Experimental analysis of natural convection in a differentially heated cavity

Abstract

Turbulent heat fluxes (THFs) estimation is of paramount importance in the determination of heat transfer in fluids. Numerical models for low Prandtl number fluids are still unreliable and their experimental evaluation is a challenging task since it requires simultaneous measurement of fast velocity and temperature fluctuations. In nuclear applications, a better understanding of THFs in liquid metals could lead to more precise predictions of the primary coolant temperature for the evaluation of nominal operation (forced convection regime) and accidental conditions (mixed and/or natural convection regime). The first part of this work focuses on the selection of the measurement techniques suitable for water, GaInSn and LBE and the thorough literature review required. Tests in different setups led to the choice of sheathed type K thermocouples and fiber Bragg gratings for temperature measurements and Ultrasound Doppler Velocimetry and Hot Wire Anemometry for velocity measurements. The comparison carried out among the different techniques underlines advantages and limitations of each of them. Calibration of each technique is performed and cross-effects of temperature and velocity are evaluated. Uncertainty analyses are also carried out. To conclude, first results obtained in a differentially heated cavity made of stainless steel 316L with an edge of 60 mm are presented. DNS numerical simulations are performed to know the ranges of the quantities to be measured and to have results available for comparison with experiments.

Keywords

• natural convection
• differentially heated cavity
• turbulent heat flux
• nuclear power reactors and nuclear fuel cycle