Inlet, free-stream, conditions for RSM model
 

Hello guys,
I am solving a turbulent flow through duct problem.Can any one tell me the best inlet and boundary conditions for the Reynolds' stresses i.e. u'u', u'v', u'w' and so on.
Thanks a lot.

A simple and commonly used boundary condition at outlet is zerogradient. At the inlet, the boundary conditions have to be consistent with the flow physics. At wall and symmetry its straight forward.

Wall Boundary Conditions
 

Thanks (harishg) for your precious reply. I am using OpenFOAM, Launder-Gibson RSM model. OF automatically sets a wall and inlet value of 1e-15 for u'u', v'v' and w'w'. This may be understood. But the trouble with other stresses u'v', u'w',... etc. They are zeros at inlet but I can't find out how OF sets their values at the wall :confused:. Surely k is given a fixed value of 1.5e-15 at both inlet and walls. While epsilon and nu are calculated from wall functions.
Please help me solve this annoying problem.

All the Reynolds stresses are zero at the wall. However if yplus is in the log layer, the wall functions have to be calculated accurately even for the stresses. The values cannot be set to 1e-15 at the wall. Check the book of Wilcox.

you can use experimental data for a duct or pipe.for free stream this components are very amall.
as a suggestion if you have a 3d geometry first use a simpler model like k-e then switch to RSM model to have better convergence.