In this paper, experimental and numerical studies of heat transfer in a test local of side $H=0.8\text{m}$ heated from below are presented and compared. All the walls, the rest of the floor and the ceiling are made from plywood and polystyrene in sandwich form ( $3\text{mm}$ plywood- $3\text{cm}$ polystyrene- $3\text{mm}$ plywood) just on one of the vertical walls contained a glazed door ( $2\text{H}/3\times 0.15\text{m}$). This local is heated during two heating cycles by a square plate of iron the width $L=0.6\text{H}$, which represents the heat source, its temperature $T_h$ is controlled. The plate is heated for two cycles by an adjustable set-point heat source placed just down the center of it. For each cycle, the heat source is switched “on” for 6 h and switched “off” for 6 h. The outdoor air temperature is kept constant at a low temperature $T_c<T_h$. All measurements are carried out with k-type thermocouples and with flux meters. Results will be qualitatively presented for two cycles of heating in terms of temperatures and heat flux densities $\phi$ for various positions of the test local. The temperature evolution of the center and the profile of the temperature along the vertical centerline are compared by two dimensions simulation using the lattice Boltzmann method. The comparison shows a good agreement with a difference that does not exceed $\pm 1$°C.

This work presents a simulation of the phenomena of natural convection in an enclosure with a variable heating regime by the lattice Boltzmann method (LBM). We consider a square enclosure of side H filled with air (Pr = 0.71) and heated from below, with a hot portion of length L = 0.8 H, by imposing a sinusoidal temperature. The unheated segments of the bottom wall are treated as adiabatic, and one of the vertical walls features a cold region, while the remaining walls remain adiabatic. The outcomes of the two-dimensional (2D) problem are depicted through isotherms, streamlines, the temperature evolution within the enclosure, and the Nusselt number. These visualizations span various amplitude values “a” in the interval [0.2, 0.8], and of the period T₀ for Ra = 10⁷. The amplitude and period effect on the results is evaluated and discussed. The amplitude of the temperature at the heart of the enclosure increases with the increase in amplitude. This also increases with the period (T₀) of the imposed temperature, something that is not observable on the global Nusselt number.