View Issue Details

IDProjectCategoryView StatusLast Update
0001245OpenFOAMBugpublic2015-02-24 17:02
Reporteruser897Assigned Tohenry  
PrioritynormalSeverityminorReproducibilityhave not tried
Status closedResolutionno change required 
PlatformLinuxOSOpenSuseOS Version11.3
Summary0001245: omegaWallFunctionFvPatchScalarField: omegaVis: constant may be incorrect.
DescriptionWhile working on a custom omegaWall function, I noticed that line 241:

scalar omegaVis = 6.0*nuw[faceI]/(beta1_*sqr(y[faceI]));

The constant, 6.0, is reported as 60.0 the sources I have read:

Knopp, T., Eisfeld, B. & Calvo, J.B., (2009): A new extension for k–w turbulence models to account for wall roughness. International Journal of Heat and Fluid Flow, 30(1), pp.54–65.

as well as on the following webpage (under the recommended boundary conditions):
TagsNo tags attached.



2015-02-08 17:39

updater   ~0003721

I've done a quick search for the paper mentioned in the wall function "src/turbulenceModels/incompressible/RAS/derivedFvPatchFields/wallFunctions/omegaWallFunctions/omegaWallFunction".
I found instead this thread:
There in post #3 you can find a link to the old technical paper from the same author (if I have not misunderstood it).

On that older technical paper "Improved Two-Equation k-omega Turbulence Models for Aerodynamic Flows", by Florian R. Menter, October 1992, there are exactly the 2 equations (23) and (24), where one is 10x of the other.
On that same subsection is written this:
«It should be noted that models based on the omega-equation give accurate results if the near wall values of omega are sufficiently large. Both, equations 23 and 24 satisfy this demand. The results are not sensitive to the factor (10) used in equation 24.»

I'm not sure what's written on the paper indicated on the omega wall function header, but possibly the factor 10 could somehow be optional? Although it would be another source of confusion, if the factor is not documented accordingly to when it should be used...


2015-02-08 22:18

manager   ~0003727

Given that different references provose different values for this coefficient it is not clear which we should use. If it is generally considered that 60 is preferable I am happy to put it in but I have not seen a paper which makes this statement. Alternatively it could be an optional input but then what should the default be?

Can anyone provide results showing that 60 is better than 6 either in accuracy or stability?


2015-02-24 17:02

manager   ~0003889

Orphaned report

Issue History

Date Modified Username Field Change
2014-03-28 02:12 user897 New Issue
2015-02-08 17:39 wyldckat Note Added: 0003721
2015-02-08 22:18 henry Note Added: 0003727
2015-02-24 17:02 henry Note Added: 0003889
2015-02-24 17:02 henry Status new => closed
2015-02-24 17:02 henry Assigned To => henry
2015-02-24 17:02 henry Resolution open => no change required