turbulent flow....
When using 'the kех model', I think, I had to set the quantity of the turbulent kinetic energy, k, turbulent kinetic dissipation rate, e (or characteric length, l).
We use the code of "CFDACE". How to set the k, e, l ......???? Ah, what is the eddy viscosity......and that quantity ??? 
Re: turbulent flow....
When the flow becomes turbulent, no one knows how to solve it. Once upon a time, there was a smart person who lived in a far away land . One day, he invented a set of timeaveraged equations from the NavierStokes equations, and claimed that this set of equations can give you the answers of the turbulent flow. Inside this set of equations, there are unknown variables called Reynolds Stresses terms, which must be modelled and calculated before you can find the answers to the equations. In the last thirty years, researchers have been related these terms to eddy viscosity and velocity gradients. And the eddy viscosity also was further related to TKE ( turbulence kinetic energy) and TKEdissipation. To find the answers to TKE and TKEdissipation, you now have to solve a new set of equations , one for TKE and the other for TKEdissipation. To go from here, a good book on turbulence modelling would be a better place to start. ( TKEdissipation contains the length scale information of the turbulence model. A length scale is just a adjustable constant which can change from place to place in the turbulent flow. You can do whatever you like as long as you can get good results. A really good idea.)

All times are GMT 4. The time now is 07:22. 