[alimea]

Hi

How can I convert a "volScalarField" to an "fvScalarMatrix"?

I want to add a volScalarField term to energy Eqn.

Thanks

[Zeppo]

Adding a volScalarField to a fvScalarMatrix is as simple as this:

volScalarField source = ...;
fvScalarMatrix equation = ...;

equation == source;

This way volScalarField goes directly to the rhs (source) of fvScalarMatrix.

[alimea]

Originally Posted by Zeppo

Adding a volScalarField to a fvScalarMatrix is as simple as this:

volScalarField source = ...;
fvScalarMatrix equation = ...;

equation == source;

This way volScalarField goes directly to the rhs (source) of fvScalarMatrix.

Thank you for your answer.
Yes, I did it by chance and it didn't give me any error😁
I thought that these kind of variables are different.
Again thanks

[Zeppo]

Even though types of variables are different they can be operated that way just fine giving you the result that you might expect.

[rarnaunot]

Hi Foamers,

I would like to save the values of the fvScalarMatrix after solving it. But the operator "==" doesn't work upside down, I mean, assigning the values of the fvScalarMatrix to a volScalarField.

Do you know how to do it?

Thanks,

[agustinvo]

Hi,


you can't do this because a volScalarField is a scalar variable. fvScalarMatrix is an equation for a scalar variable. What do you want to save from it?

[rarnaunot]

Hi agustinvo,

I would like to save the solution of the fvScalarMatrix and then, pass the solution as a volScalarField to another fvScalarMatrix something like:

fvScalarMatrix equation0 ();

equation0.solve();
fvScalarMatrix equation1 (blablablah=solution of equation0)

Thanks,

[agustinvo]

Well, when you solve your fvScalarMatrix equation0 you compute the value of a certain variable var0. Then just add it as a source term into equation1.

[rarnaunot]

The problem is that the dependent variable of equation0 is not the same as the one in equation1...

Thanks for your quick reply,

[agustinvo]

That is not a problem! Look:

fvScalarMatrix TEqn
(
     fvm::div(phi, T)
   - fvm::laplacian(alphaEff, T)
   ==
     fvOptions(T) 

);

Here you solve for T, which is a volScalarField


You can add later on T in a source term of any other equation, like

tmp<fvVectorMatrix> tUEqn
 (
    fvm::div(phi, U)
 + turbulence->divDevReff(U)
 ==
    g*beta*(T-TRef)
 + fvOptions(U)
);

I hope it helps a bit. If you share your code we can help you better

[rarnaunot]

My code is something like:

        fvScalarMatrix STransfer
                (
                    K*S2Sat - fvm::Sp(K,S2)  
                );

        STransfer.solve();
 
        fvScalarMatrix S2Eqn
        (
            fvm::ddt(alpha2, S2)
          + fvm::div(alphaPhi2, S2)

           ==
            STransfer
          );
        S2Eqn.solve();


        fvScalarMatrix S1Eqn
        (
            fvm::ddt(alpha1, S1)
          + fvm::div(alphaPhi1, S1)
        ==
         -STransfer
        );

        S1Eqn.solve();

The problem is at S1Eqn as the dependent variable is S1 in the left side of == and S2 in the right-hand side ==.

That's one of the problems and probably I would face more after solving this one...

Thanks in advance,

PD: Do not hesitate to sent my a private message.

[Zeppo]

You may want to use S2 field for which S2Eqn equation has been solved for

        fvScalarMatrix S1Eqn
        (
            fvm::ddt(alpha1, S1)
          + fvm::div(alphaPhi1, S1)
        ==
         -(K*S2Sat - K*S2)
        );

        S1Eqn.solve();

[agustinvo]

Indeed, this is what it was planned to add.

[rarnaunot]

Thanks for your help Zeppo and agustinvo.

I have compiled that code and I'm trying it at a simple case.