Non-adiabatic annular flow experiments in a subchannel geometry with spacer grid

Abstract

Boiling annular flow in a subchannel geometry with functional spacers is studied using X-Ray tomography in the Dryout Tomography Experiment (DoToX) facility at ETH Zürich. The liquid film distribution on the walls and the liquid holdup within the gas core of the annular flow in the subchannel, both under the influence of the spacer vanes, are extracted from tomographic images. The subchannel cross section consists of a main subchannel in the center surrounded by four heating flow sections representing fuel rods and connected to four partial subchannels with simplified geometry. This effectively removes unnatural no-slip boundary conditions at the narrow gap of the main subchannel.Chloroform is used as a working fluid because of its low enthalpy of evaporation and a low boiling point around 60 °C close to atmospheric pressure. A liquid heating system with water at slightly above 100 °C provides the heating power for the boiling process. This ensures continuous operation of the facility even in case of a dryout scenario, avoiding structural damages by imposing a maximum system temperature. The high thermal capacity of the water allows for an almost isothermal boundary condition. Although not identical to real BWR settings, this experiment is well suited to generate data for model validation.Originally designed for tomographic imaging with both cold neutrons and X-rays, this paper focuses on results obtained by X-rays. Since the X-ray contrast between the working fluid chloroform and the structural material of the channel, aluminum, is very small, differential imaging with adaptive centering of reconstructions with and without the liquid film were required. The measurement campaign consisted of three measurements with constant chloroform mass flows at 5.00, 4.05 and 3.25 kg/min and constant heating liquid temperature of 100–104 °C, in addition to the empty channel reference measurement. The thermal power transferred to the working fluid under these conditions was about 10 kW of heating power. The resulting liquid film thicknesses (LFT) upstream of the spacer decreased from 91 to 49 µm. The effect of the spacer vanes is visualized and quantified over an axial section of 10 hydraulic diameters. Reference samples of the film thickness and the liquid holdup were taken upstream of the spacer. Dryout was nearly reached in characteristic regions of reduced film thickness due to turbulent shear stresses, while other regions show a significant local increase of the liquid film thickness. The deposition of about 30% of the liquid holdup from the gas core contributes to the increase. The results are consistent with previous adiabatic experiments with similar geometries.