LES of Twin High-Re Rectangular Jets for Benchmarking Model Order Reduction Methods

Abstract The study of turbulent mixing in non-isothermal coolant streams is crucial in understanding thermal striping, an oscillatory mixing pattern that can lead to thermal fatigue and degradation of internal components in advanced nuclear reactors. Thermal striping is linked closely to local 3D fine scales associated with turbulence. Historically, Large Eddy Simulation (LES) or Direct Numerical Simulation (DNS) were required to resolve these scales. Our investigation centers on the mixing dynamics within the Reactor Cavity Cooling System (RCCS) separate effects test facility. We study the interaction of two parallel-planar jets within a confined plenum to uncover flow mixing characteristics. Initially, LES simulations are employed to generate velocity and temperature statistics and time series data for comparison against future ROM approaches. Analysis of the velocity time series data using Power Spectrum Density (PSD) reveals a distinct low frequency mixing mode indicative of thermal striping. Subsequently, we utilize Proper Orthogonal Decomposition (POD) to extract dominant flow structures from 600 high-fidelity instantaneous velocity snapshots. These POD modes are utilized to construct a reduced-order model (ROM) capable of replicating the low frequency mode associated with thermal striping. By leveraging LES data with ROM techniques, we can efficiently and accurately predict the low frequency mode.