[anuragm]

Hello,

I am quite new to OpenFOAM and have run into some problems. I am trying to implement a solver for a multiphase system coupled with population balance (DQMOM). For this purpose, I am using a modified twoPhaseEulerFoam solver. The code was not written by me, I am just trying to edit it and implement if for my problem. The solver code was originally written for version 1.6 and I am using for OpenFOAM 2.3.0. However, whenever I run the solver on my test case, I get the following error:

--> FOAM FATAL ERROR:
dimensions of phi are not correct

    From function CrankNicolsonDdtScheme<Type>::fvcDdtPhiCorr
    in file finiteVolume/ddtSchemes/CrankNicolsonDdtScheme/CrankNicolsonDdtScheme.C at line 1162.

FOAM aborting

#0  Foam::error::printStack(Foam::Ostream&) in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#1  Foam::error::abort() in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#2  Foam::fv::CrankNicolsonDdtScheme<Foam::Vector<double> >::fvcDdtPhiCorr(Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh> const&, Foam::GeometricField<Foam::Vector<double>, Foam::fvPatchField, Foam::volMesh> const&, Foam::GeometricField<double, Foam::fvsPatchField, Foam::surfaceMesh> const&) in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libfiniteVolume.so"
#3
 in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/bin/multiPhasePbeFoam"
#4
 in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/bin/multiPhasePbeFoam"
#5  __libc_start_main in "/lib64/libc.so.6"
#6
 in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/bin/multiPhasePbeFoam"
Aborted

The same error also pops up if I change the scheme from CrankNicolson to Euler. I would be glad if someone helps or at least points me in the right direction. I checked the initial conditions and the pEqns.H file to check if there is a problem with the dimensions of phi. But no luck and I have no clue where to proceed at this stage.

The multiphase solver (multiPhasePbeFoam) can be found here and the test case (using 3 moments) is here.

Regards,
Anurag

PS: Because of the version differences, I replaced ddtPhiCorr with ddtCorr in pEqn.H as per this post. The solver would not compile without this change.

[anuragm]

To add to the above post, the solver is a modification of twoPhaseEulerFoam - it has been modified to solve for n (dispersed) + 1 (continuous) phases.

[wyldckat]

Greetings Anurag,

It requires a lot of coding changes, not simply a replacement of "ddtPhiCorr with ddtCorr"

Given the complexity of the code, I'll only be able to give you a few indications on how to proceed:

1. Download this file: https://github.com/OpenFOAM/OpenFOAM...lerFoam/pEqn.H - click on the "Raw" button to download it. This is the file "pEqn.H" for the solver twoPhaseEulerFoam in OpenFOAM 2.2.x.
2. Place the file in the same folder "multiPhasePbeFoam" with a new name, e.g. "pEqn22x.H".
3. Now, from the folder "multiPhasePbeFoam", run this command:
   

   diff -Nur pEqn22x.H $FOAM_SOLVERS/multiphase/twoPhaseEulerFoam/pEqn.H > 22xVS230.patch

4. If all goes well, you should now have a file named "22xVS230.patch" with a complete list of the changes that were made from OpenFOAM 2.2.x to 2.3.0. And you'll see why it isn't a quick/simple solution I could explain to you here. But you can now use that file as a reference guide as to which changes you need to do in the "pEqn.H" of your solver.

Good luck! Best regards,
Bruno

[anuragm]

Hi Bruno,

Thank you for the reply. Your suggestion was helpful. I am trying to see how different are the new pEqn.H files from the older ones (for both twoPhaseEulerFoam and multiPhaseEulerFoam). In my case, the twoPhaseEulerFoam solver was modified to solve for n+1 phases. So you are correct, the corrections will not be straightforward.

I will post again in case I encounter more difficulties.

Regards,
Anurag

[anuragm]

Hi Bruno,

I compiled the solver again making some changes in the pEqn.H file. This time I receive the following error message.

--> FOAM FATAL ERROR:
Different dimensions for +=
     dimensions : [1 0 -1 0 0 0 0] = [0 3 -1 0 0 0 0]

    From function dimensionSet::operator+=(const dimensionSet&) const
    in file dimensionSet/dimensionSet.C at line 185.

FOAM aborting

#0   Foam::error::printStack(Foam::Ostream&) in  "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#1  Foam::error::abort() in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#2   Foam::dimensionSet::operator+=(Foam::dimensionSet const&) const in  "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/lib/libOpenFOAM.so"
#3   Foam::DimensionedField<double,  Foam::surfaceMesh>::operator+=(Foam::DimensionedField<double,  Foam::surfaceMesh> const&) at  ~/OpenFOAM/OpenFOAM-2.3.0//src/OpenFOAM/lnInclude/DimensionedField.C:508
#4   Foam::GeometricField<double, Foam::fvsPatchField,  Foam::surfaceMesh>::operator+=(Foam::GeometricField<double,  Foam::fvsPatchField, Foam::surfaceMesh> const&) at  ~/OpenFOAM/OpenFOAM-2.3.0//src/OpenFOAM/lnInclude/GeometricField.C:1215
#5   Foam::GeometricField<double, Foam::fvsPatchField,  Foam::surfaceMesh>::operator+=(Foam::tmp<Foam::GeometricField<double,  Foam::fvsPatchField, Foam::surfaceMesh> > const&) at  ~/OpenFOAM/OpenFOAM-2.3.0//src/OpenFOAM/lnInclude/GeometricField.C:1215
#6
 at ~/OpenFOAM/misra-2.3.0/applications/solvers/multiphase/multiPhasePbeFoam/pEqn.H:25
#7  __libc_start_main in "/lib64/libc.so.6"
#8
 in "/home/misra/OpenFOAM/misra-2.3.0/platforms/linux64GccDPOpt/bin/multiPhasePbeFoam"
Aborted

Specifically, the offending line seems to be

phiDragb   += fvc::interpolate((alpha[i]/rhob)*K[i]*rUbA)*phia[i];

in the following pEqn.H file

http://pastebin.com/HEEe22Ff

Do you have any ideas where to go from this point?

Regards,
Anurag

[anuragm]

I managed to solve the above error by adding the following lines to my pEqn.H file

surfaceScalarField phiDragb
    (
        IOobject
        (
            "phiDragb",
            runTime.timeName(),
            mesh,
            IOobject::NO_READ,
            IOobject::AUTO_WRITE
        ),
        mesh,
        dimensionedScalar("phiDragb", dimArea*dimVelocity, 0)
    );

In essence, I defined the offending variable (phiDragb) to have the required dimensions.

Right now I am trying unsuccessfully calculate the deformation rate tensor. The problem seems to be similar - related to the dimensions. I am trying to solve this one and will post here again in case I hit a wall.

[vasava]

@anuragm: Can you post your solver files here please??

[anuragm]

Hi Vasava,

I am sorry but I saw your post only now. I figured I needed to understand the basics behind my problem before I could seek further help here on this forum. Although I could make the solver run, I had some problems with my results. So I am trying a different approach. I have been working on implementing DQMOM in the multiphaseEulerFoam solver in OpenFOAM 2.1.1/2.3.x on my own.

I dont know if you were looking for help or offering it.

PS: My solver files were uploaded and linked in the first post.

[vasava]

Thanks anuragm!. At that time I was exploring possibilities of using population balance within openfoam and was keen about your solver and problem. Can you re-post the solver if you have made any modifications. Also a test example would also be of great help.

Thanks for replying!!

[anuragm]

Hi Vasava,

I will upload the files this week. Just to let you know, I was having problems with the effect of gravity. Let me know if you are able to validate the results using these files.

[anuragm]

Hi Vasava,

After talking to my colleagues at work, I'm afraid I cannot send you the files because I do not own the code. It is best you contact the original authors of the paper.

Development and implementation of a polydispersed multiphase flow model in OpenFOAM

[vasava]

Thats ok. I will send them message my self.

[Tushar@cfd]

Originally Posted by anuragm

Hi Vasava,

After talking to my colleagues at work, I'm afraid I cannot send you the files because I do not own the code. It is best you contact the original authors of the paper.

Development and implementation of a polydispersed multiphase flow model in OpenFOAM

Dear Anurag,

I think the problem is with the dimensions of the code line. Since, you cannot share your data due to some constraints. I would advice you to check the dimension of both the RHS terms of the below mentioned equation.

        phiDraga[i] = fvc::interpolate(beta/phasea[i].rho()*K[i]*rUaA)*phib
                    + rUaAf[i]*(g & mesh.Sf());

I hope it will resolve your error.

-
Best Luck!