CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   Phoenics (https://www.cfd-online.com/Forums/phoenics/)
-   -   Turbulence Model (https://www.cfd-online.com/Forums/phoenics/52255-turbulence-model.html)

Ross April 14, 2008 11:31

Turbulence Model
 
Can anyone please give me a good definition of a tubulence model and what its job is?

Thanks

Ross

PattiMichelle Sheaffer May 6, 2008 15:27

Re: Turbulence Model
 
If you imagine a fluid flowing over an object, downstream there will be a vortex field, enhancing the mixing by its very nature. The problem is that in 3-dimensions, these vortexes break up at a very large rate so that there is a statistical ensemble of sizes of eddies, ranging from the size of the object down to molecular scales. These cannot be resolved on a computer grid because of the huge size range - you would need orders-of-magnitude more cells (and computer power) than are currently possible. Also, grids and solvers are "leaky" and at small scales (tiny eddies), it's hard to tell if a solution is accurate. So instead, you use a model that makes some mathematical assumptions about how the eddies affect mixing, diffusion, and dissipation, and then just solve these equations. The problem is that no set of equations is valid under all conditions. Read the Phoenics Encyclopedia and docs under "turbulence" - it has some good entries on the models and their performance.

PattiMichelle Sheaffer May 6, 2008 15:42

Re: Turbulence Model
 
...I forgot to mention that the turbulence model equations are solved on a normal (coarse) mesh. For instance, since turbulence enhances mixing, a turbulence model would increase the *effective* diffusion rate of species, but you would never "see" any eddies in the output of PHOENICS (or any other code) because it is only modeling the *bulk effects* of the eddies (e.g., increasing effective diffusion rate), not the eddies themselves.

Hope this helps,

Patti


All times are GMT -4. The time now is 11:56.