CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM (https://www.cfd-online.com/Forums/openfoam/)
-   -   Wall boundary conditions for turbulence models (https://www.cfd-online.com/Forums/openfoam/117957-wall-boundary-conditions-turbulence-models.html)

greenleader May 18, 2013 02:56

Wall boundary conditions for turbulence models
 
CFD + OpenFOAM Newbie question:
I'm confused about how to specify boundary conditions on walls for turbulence models.
(I have looked through the manuals, the forums and examples, and sadly I'm still confused!)

Can anyone offer a concise explanation of the entries that should be specified in the /0/k /0/epsilon and /0/omega dictionaries, and the general approach to selecting values of k/epsilon/omega on the walls?

(I am reasonably happy with how to set up the inlet/outlets, but I am struggling with the walls!)

The specific case I am working on is a circular duct with air flowing over an obstacle in the duct. Initially I want to use wall functions.

I see things like this in the examples:
Code:

    DuctWalls
    {
        type            omegaWallFunction;
        value          $internalField;
    }

or
Code:

  DuctWalls
    {
        type            epsilonWallFunction;
        value          uniform 0.01;
    }

What's the difference between value $internalField and value uniform? What do these keywords tell OF to do? Sometimes I see the value set to uniform 0.0 other times to some value.

If I wanted to do the near-wall approach and not use wall functions, what would I do differently in setting these parameters?

Any comments are appreciated....

msuaeronautics May 18, 2013 23:55

The $internalField keyword is essentially telling OF that you would like that particular boundary to take on the value that is specified next to the "internal field" closer to the top of the file. The "internal field" value is just a starting parameter describing what's going on inside the domain just before the simulation begins, and gets overwritten at the first step. It is normally acceptable to use 0 for this.

If I'm not mistaken, you should use large values of omega at the walls (in the thousands). The uniform keyword allows you to specify a particular value (essentially saying that the value you specify is uniform across that entire boundary and that there is no gradient or any type of variation).

greenleader May 27, 2013 10:01

Sorry msuaeronautics, I missed your reply... thanks for your comments.


All times are GMT -4. The time now is 07:22.