Grid refinement studies of fully developed turbulent flow using second order turbulence models

Abstract

Paper presented at the 7th International Conference on Heat Transfer, Fluid Mechanics and Thermodynamics, Turkey, 19-21 July, 2010.

Estimation of errors involved in the numerical prediction of fluid flow and heat transfer problems is an essential element in CFD. The errors involved in the numerical simulations can be categorized into two groups: modeling errors and numerical errors. For complex flows such as the turbulent flows, models which are simplified mathematical representation of the flow are prone to errors in realizing the flow behavior. Numerical errors are caused by the discretization of the governing equations and approximations of boundary conditions, non­-convergence of iterative solution procedures, and round-off errors. It is crucial to minimize the numerical errors and evaluate the total errors precisely. The main objective of this paper is provide benchmark values for investigation of new turbulence models and to investigate the modeling and numerical erros encountered in the numerical simulation of the fully developed plane-channel flow using RSM with LRR and SSG pressure- strain second order turbulence models. Mesh refinement studies are reported with a fourth-order extrapolation scheme for flows with Reynolds numbers of 50000, 200000, and 500000. This study exclusively investigates the performance of staggered and colocated variable arrangements in providing stronger stress-strain rate couplings. The near wall region is overpassed by implementation of the standard wall-function method. The results obtained are compared against the Laufer's experimental data and DS numerical results which enables evaluation of the accuracy of the second order turbulence models and grid arrangements as well as near-wall grid point location quantitatively for turbulent flows