k and epsilon on inlet boundaries
 

Hello,

Does anybody know any rules for setting the boundary conditions for these two parameters of turbulence? how can I know their values at the inlet from a practical point of view? would be thankful for any suggestions.

Daniel

Re: k and epsilon on inlet boundaries
 

Hi!

There is an example in Tutorial 1.2

Good luck,

Julie

Re: k and epsilon on inlet boundaries
 

Hi Daniel

You can almost never set the right value of k and epsilon at the inlet. You must therefor have your inlet far away from the area you are interested in so the model will build it own boundary layer.

Re: k and epsilon on inlet boundaries
 

Daniel,

as Lennart says, always try and set your boundaries as far away as you can.

Since it is difficult to measure k and epsilon experimentally (and because the CFD epsilon bears little realtion to the physical turbulent dissipation) most people set up their turbulent boundary conditions in terms of Turbulent intensity (Ti) and length scale (L).

Generally people set Ti at 5 to 10% (0.05 to 0.1). For internal flows L is usually calculated as about 10% of the inlets hydraulic diameter. For external flows choose L so that the ratio of turbulent to molecular viscosity at the inlet is in the range 10 to 100.

These are only guidelines, and strictly speaking you should always carry out a sensitivity test on the values you specify, but in truth, in the hard world of industrial CFD people rarely do.

Good luck

Stephen