中国物理B ›› 2006, Vol. 15 ›› Issue (1): 149-155.doi: 10.1088/1009-1963/15/1/024
胡非1, 程雪玲1, 姜金华1, 马晓光1, 李磊2
Li Lei (李磊)ab, Hu Fei (胡非)a, Cheng Xue-Ling (程雪玲)a, Jiang Jin-Hua (姜金华)a, Ma Xiao-Guang (马晓光)a
摘要: In this study, the Reynolds-averaged Navier--Stokes (RANS) method is employed to simulate the flow within and over an intersection model with three kinds of $k$--$\varepsilon$ turbulence closure schemes, namely, standard model, renormalization group (RNG) model and realizable $k$--$\varepsilon$ model. The comparison between the simulated and observed flow fields shows that the RANS simulation with all the three turbulence models cannot completely and accurately reproduce the observed flow field in all details. A detailed comparison between the predicted profiles of wind velocities and the measured data shows that the realizable $k$--$\varepsilon$ model is the best one among the three turbulence closure models in general. However, the extent to which the improvement is achieved by the realizable $k$--$\varepsilon$ model is still not enough to completely and accurately describe the turbulent flow in a relatively complex environment.
中图分类号: (Navier-Stokes equations)