Lattice Boltzmann method simulated effect of nanoparticle size on natural convection patterns of nanofluids

In this work, numerical simulation of natural convection of nanofluids within a square enclosure are conducted by using the non-dimensional lattice Boltzmann method (NDLBM). The effects of key governing parameters Knudsen number (10(-6) <= Kn(f,s) <= 10(4)), Rayleigh number (10(3) <= Ra-f,Ra-L <= 10(6)), and nanoparticle volume fraction (10(-2) <= phi(s) <= 10(-1)) on the heat and mass transfer of nanofluids are discussed. The results show that in the low conduction dominated regime, the nanoparticle size has little effect on heat transfer, whereas in the high Ra-f,Ra-L convection dominated regime, larger nanoparticle size significantly enhances flow intensity and heat transfer efficiency. For fixed Ra-f,Ra-L and phi(s), the heat transfer patterns change from conduction to convection dominated regime with Kn(f,s) increasing. The influence of nanoparticle volume fraction is also investigated, and in the convection-dominated regime, the maximum heat transfer efficiency is achieved when phi(s) = 8%, balancing thermal conduction and drag fore of nanofluid. Additionally, by analyzing the full maps of mean Nusselt number ((Nu) over bar (f,L)) and the enhancement ratio related to the base fluid (Re-n,Re-f), the maximum value of (Nu) over bar (f,L) and Re-n,Re-f occur when the nanoparticle size Kn(f,s) = 10(-1) is for both conductive and convection dominated regime. To ascertain the effects of all key governing parameters on (Nu) over bar (f,L), a new empirical correlation is derived from the numerical results, providing a more in-depth insight into how these parameters influence on heat transfer performance.