Scale-adaptive simulation of turbulent mixed convection of nanofluids in a vertical duct

Abstract

The present paper focuses on the turbulent mixed convection of nanofluids through a vertical square duct. The prediction accuracy of scale-adaptive simulation (SAS) approach is investigated versus RANS-based models (k − e and k − ω), in terms of Nusselt number and friction factor. A thermal-dependent model is considered to determine the effective thermal conductivity and effective dynamic viscosity of nanofluids. The present numerical simulations are performed for CuO–water and SiO2–water nanofluids and compared with various experimental data. Results indicate that the SAS approach can predict the unsteady flow and heat transfer of nanofluids more accurately than the k − e and k − ω models. Moreover, it is found that the turbulent velocity fluctuations enhance in streamwise, spanwise and wall-normal directions with an increasing nanoparticle volume fraction, whilst this increment is higher in streamwise direction. Also, in the near-wall region the effect of the presence of nanoparticles on the turbulent velocity fluctuations is more considerable, which increases the turbulence content of the flow field.