Experimental Flow Visualization and Temperature Measurements for Validation in Hemispherical Upper Plenum and Pebble Bed Geometries

Abstract

Advanced reactor and Small Modular Reactor (SMR) systems present a wide range of compelling fluid flow behaviors and complex flow structures. The hemispherical plenum features heavily in high temperature gas cooled reactors (HTGRs) as well as liquid metal reactor systems. In a loss of flow (LOF) accident scenario in a HTGR, flow direction is reversed and is driven by natural convection upward from the core. Liquid metal reactor vessels are subject to high thermal gradient and fluid level oscillations, which cause extreme thermal stresses and fatigue failure of the hemisphere wall. As such, a hemispherical upper plenum test section was chosen to provide necessary high-fidelity experimental data for simulation and modelling validation. Experiments have been conducted to measure the flow field and temperature data within the plenum and along its surface in buoyancy driven and thermally stratified states.

Pebble bed reactors are associated with a high degree of complexity in their flow and heat transfer characteristics. This is primarily due to the packing of the spherical fuel elements having a great degree of variance throughout the bed. In order to provide useful data for simulation validation in these systems, it is critical to not only provide high fidelity measurements, but also to provide the geometry on which the experiment was conducted for direct comparison of experiment and simulation data. Experiments were conducted to provide flow field and temperature data within an isothermal and non-isothermal packed bed as well as a three-dimensional reconstruction of the geometry by synchronized laser scanning.

Experimental methods include particle image velocimetry (PIV) for flowfield measurement, planar laser induced fluorescence (P-LIF) for fluid temperature measurement, and distributed temperature sensing (DTS) using optical fiber for surface temperature measurement. This data, and subsequent reduction employing Proper Orthogonal Decomposition (POD), allows for comparison of a wide range of parameters between experimental and simulation data including multiscale-multiphysics modelling techniques for non-Light Water Reactor (LWR) advanced reactors.