Abstract
The natural or free convection flow in an open cavity filled with ethylene glycol-Al2O3 nanofluid has been integrated numerically using the multiple-relaxation-time (MRT) lattice Boltzmann method (LBM) with the Graphics Process Unit (GPU) acceleration. The applications of this type of problem frequently appear in opendoor ovens and refrigerators and in the cavities of central solar receptors. The ceiling and bottom walls of the cavity are thermally adiabatic. The right side of the cavity is fully opened while heat flux condition is imposed on its opposite wall. The thermal conductivity of the nanofluid varies with fluid temperature. Numerical simulations have been conducted for a fixed Prandtl number, Pr = 16.6, with varying other controlling parameters, i.e., the Rayleigh number from Ra = 105 to 107, and nanoparticle volume fraction from 0 to 5%. The streamlines, isotherms, mid-line temperature and velocity distributions, and the rate of heat transfer regarding the Nusselt number have been presented in this article. The isolines and tabular form give the local and total entropy generation due to the fluid friction and temperature gradients. The tangential and normal velocity components grow dramatically as Ra increases; however, the temperature diminishes with Ra. Also, the locally distributed Nusselt number increases in magnitude with nanofluids' volume fraction and Rayleigh number. The total entropy enhances significantly for the augmentation of the nanoparticles volume fraction, and the entropy due to fluid friction dominates over the entropy generated by the temperature gradients.