Effects of temperature-dependent properties on natural convection of nanofluids in a partially heated cubic enclosure

In this paper, the effects of temperature-dependent properties on natural convection of nanofluids in a partially heated cubic enclosure are investigated in detail with lattice Boltzmann method. To improve the computational efficiency, all simulations are performed on the Graphics Processing Unit (GPU) using NVIDIA's CUDA. The fluid in the cuibc cavity is a water-based nanofluid containing Al2O3 nanoparticles. The effects of thermal Rayleigh number (104⩽Raf⩽106), diameter of nanoparticle (25nm⩽ds⩽100nm), nanoparticle volume fraction (0.0⩽ϕ⩽0.04), temperature of the cooled sidewall (315K⩽Tc⩽335K), temperature difference between the sidewalls (10K⩽ΔT⩽50K), aspect ratio (0.50⩽AR⩽1.00) and heating location on temperature field and fluid flows are investigated. The results reveal that the average Nusselt number is decreased with the increase of nanoparticle volume fraction. In addition, it is also observed that there is an optimal volume fraction ϕmax at which the maximum heat transfer enhancement is obtained, and the value of ϕmax is found to increase slightly with decreasing the nanoparticle diameter, and to increase remarkably with increasing the temperature of Tc or ΔT. Moreover, we also find that the average Nusselt number and the heat transfer enhancement are decreased with the increase of aspect ratio.