[andreape]

Hello!

I'm trying to run some simulations with some rough surfaces but unfortunately I don't see any effect, i.e. the solution doesn't change if I vary the Ks number. Here an example in how I set the BC for nut:

PHP Code:

/*--------------------------------*- C++ -*----------------------------------*\
| =========                 |                                                 |
| \\      /  F ield         | OpenFOAM: The Open Source CFD Toolbox           |
|  \\    /   O peration     | Version:  4.x                                   |
|   \\  /    A nd           | Web:      www.OpenFOAM.org                      |
|    \\/     M anipulation  |                                                 |
\*---------------------------------------------------------------------------*/
FoamFile
{
    version     2.0;
    format      ascii;
    class       volScalarField;
    object      nut;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 2 -1 0 0 0 0];

internalField   uniform 0;

boundaryField
{
    farfield
    {
        type            freestream;
        freestreamValue uniform 0;
        value           uniform 0;
    }

    frontAndBackPlanes
    {
        type            empty;
    }

    wall_airfoil
    {
        type            nutUSpaldingWallFunction;
        value           uniform 0;
    }
    wall_airfoil_rough
    {
        type           nutkRoughWallFunction;
        Ks              uniform 0.2;
        Cs              uniform 0.5;
        value          uniform 0.0;


    }
}

// ************************************************************************* // 


Do you know what coulb be the error? I have to change only the BC in nut right?

Thank you in advance!

[RobertHB]

A few ideas without knowing the exact solution.
- The wallfunction seems to be best suited for flows with a large inherent nut value. I tried differen Ks values in a lamiar flow with the average nut being < 1e-10 and there are no visible changes.
- How does Ks = 0.2 relate to your geometry? If your airfoil is tenth of meters long, Ks = 0.2 might not do much. You could try to use a much larger value.
- I tested the wallfunction with the channel395 case and Ks = 1. Graphically, in paraview, there were no visible changes, but plotting nut and U over the distance from the wall showed minimal changes.

[andreape]

No the airfoil is only 1 meter long. Now I tried for greater Ks but seems to be always a little variation. Could be a problem of meshing? Or is it beacause it's in 2D?

For surface roughness problem the only thing that I have to change is the BC at wall in 0/nut right?

Thank you!

[RobertHB]

Meshing, that might be it. The BC does not work if ks is larger than your first cell height.

[andreape]

In fact, adjusting the first cell height I get some results that makes sense. Thank you. Now I had some difficulties to get the right drag coefficient. Do you have any advice?? Finer mesh behind the airfoil?