[Santiago]

So, I was wondering how to do the inner product between a dimensionedVector and a vectorField other than looping directly over the vectorField itself.

[me.ouda]

Did you try just using "&" operator.

I've used it and I get the expected values as if I dot the vector by each vector in the vector field. Here is the code:

    object      U;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

dimensions      [0 1 -1 0 0 0 0];

internalField   nonuniform List<vector> 
9
(
(-9.66371e-05 0 -0.133191)
(-0.000274768 0 -0.132876)
(-0.000411 0 -0.131537)
(-0.000544201 0 -0.129738)
(-0.000672783 0 -0.12728)
(-0.00077639 0 -0.124466)
(-0.000874089 0 -0.121203)
(-0.000942926 0 -0.117627)
(-0.00100253 0 -0.113825)
)
;

dimensionedVector v1 ("v1", dimVelocity, vector (0.2, 0.3, 0.4));

Info << (v1.value() & U ) << endl;

The result field:

dimensions      [0 1 -1 0 0 0 0];

internalField   nonuniform List<scalar> 9(-0.0532957 -0.0532054 -0.052697 -0.052004 -0.0510466 -0.0499417 -0.048656 -0.0472394 -0.0457305);

The old boundary conditions was "zeroGradient", for the new field it becomes "calculated":

outlet
    {
        type            calculated;
        value           nonuniform List<scalar> 3(-0.052697 -0.0499417 -0.0457305);

May be someone can help on how the resultant boundary condition is defined when we multiply two fields

[Santiago]

Hi @me.ouda


Well, I though so as well, that the & operator should be overloaded. But the thing is that my vectorField is not a volVectorField, but rather is over a patch. When I do the operation, the compiler throws a type-cast error. Ok, to be precise let me show what I'm trying to do:

    vectorField topPatchVelocity = 
        interpolatorBotToTop.faceInterpolate
        (
            UBot.boundaryField()[botPatchID]
        );

    topPatchVelocity = momentumDensityRatio & topPatchVelocity;
    topBC = topPatchVelocity;

This piece of code throws me an error in the last-but-one line. Note that momentumDensityRatio is a dimensionedVector.

[Zeppo]

Originally Posted by Santiago

topPatchVelocity = momentumDensityRatio & topPatchVelocity;

This won't compile unless momentumDensityRatio's dimension is dimLess. Dimensions to the left of = should be the same as dimensions on the r.h.s.

Extracting the non-dimensional value should work too:

topPatchVelocity = momentumDensityRatio.value() & topPatchVelocity;

[Santiago]

Originally Posted by Zeppo

This won't compile unless momentumDensityRatio's dimension is dimLess. Dimensions to the left of = should be the same as dimensions on the r.h.s.

Extracting the non-dimensional value should work too:

topPatchVelocity = momentumDensityRatio.value() & topPatchVelocity;

Nope, It didn't work. Now I get a type-cast error on the assignment operator:

setDirichlet.H: In function ‘int main(int, char**)’:
setDirichlet.H:10:19: error: no match for ‘operator=’ (operand types are ‘Foam::vectorField {aka Foam::Field<Foam::Vector<double> >}’ and ‘Foam::tmp<Foam::Field<double> >’)
  topPatchVelocity = momentumDensityRatio.value() & topPatchVelocity;

[Santiago]

Guys, I saw what the problem was: I was doing the contraction of two vector, and I needed a linear application on a vector (to scale it of course). I changed the scaling term to a dimensionedTensor and everything runs smoothly. Now I'll run the code and see what comes out.

Thanks