Improvement Of Turbulent Models In Typical Shear Flows

Reynolds-Averaged Navier-Stokes simulation,one of computational fluid dynamics methods,has the advantages of fast computational speed,high precision and less resource consumption.It's an economical choice for solving turbulence problems.Many previous researchers proposed turbulent models,in which standard k-?model is used extensively,and applied them in Reynolds-Averaged Navier-Stokes simulation for engineering problems.However,compared with the results of experiments or direct numerical simulation(DNS),the prediction of the standard k-?model is still not accurate enough for many flow fields(jet,far wake,etc.),and further optimization is still needed in these flow fields.Combined with the scale-by-scale energy budget equation,the turbulent viscosity hypothesis and other related analysis,the present study optimizes the standard k-?model.By the analysis,we get the constants suitable for jet and plane wake.We derive the theoretical expression of C_?1in round jet.Moreover,we calculate the distributions of C_?in round jet and plane wake.The optimized constants for plane wake are extended to the temperature field from the velocity field,and the validity of the values in the temperature field is verified.In addition,we also discuss other models,such as Bradshaw one-equation model,Spalart-Allmaras model(SA model),RNG k-?model,and (?) model for temperature.Combined with the Yaglom equation,we optimize the (?) model,and obtain the theoretical expressions of the model parameters.The model constants are improved on the basis of the experimental data of plane wake.Besides,this paper also extends SA model,and RNG k-?model to jet flow and plane wake respectively.We obtain the new model constants suitable for corresponding flow on the basis of the experimental data and DNS results.The improved constants have high accuracy in SA model's simulation for jet,and the prediction for plane wake under new constants of RNG k-?model approaches the results of experiments and DNS more closely.