---

In this paper, natural convective heat transfer of nanofluids in a uniform magnetic field between the square cavity and inner cylinder, was simulated via Lattice Boltzmann Method. The inner cylinder in square shape, diamond, and circular has been examined. Square cavity walls and inner cylinder surfaces are at a constant cold and warm temperature, respectively. The flow, temperature, and magnetic field is calculated with solving flow, temperature, and magnetic distribution functions simultaneously. D2Q9 lattice arrangement for each distribution function is used. The results clearly show the behavior of fluid flow and heat transfer between the cavity and the cylinder. The results have been validated with available valid results showing relatively good agreement. The effects of Rayleigh number, Hartmann number, void fraction and type of nanoparticles on natural convective heat transfer are investigated. This study shows that for all three geometries used with the same void fraction, type of nanofluid, and Rayleigh number, natural convective heat transfer decreases with Hartmann number. Also, when Hartmann number was had fixed, natural convective heat transferwas increased with Rayleigh number. Thus, to select the right geometry for optimum natural convective heat transfer, our needs to pay special attention to Hartmann and Rayleigh numbers. In addition, viod fraction and type of nanofulid can affect heat transfer directly.

---

Numerous models have been proposed to incorporate various equations of state (EOS) into the pseudo potential model. This paper presents an investigation of different EOS types based on the Gong and Cheng model in multiphase-single component flows by the lattice Boltzmann method. Primarily, it is conducted to investigate eight EOS’s classified in four categories; the Shan- Chen EOS, the cubic EOS, the non-cubic EOS, and the cubic and non-cubic combination EOS. The results show that each EOS type results in producing relatively similar spurious currents and has a maximum achievable density ratio. Although by choosing a proper beta parameter for every EOS the simulation errors decrease dramatically, our results show it is impossible to set a constant parameter for the non-cubic EOS. Therefore, a new equation is introduced to predict an efficient beta for the cubic and the Shan- Chen EOS’s. It is also found that the non-cubic, cubic, and non-cubic and cubic combination EOS’s have a wider temperature range and larger density ratios respectively. Hence, we determine a temperature dependent function for the beta parameter prediction instead of using a fixed value for the non-cubic EOS. The results are noticeably in better agreement with those of the Maxwell construction (theoretical results).