shift WindNinja momentum solver runs from using the kEpsilon model to using the RNG kEpsilon model #525
Assignees
Labels
Comments
latwood
added a commit
that referenced
this issue
Jul 5, 2024
…ilon to RNG kEpsilon, but so far just for OpenFOAM 8. For issue #525
latwood
added a commit
that referenced
this issue
Jul 5, 2024
… 2.2.0, seems to be working, the coefficients are moving along in each time directory and simpleFoam reports it is using RNG kEpsilon now, but sample doesn't specify what turb model is used anymore so needs some more checking. Also need to study the code just a hint to make sure it is really using the values, it seems to be using them though. Then, probably want to modify the inletBC code so that instead of using a hard coded value for Cmu, 0.09 for kEpsilon or 0.085 for RNG kEpsilon, that it fetches it from an appropriate place instead, so it autoupdates itself without recompiling the inletBCs. For now this seems to be working though for issue ##525
latwood
added a commit
that referenced
this issue
Sep 10, 2024
…looks great! And adding small updates to the commented methods for getting/updating the Cmu and kappa coefficients in the inlet tke BC functions, just to keep track of the methods. Did notice that one difference, foam 8 vs 2.2.0, is that in foam 8, the epsilon wall function grabs its coefficients from the nut wall function rather than supplying them itself like in foam 2.2.0, this leads to those coefficients being dropped in the printing for the epsilon wall function during decomposePar and later timesteps, but looking through the code, since the nut wall function has the proper coefficients, the epsilon wall function is truly using the proper RNG coefficients, it's just not printing them for foam 8. Foam 2.2.0 prints them every timestep, I kind of prefer that method tbh. for issue #525, also related to issue #458
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
As we found in our soon to be published Perdigao paper, the RNG kEpsilon model performed better for turbulence outputs than the kEpsilon model, so now we are making RNG kEpsilon the primary run turbulence model for WindNinja momentum solver simulations in OpenFOAM.
It's straightforward to make the change in OpenFOAM 8, but may need a bit more tweaks for OpenFOAM 2.2.0, specifically with the inlet boundary condition application files.
The text was updated successfully, but these errors were encountered: