[rvl565]

Hi,

I am a little confused about the omega wall boundary condition.

I am running a turbulent flow over flat plate ( simpleFoam,komegaSST, 10m/s, 1%Tu intensity) and comparing the value of cf with a theoretical correlation.

When i use the Menter's wall boundary condition the cf is underpredicted ( about 70% error i think)

but when i use the omegawall functions value at the wall ( i run the case with the wall function for a few iterations and see what value it uses( which is a constant along the wall) and then use that particular value as 'fixedValue'), i get very good results ( <4% error)

To understand why this was happening, I looked at the omega wall function code and saw the following line:
scalar omegaVis = 6.0*nuw[faceI]/(beta1_*sqr(y[faceI]));

instead of the 60 ( as in menter's paper) there is a 6.0 in the wall function code.

Why is this so? Which one is correct?

Thanks for your help!!

[nimasam]

im not familiar with this model, but maybe its a bug, you can report it in bug report

[immortality]

hi
Its 6 in some new versions.its an approximation assymtotic value not an exact one.
Which type of grid you use near walls?

[rvl565]

I am using a fine mesh near the wall... y+=1

[immortality]

you should use a very low value for k like 1e-18
for epsilon and omega test two models:both 1e-18 or both high epsilon 100000 and omega 10000 and see which is better for your case.