[chanchihin]

Hi FOAMers,

I've been trying to validate the boundary layer profile of my body by plotting the eddy viscosity ratio near my wall. I am simulating a high Re case ~ 3m and can't afford to have well-resolved mesh so my current mesh has first layer y+ ~ 25. The simulation ran and converged and I noticed that my eddy viscosity (nut) from the first layer at the wall is 0 and grows to almost 5000 times my kinematic viscosity (nu) in the freestream.

Based on my simple understanding of turbulence, shouldn't eddy viscosity (nut) be non-zero at the inertial layer (my first layer) and at 30-50 times kinematic viscosity (nu), while decreasing to 0 outside the boundary layer?

I am basically trying to validate this:
https://www.computationalfluiddynami...oundary-layer/

I have attached my plots. Am i making mistakes for my simulation, I.C or B.Cs?

[Tobermory]

Check your inlet boundary condition for nut, since your high free stream value for nut is probably due to that. Otherwise, it's not really possible to comment further without you sharing the case files ... for example, what boundary condition are you applying at the wall? what turbulence model are you applying? what are your inlet boundary conditions?

[chanchihin]

Hi, thank you for your suggestion. You are absolutely right that my inlet conditions had too high K, and omega values - Nut was calculated field. I've adopted the same inlet and initial conditions (lower values) from this paper and got the nut/nu qualitative similar.

I've attached my results with 2 different meshes, y+~5 (natural B.C) and y+~30 (NutWallFunction, OmegaWallFunction, kqRWallFunction)

What is interesting is that y+~30 showed a stronger Nut in the inertial layer compared to y+~5. What is not obvious to me is that the first layer of the y+~30 mesh showed 0 Nut - which I thought should not non-zero since it is embedded in the inertial layer with maximum turbulent production reflecting non-zero Nut? Do you have any opinions on that?

[Tobermory]

Originally Posted by chanchihin

What is interesting is that y+~30 showed a stronger Nut in the inertial layer compared to y+~5. What is not obvious to me is that the first layer of the y+~30 mesh showed 0 Nut - which I thought should not non-zero since it is embedded in the inertial layer with maximum turbulent production reflecting non-zero Nut? Do you have any opinions on that?

I think you have this the wrong way around. Looking at your pictures, the mesh with y+~30 gives a non-zero nut in the cell next to the wall, as expected, while the mesh with y+~5 (ie in the laminar sublayer) gives a zero nut in the cell-adjacent wall, again as expected. nut should be zero in the laminar sublayer, and then increases through the BL to a peak at around the midpoint. Seems like that's what is happening in your simulation.

PS. I would suggest aiming for y+ ~ 1 or less if you really want to run lowRe.