[fish man]

Line 242 of force.cpp is used to calculate the frictional drag using the laminar flow model in OpenFOAM.

line 242 of force.cpp
-rho()*laminarT.nu*dev(twoSymm(fvc::grad(U)))

The fvc::grad(U) included in this expression is
postProcess -func “grad(U)”
is the same as the one calculated by postProcess -func “grad(U)”?

Thank you in advance.

[Tobi]

Should be yes. The postProcess uses a pre-defined file. Hence we execute the grad function object which indeed does the following (gradTemplate.C):

template<class Type>
bool Foam::functionObjects::grad::calcGrad()
{
    typedef GeometricField<Type, fvPatchField, volMesh> VolFieldType;
    typedef GeometricField<Type, fvsPatchField, surfaceMesh> SurfaceFieldType;

    if (foundObject<VolFieldType>(fieldName_, false))
    {
        return store
        (
            resultName_,
            fvc::grad(lookupObject<VolFieldType>(fieldName_)),
            mesh_.changing() && mesh_.cache(resultName_)
        );
    }
    else if (foundObject<SurfaceFieldType>(fieldName_, false))
    {
        return store
        (
            resultName_,
            fvc::grad(lookupObject<SurfaceFieldType>(fieldName_)),
            mesh_.changing() && mesh_.cache(resultName_)
        );
    }

    return false;
}

As you can see it is fvc::grad(whatEverField).

[fish man]

Hello
Thank you very much.
I now have a better understanding of "grad".

[fish man]

grad(U) derived by postProcess -func “grad(U)” and fvc::grad(U) in the source code had an average error of about 2.9 times when comparing values

It would be great if you could tell me why this is....

[Tobi]

How you compare the data? How are you checking the grad from source code (what ever you mean here).

Make sure you have the same discretization schemes active

[fish man]

We modified force.c and force.h in

src/functionObjects/forces/forces

to visualize gradU during analysis.
Then, after the analysis was finished, gradU was derived with

postProcess -func “grad(U)”

.
The above two gradUs were compared.