[pechwang]

Hello everyone，

I'm new to OpenFOAM. I have some experience with simpleFoam. Now I want to move to MRFSimpelFoam. What I'm simulating is an annular ring. In order to reduce the computational cost, I only simulate part of it. The domain is shown in the attachment. There are two walls, one is rotating and the other is stationary. When I select the rotating zone, I put all the cells in the rotaing zone. But the results are not correct. I know there must be something wrong, but I don't know where.

This is how I generate the mesh
convertToMeters 0.001;

vertices
(
(91.89382313 -9.658430006 0) //point 0
(103.7286337 -10.90231872 0) //point 1
(103.7286337 10.90231872 0) //point 2
(91.89382313 9.658430006 0) //point 3
(91.89382313 -9.658430006 0.127) //point 4
(103.7286337 -10.90231872 0.127) //point 5
(103.7286337 10.90231872 0.127) //point 6
(91.89382313 9.658430006 0.127) //point 7
);

blocks
(
hex (0 1 2 3 4 5 6 7) (20 27 10) simpleGrading (1 1 1)
);

edges
(
arc 1 2 (104.3 0 0)
arc 3 0 (92.4 0 0)
arc 5 6 (104.3 0 0.127)
arc 7 4 (92.4 0 0.127)
);

boundary
(
movingWall
{
type wall;
faces ((4 5 6 7));
}
fixedWall
{
type wall;
faces ((0 1 2 3));
}
inlet
{
type patch;
faces ((0 4 7 3));
}
outlet
{
type patch;
faces ((2 6 5 1));
}
side_half0
{
type cyclic;
neighbourPatch side_half1;
faces ((0 1 5 4));
}
side_half1
{
type cyclic;
neighbourPatch side_half0;
faces ((3 7 6 2));
}
);

Then I use topoSet to collect the cells. And it is my first time to use topoSet.
convertToMeters 0.001;

actions
(

// Get all cells
{
name rotor;
type cellSet;
action new;
source cylinderToCell;
sourceInfo
{
p1 (0 0 0); // start point on cylinder axis
p2 (0 0 0.127); // end point on cylinder axis
radius 104.4;
}
}
{
name rotor;
type cellZoneSet;
action new;
source setToCellZone;
sourceInfo
{
set rotor;
}
}

);
Since the simulation can be run, that means at least the mesh can provide enough information. And last is the U file.

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

internalField uniform ( 0 0 0 );

boundaryField
{
inlet
{
type zeroGradient;
}
outlet
{
type zeroGradient;
}
movingWall
{
type fixedValue;
value uniform ( 0 0 0 );
}
fixedWall
{
type rotatingWallVelocity;
origin (0 0 0);
axis (0 0 1);
omega constant -104.72; //rad per sec (1000 rpm)
}
side_half0
{
type cyclic;
}
side_half1
{
type cyclic;
}
}

I know maybe this problem is simple and stupid. But I really need helps. Thank you so much.

[nimasam]

well
1- create a test case and upload it here
2- do you think, you need MRF method at all!, it seems only one patch is moving then you can assign it an appropriate BC, where is your MRFzone?

[pechwang]

Hi Nimasam，

Thank you for your reply. The test case is in the attachment. The reason why I have to use rotating reference frame is because later the geometry will become complex. There will be some grooves on the wall. As to MRF or SRF, later I will move to interFoam. And there is only MRFinterFoam, there is no SRFInterFoam. So I think I have to use MRF method.

[pechwang]

The test case is in the attachment.

I first use blockMesh and then use topoSet to get the rotating zone.