblies

Abstract

One of the main issues in the context of the safety assessment of liquid metal-cooled reactors is flow blockages in fuel sub-assemblies. CFD modelling may be used to predict the temperature and velocity fields inside a fuel assembly under blocked and unblocked conditions. However, the computational cost requirements make it infeasible to model the complete assembly using the resolved RANS, LES or DNS approaches. To this end, a reduced-resolution RANS approach of an assembly is presented and validated in this paper.In order to reproduce and compare with experimental data, the 19-pin hexagonal fuel assembly of the NACIE-UP facility at ENEA in Italy has been selected for the present simulation study. For the flow blockage, a 6-pin edge-type blockage is selected similar to that used in the experimental facility. During the pre-test simulation study, wall-resolved RANS simulations are performed for the unblocked flow conditions. Using the wall-resolved RANS simulations as reference, a mesh sensitivity is performed by subsequently reducing the wall and bulk resolutions. As the blocked flow condition requires transient simulations, the resolution study is only performed for the unblocked flow condition to reduce the computational effort. It is shown that the approach may be used to reproduce the temperature and velocity fields similar to that in wall-resolved simulations within tolerance limits, but at a fraction of the computational cost.Experimental data for unblocked and blocked flow conditions has been made available for the Blocked Fuel Pin Simulator (BFPS) section at the NACIE-UP facility. These experimental data are used to validate the reduced-resolution approach in the post-test analysis. The difference between the present results and the experimental data is attributed for and quantified. Consistent with the literature, two main effects of flow blockages are identified. A local temperature hotspot is observed in the wake of the blockage caused by the coolant recirculation zone. The global effect of temperature peak is observed at the end of the active length downstream of the blockage, caused by the lower mass flow rates in the blocked sub-channels. For the unblocked flow condition, the maximum difference compared to the experimental data is estimated to be 11%. While the error in prediction of the local temperature hotspot and peak temperature at the end of active length is estimated to be 25% and 12%, respectively, for the partially-blocked flow condition. Finally, based on the 19-pin model, the feasibility of the reduced-resolution approach for the complete 127-pin assembly of ALFRED demo reactor is demonstrated. In addition to the unblocked flow condition, the approach is also used for blocked grid spacer conditions. Consistent with the results of NACIE bundle, the two main effects – local and global – of the flow blockage are also observed here. The local effect of temperature hotspot behind the blockage is observed to be dominant for the larger blockage, while the global effect of peak temperature downstream of the blockage is observed to be proportional to the size of the flow blockage.