Volumetric Measurements of Mean Velocity Vector and Reynolds Stress Tensor for CFD Validation: Magnetic Resonance Velocimetry in a Nuclear Fuel Assembly Model with Mixing Grids

Abstract

Experimental studies in industrial-relevant geometries are of great value for validating computational fluid dynamics (CFD). This study provides such data using Magnetic Resonance Velocimetry (MRV) in a replica of the single-phase and isothermal OECD/NEA-KAERI rod bundle benchmark exercise based on the MATiS-H test facility at the Korea Atomic Energy Research Institute (KAERI). The geometry is a 5 × 5 nuclear fuel assembly model of a pressurized water reactor with a split-type mixing grid inducing a swirling flow in each sub-channel. The Reynolds number based on the hydraulic diameter is 50,250. Recent studies demonstrated that MRV enables a comprehensive validation of CFD results in industrial-relevant test cases by providing time-averaged, three-dimensional measurement data from complex opaque structures. Nevertheless, there was still some potential left to improve the accuracy of the measurement. This study uses a newly developed MRV method to accurately measure the mean velocity vectors and the Reynolds stress tensor in three dimensions. The measurement volume reaches from shortly upstream to 10 times the hydraulic diameter downstream of the mixing grid. The estimated mean measurement uncertainty of the velocity data is 1.9% based on the bulk velocity of 1.72 m/s. In the case of the Reynolds stress data, the estimated mean uncertainty for each component is between 0.7 and 1.8% based on the square of the bulk velocity. The comparison to previously published Laser Doppler velocimetry measurements confirms the high accuracy of the reported 3D MRV data. The study includes a detailed description of the technique and boundary conditions. The measurement data is available to interested parties upon request.