Prediction of the Drag Coefficient of a Rectangular Cylinder using four discretization schemes

ABSTRACT Predictions are reported of the two-dimensional turbulent flow around a rectangular cylinder at high Reynolds numbers. Computational Fluid Dynamics (CFD) is widely used to investigate heat transfer, fluid flow, chemical reaction and mass transfer phenomenon. While solving the Navier-Stokes equations, the convection term is always prone to numerical instability and therefore the discretization of the convection term needs special attention. The performance of different convection schemes had been previously assessed on one-dimensional convection-diffusion problem. Nevertheless, the numerical errors of these convection schemes are more pronounced in higher-dimensional problems especially those involving pressure term and flow recirculation. In this report, the performances of four convection schemes, namely, the first order upwind differencing, hybrid differencing, Quadratic Upstream Interpolation for Convective Kinematics (QUICK) and power-law, are investigated on the flow past a rectangular cylinder of aspect ratio 5:1. Two turbulence closure models have been used to carry out computations. The computations were performed using EasyCFD software and the preliminary results were validated using research data on the fluid flow past rectangular cylinders with sharp corners before extending the present computational research. The effects of Reynolds number, four discretization schemes and two turbulence models on the drag coefficient (CD) of the cylinder are studied here. The consistency, efficiency and accuracy of the results due to different convection schemes, along with two turbulence models, are compared. It is found that although the QUICK scheme performed better than other schemes in predicting the various integral parameters and in terms of its convergence rate is the lowest. Keywords – Drag coefficient, turbulent flow, Rectangular Cylinder, Turbulence models, Discretization schemes.