Analysis of the K-Epsilon Turbulence Model


Book Description

Aimed at applied mathematicians interested in the numerical simulation of turbulent flows. Centered around the k-&epsis; model, it also deals with other models such as one equation models, subgrid scale models and Reynolds Stress models. Presents the k-&epsis; method for turbulence in a language familiar to applied mathematicians, but has none of the technicalities of turbulence theory.













An Improved K-Epsilon Model for Near-Wall Turbulence and Comparison with Direct Numerical Simulation


Book Description

An improved k-epsilon model for low Reynolds number turbulence near a wall is presented. The near-wall asymptotic behavior of the eddy viscosity and the pressure transport term in the turbulent kinetic energy equation is analyzed. Based on this analysis, a modified eddy viscosity model, having correct near-wall behavior, is suggested, and a model for the pressure transport term in the k-equation is proposed. In addition, a modeled dissipation rate equation is reformulated. Fully developed channel flows were used for model testing. The calculations using various k-epsilon models are compared with direct numerical simulations. The results show that the present k-epsilon model performs well in predicting the behavior of near-wall turbulence. Significant improvement over previous k-epsilon models is obtained. Shih, T. H. Glenn Research Center NASA-TM-103221, ICOMP-90-16, E-5635, NAS 1.15:103221 NASA ORDER C-99066-G; RTOP 505-62-21...










Elliptic Flow Computation by Low Reynolds Number Two-Equation Turbulence Models


Book Description

A detailed comparison of ten low-Reynolds-number k-epsilon models is carried out. The flow solver, based on an implicit approximate factorization method, is designed for incompressible, steady two-dimensional flows. The conservation of mass is enforced by the artificial compressibility approach and the computational domain is discretized using centered finite differences. The turbulence model predictions of the flow past a hill are compared with experiments at Re = 10 exp 6. The effects of the grid spacing together with the numerical efficiency of the various formulations are investigated. The results show that the models provide a satisfactory prediction of the flow field in the presence of a favorable pressure gradient, while the accuracy rapidly deteriorates when a strong adverse pressure gradient is encountered. A newly proposed model form that does not explicitly depend on the wall distance seems promising for application to complex geometries. Michelassi, V. and Shih, T.-H. Unspecified Center NASA ORDER C-99066-G; RTOP 505-62-21...