CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Mixer,ICEM and AMI (http://www.cfd-online.com/Forums/openfoam-solving/114031-mixer-icem-ami.html)

sharonyue March 4, 2013 04:27

Mixer,ICEM and AMI
 
4 Attachment(s)
Hi foamers;

I am using pimpleDyMFoam,the mesh was generated by ICEM.I find my result is a little wired.Anyone can help me?I dont know where is going wrong.

first I copy the fluent.msh into /incompressible/pimpleDyMFoam/mixerVessel2DAMI



1,Stationary Mesh.

The image is attached.

2,Rotate mesh.

The image is attached.

3,Merge mesh

I use ICEM to merge mesh,the image is attached.
AMI1 and AMI2 is actually the same.but there are two boundaries.

4,fluentMeshToFoam -writeZones fluent.msh

After this I get the polyMesh.and there are two parts in cellzones dictionary.


5,Change boundary dictionary.

Code:

5
(
    AMI1
    {
        type            cyclicAMI;
        nFaces          116;
        startFace      6944;
neighbourPatch  AMI2;
        transform      noOrdering;
    }
    IMPELLER
    {
        type            wall;
        nFaces          116;
        startFace      7060;
    }
    AMI2
    {
        type            cyclicAMI;
        nFaces          156;
        startFace      7176;
neighbourPatch  AMI1;
        transform      noOrdering;
    }
    WALLS
    {
        type            wall;
        nFaces          156;
        startFace      7332;
    }
    frontAndBackPlanes
    {
        type            empty;
        nFaces          7216;
        startFace      7488;
    }
)

6,dynamicMeshDict
Code:

dynamicFvMesh  solidBodyMotionFvMesh;

motionSolverLibs ( "libfvMotionSolvers.so" );

solidBodyMotionFvMeshCoeffs
{
    cellZone        ROTATE;

    solidBodyMotionFunction  rotatingMotion;
    rotatingMotionCoeffs
    {
        CofG        (0 0 0);
        radialVelocity (0 0 5000); // deg/s
    }
}

7,p
Code:

internalField  uniform 0;

boundaryField
{
    AMI1
    {
        type                cyclicAMI;
        value                $internalField;
    }

    AMI2
    {
      type                cyclicAMI;
        value                $internalField;
    }

    IMPELLER
    {
        type            zeroGradient;
    }

    WALLS
    {
      type            zeroGradient;
    }

 
}

8, U
Code:

dimensions      [0 1 -1 0 0 0 0];

internalField  uniform (0 0 0);

boundaryField
{


    AMI1
    {
        type                cyclicAMI;
        value                $internalField;
    }

    AMI2
    {
      type                cyclicAMI;
        value                $internalField;
    }

    IMPELLER
    {
        type            movingWallVelocity; // fixedValue;
        value          uniform (0 0 0);
    }

    WALLS
    {
        type            movingWallVelocity; // fixedValue;
        value          uniform (0 0 0);
    }
}

9,pimpleDyMFoam


the result is here.Looks like the velocity in the stationry mesh and rotate mesh is totally different, in the rotate mesh the velocity is big. in the stationary is small.

Looks like the velocity can not penetrate the AMI boundary.

Anyone know how to handle this?

sharonyue March 4, 2013 22:48

but in interDyMFoam, it looks fine. check it out here:http://www.cfd-online.com/Forums/ope...tml#post411534

be_inspired July 24, 2013 10:06

2 Attachment(s)
Hi,

Did you solve the issue with pimpleDyMFoam?

I am facing a similar problem. I have merged the two zones using Fluent. The AMI interfaces were set as "wall" inside Fluent and then, after conversion to OF with fluent3DMeshtoFoam, I change the type to cyclicAMI.

I have run simpleFoam previously to have a field initialized with no problems in the AMI patches. Then I changed to pimpleDyMFoam and it fails. The max Courant goes to 5.86138e+07. If I start from time 0, the solver runs but after some pimple iterations, the number of iterations required to converge Ux goes up to 1001 and it crashes.

Comparing my case with the propeller case(tutorial) I do not know if it is needed to do all those steps if previously I have already generated the AMI patchs with ICEM and I have already two regions.

Thank you very much

sharonyue July 24, 2013 10:18

Quote:

Originally Posted by be_inspired (Post 441709)
Hi,

Did you solve the issue with pimpleDyMFoam?

I am facing a similar problem. I have merged the two zones using Fluent. The AMI interfaces were set as "wall" inside Fluent and then, after conversion to OF with fluent3DMeshtoFoam, I change the type to cyclicAMI.

I have run simpleFoam previously to have a field initialized with no problems in the AMI patches. Then I changed to pimpleDyMFoam and it fails. The max Courant goes to 5.86138e+07. If I start from time 0, the solver runs but after some pimple iterations, the number of iterations required to converge Ux goes up to 1001 and it crashes.

Comparing my case with the propeller case(tutorial) I do not know if it is needed to do all those steps if previously I have already generated the AMI patchs with ICEM and I have already two regions.

Thank you very much

Hi,

If it works well in simpleFoam and crashes with pimpleDyMFoam, it's probably incurred by the AMI mesh, in your pimpleDyFoam, run this solver:

moveDynamicMesh

and see if the mesh is rotating.

Regards.


All times are GMT -4. The time now is 06:20.