Numerical Investigation on Effect of Turbulence Model Selection for Aerodynamic Prediction of Axial Flow Fan

Abstract

Axial fans are one of the major power consuming cooling systems in most of the devices being used in daily basis. So, it is very important to predict the aerodynamic performance of the fan design. One way of estimating the fan performance in full 3D scale is by using Computational fluid dynamics. Turbulence models plays an important role in application of CFD for Axial flow fans. It is important to choose proper turbulence model in order to get the most accurate results. In this study, a baseline model experiment was chosen and validated. Both steady and transient studies have been carried out. Steady state simulation was performed using Reynolds Averaged Navier Stokes turbulence models such as Spalart Allmaras, RNG k-ε with swirl dominant flow, k-ω SST model and Transition SST model. Transient simulation is performed using RANS model such as Spalart Allmaras, Standard k-ε with standard wall function, RNG k-ε with swirl dominant flow, k-ω SST model, Detached Eddy Solver and Standard k-ω with embedded Large Eddy Solver. Both studies reveal that with coarse mesh URANS model can predict more accurate than Hybrid URANS models. Out of all the models DES showed a very good prediction in terms of pressure drop.