Characterization of thermal stratification using validated liquid metal pool mixing models

Abstract

Weakly stratified thermal fronts in the upper plena of liquid metal cooled reactors can pose prediction complexities for safety analyses. The strong thermal diffusivity of the coolant changes gravity’s influence on the flow by reducing horizontal density gradients at time scales faster than gravity can act on them. While damped out thermal behavior results in stratification, the underlying convection currents continue contributing to enhanced mixing. Experimental turbulence spectral information, captured by distributed velocity (ultrasonic) and temperature (fiber-optic) measurements, is used to construct microscale parameters and quantify the amount of enhanced vertical thermal diffusivity in a scaled gallium surrogate of a pool type sodium fast reactor upper plenum. These constructed microscale parameters are used to validate empirical models for use in liquid metal pools. This experimental data and, separately, the validated models, are fed into a simple 1-D transport model to quantify the impacts to reactor safety and allow for testing of the efficacy of using the more easily available bulk parameters (i.e., core barrel Reynolds and Richardson numbers) to estimate the enhanced diffusivity. The 1-D model represents the experimental fronts’ progressions most closely after accounting for these characteristics of mixed convection unique to liquid metals.