hi:

I want to know if there is a proper turbulence model to simulate the low Reynolds number flow fields.(The Reynolds number maybe less 20 thousand or more) I know there are so many two eqations turbulence models with Low-Reynolds correction today used widely and successfully in simulating subsonic/supersonic flow fields,Can I use one of tuebulence models to simulate such low Renolds number problem derictly? I means no any modification to the compressible NS code,just use it to deal with low Reynolds number problem derictly.Is it suitable??? low Mach number may also cause some problem ,but now I don't want to consider it (a precondining method may be helpful,i known) Any advice will be great appreciated!

I am not a great expert in low Re number turbulent simulations, but I have read that the k-omega transitional model is better suited for low Re flows than the low Re versions of the k-epsilon model.

Spalart-Almaras is another model that can be used for low Re simulations.

Thank you very much! Now I'm trying to test some 2-eqs. turbulence models, Chien ,Shih-Lumley,and CMOTT k-e model( "CMOTT turbulence module for NPARC",NASA Contractor Report 2043,ICOM-97-10;CMOTT-97-05), I gained these models from a standard k-e model, and i had never used any k-w and S-A model,so it was so difficult for me to code such new models now.... I have made the Chien & Shih-Lumley model run sucessfuly in my code(maybe there still are some bugs),but the CMOTT k-e model failed,because some terms i didn't know how to derrive becauz the limit of Exp data,I only campared the computatinal results of flow pass a flat with Spalding teory(I still not known if it was proper to calculate the friction coefficent using this formular) for high REnolds number such as 6*10**6,the Chien model gives resonable result but Shih-Lumley model give a much more bigger result, but when i set Reno=0.2*10**5,Shih-Lumley model give a better result ,I don't know why! Maybe there's still too much thing i don't know!

[hamad]

i am try to simulate jet imingment on flate plate.
i suggest to use k-epsilom model. use option RNG. then differential viscosity model. For wall treatment: you can select stardard wall function or scalable wall function.

good luck