[The_Kaiser]

Hey Foamers!

I wand to compare the behaviour of walls with the S-A turbulence model in a DES simulation. My first calculation went for 200s to initialize the flow, then I started averaging my fields.
My BC for wall modelling were:

nut:
wall {
     type    nutUSpaldingWallFunction;
     value   uniform 0;
}

nuSGS:
wall {
     type      nutUSpaldingWallFunction;
     value     uniform 0;
}

Now I want to start a calculation from the 200. second. Would it be a good idea, to change nuSgs to zeroGradient? or to fixed value 0 ? my y+ < 1 everywhre and I want to know, if the results will be the same as with wall modelling
And what should I do with "nut"? Actually, it's not being written out in the next timesteps, I had it only in "0". Does it mean, I need it only for initializing the flow?

[The_Kaiser]

Did nobody a DES with Spalart-Allmaras?


Sent from my iPhone using CFD Online Forum mobile app

[Bazinga]

I am just now starting with S-A DES, DDES and IDDES - till now I always used Spaldings law. Here is a thread where someone used wall functions and no wall functions for a LES, maybe it helps:
LES WallFunctions in O.F. 2.2.0 (nutUSpaldingWallFunction)

Which version of OpenFOAM are you running? Because I use 4.0 and used the motorbike tutorial as a foundation for my simulations. I did not have a nuSGS file but nut and nuTilda. nut was saved in each time step as well as nuTilda for me.

[The_Kaiser]

I was running it with OF 2.3. nut was only needed in 0, for the simulation. the stramnge thing was, when I run for example yPlusLES, I had to copy the nut file into the timestep, that interested me