What Rayleigh numbers are achievable under Oberbeck–Boussinesq conditions?

Abstract

The validity of the Oberbeck–Boussinesq (OB) approximation in Rayleigh–Bénard (RB) convection is studied using the Gray & Giorgini (Intl J. Heat Mass Transfer, vol. 19, 1976, pp. 545–551) criterion that requires that the residuals, i.e. the terms that distinguish the full governing equations from their OB approximations, are kept below a certain small threshold $\hat {\sigma }$ . This gives constraints on the temperature and pressure variations of the fluid properties (density, absolute viscosity, specific heat at constant pressure $c_p$ , thermal expansion coefficient and thermal conductivity) and on the magnitudes of the pressure work and viscous dissipation terms in the heat equation, which all can be formulated as bounds regarding the maximum temperature difference in the system, $\varDelta$ , and the container height, $L$ . Thus for any given fluid and $\hat {\sigma }$ , one can calculate the OB-validity region (in terms of $\varDelta$ and $L$ ) and also the maximum achievable Rayleigh number ${{Ra}}_{max,\hat {\sigma }}$ , and we did so for fluids water, air, helium and pressurized SF $_6$ at room temperature, and cryogenic helium, for $\hat {\sigma }=5\,\%$ , $10\,\%$ and $20\,\%$ . For the most popular fluids in high- ${{Ra}}$ RB measurements, which are cryogenic helium and pressurized SF $_6$ , we have identified the most critical residual, which is associated with the temperature dependence of $c_p$ . Our direct numerical simulations (DNS) showed, however, that even when the values of $c_p$ can differ almost twice within the convection cell, this feature alone cannot explain a sudden and strong enhancement in the heat transport in the system, compared with its OB analogue.