Direct Numerical Simulation of Low and Unitary Prandtl Number Fluids in Reactor Downcomer Geometry

TL;DR

This study uses direct numerical simulation to analyze low and unitary Prandtl number fluid convection in reactor downcomer geometries, revealing complex heat transfer behaviors influenced by buoyancy and flow parameters.

Contribution

It introduces detailed DNS analysis of mixed convection at low Prandtl numbers, highlighting the effects of buoyancy and flow parameters on heat transfer in reactor geometries.

Findings

Buoyancy alters velocity boundary layers and fluctuation intensities.

Convective heat transfer can be enhanced or impaired depending on parameters.

Large convective structures are suggested in transition regions from spectral analysis.

Abstract

Buoyancy effect on low-flow condition convective heat transfer of non-conventional coolants, such as liquid metal and molten salts, is a crucial safety factor to advanced reactors under transient or accidental scenarios. The distinct heat transfer characteristics of non-unitary Prandtl fluids and the inherent complexity of the low-flow mixed convection phenomena requires the development of novel turbulent and heat transfer models that are adaptive to different spatiotemporal scales involved in the mixed convection heat transfer. In this work, direct numerical simulation of low-flow mixed convection is carried out at low-to-unitary Prandtl numbers that are of industrial interest. Time-averaged statistics, turbulent Prandtl number, as well as time signals are analyzed to investigate mixed convection phenomenon. From the time-averaged statistics, buoyant plume altered velocity boundary layer as well as the intensity of the fluctuation near both walls and channel centerline. Buoyancy effect also rendered different degree of convective heat transfer enhancement and impairment depends on Prandtl and Richardson number. Analysis of time series was conducted on the sodium mixed convection case to emphasize on the low-Pr mixed convection behavior at transition region. Resulting power spectra density and wavelet spectrogram suggests possible large convective structure in transition region. Future work will focus on providing broader data coverage on Pr-Re-Ri parameter space to facilitate more comprehensive analysis of mixed convection.