Full Field Reconstruction and Uncertainty Quantification of Particle Image Velocimetry Measurements in A 5X5 Rod Bundle with Mixing Vane Spacer Grids

Abstract

One of the most commonly used methods for quantifying fluid velocity profiles is particle image velocimetry (PIV). This non-invasive measurement technique employs seeding particles in a transparent simulant fluid flowing through a geometry of interest at Reynolds number (Re) in the regime of what is expected during operation. A laser sheet is projected through the fluid to illuminate the particles, and two subsequent photographs of the particle-seeded fluid are taken with a ∆t recorded between the two images. Displacement of the particles from one image to the other, δi, which is calculated using a generalized cross correlation, along with the known ∆t, are used to determine a velocity of a fluid element in the laser plane. Efforts to expand this planar two-dimension two-component (2D2C) method to a volumetric three-dimension three-component (3D3C) PIV have been pursued, which require large computational resources for processing. This study aimed to quantify a 3D3C velocity measurement in a prototypical pressurized water reactor (PWR) geometry of rods with a mixing vane spacer grid. A new matched-index-of-refraction (MIR) facility for rod bundle and spacer grid testing was developed and constructed within the Thermal-Hydraulics Laboratory of the Nuclear Engineering Department at Texas A&M University (TAMU). The facility was designed to overcome the challenge of producing high fidelity and low uncertainty data, especially near the mixing vanes, where data is sparsely available. This investigation explores an innovation in MIR fluid and presents results for a full field 3D3C velocity measurement at Re=27,390. The full field measurements are constructed via trilinear interpolation from multiple 2D2C PIV measurements. These 3D3C ensemble-averaged velocity, root-mean-square (RMS) fluctuating velocity, Reynolds stresses, and vorticity fields are presented downstream of a prototypical PWR spacer grid in a 5×5 rod bundle. A full field uncertainty quantification (UQ) of the results is also provided. Profiles taken at different subchannel positions having similar geometrical features are compared to demonstrate coherence of the results and quality of the MIR for improved PIV. New data of the complex flow near and within the mixing vanes region of the spacer grid could be measured with the new methodology.