Lattice Boltzmann modeling of convective flows in a large-scale cavity heated from below by two imposed temperature profiles

Abstract

PurposeThe purpose of this paper is to model the convective flows in a room equipped by a glass door and a heated floor of length l = 0.8 × H and submitted to a sinusoidal temperature profile and mono alternative temperature profile.Design/methodology/approachThe paper opts for a numerical study of convective flows in a large scale cavity using the Lattice Boltzmann Method (LBM) by considering a two dimensions (2D) square cavity of side H and filled by air (Pr = 0.71). All the vertical walls, the ceiling and the rest of the floor are thermally insulated, the hot portion of length l = 0.8×H is heated with two imposed temperature profiles of amplitude values 0.2 ≤ a ≤ 0.6 and for two different periods ζ = ζ0 and ζ = 0.4×ζ0. One of the vertical walls has a cold portion θc = 0 that represents the glass door.FindingsA systematic study of the flow structure and heat transfer is carried out considering principal control parameters: amplitude “a” and period ζ for Rayleigh number Ra = 108. Effects of these parameters on results are presented in terms of isotherms, streamlines, profiles of velocities, temperature in the cavity, global and local Nusselt number. It has been found that an increase in amplitude or period increases the amplitude of the temperature in the core of cavity. The Nusselt number increases when the amplitude “a” of the imposed temperature increases, but this later is not affected by variation of the period.Originality/valueThe authors used LBM to simulate the convective flows in a cavity at high Ra, heated from below by tow imposed temperature profiles. Indeed, they simulate a local equipped by a solar water heater (SWH). The floor is subjected to a periodic heating: Sinusoidal heating (Case 1) for which the temperature varies sinusoidally (SWH without a supplement), and mono alternation heating (Case 2), the temperature evolves like a redressed signal (SWH with a supplement). The considered method has been successfully validated and compared with the previous work. The study has been conducted using several control parameters such as the signal amplitude and period in the case of turbulent convection. This allowed us to obtain a considerable set of results that can be used for engineering.