[spwater]

Hi this is Pei, I am using k-epsilon model with default wall functions supported by openfoam, it seems when my mesh is coarse (10/unit lenghth) the result is quit good wcompared to the experiment. But after I increase the mesh desity to (30/unit length), the result changed and not fit the experiment.

I read some reference said the wall functions can only be used when the first point have a distance to the wall. Is there any treatments I can use in refined mesh with k-epsilon model?

Another qustion is about edge grading , can I generate a edge like this
. . . . ...... . . . .

with blockMesh, now I can only use simple grading...

Thank you very much.

[grtabor]

For standard wall models to work you need y+>11.7 for the majority of the first cell to be in the log law region. If you can't manage this you probably need a low Reynolds number model like the Launder/Sharma low-Re k-e model.

On your second question, could you split the edge into two separate blocks?;

|. . . . ..|.... . . . . |

Gavin

[spwater]

Originally Posted by grtabor

For standard wall models to work you need y+>11.7 for the majority of the first cell to be in the log law region. If you can't manage this you probably need a low Reynolds number model like the Launder/Sharma low-Re k-e model.

On your second question, could you split the edge into two separate blocks?;

|. . . . ..|.... . . . . |

Gavin

Thank you for your reply, I already modify blockMesh utility to add a double grading functions and wall treatment function.

[behzad-cfd]

Originally Posted by spwater

Thank you for your reply, I already modify blockMesh utility to add a double grading functions and wall treatment function.

Hi pei
i have same problem with grading and couldn't solve it.
if you could solve this please help me too.
i saw your modified blockMesh utility but it doesn't work on my OF versions.i tested on 1.6 ext,2.1.1,2.2.0 and 2.2.1.
thanks in advance