[rcastilla]

Hi, everybody,


I am facing a weird behavior of blockMesh that is puzzling me .


I have vortices, a block and the axis patch defined in this way:

convertToMeters 0.001;

vertices
(
    (              49   1.37369130467505 -0.059976657627337 )  
    (              57   1.37369130467505 -0.059976657627337 )  
    (              49                 0                  0 )  
    (              57                  0                  0 )  
    (              49   1.37369130467505  0.059976657627337 )  
    (              57   1.37369130467505  0.059976657627337 ) 


edges
(
);

blocks
(
    hex (2 3 1 0 2 3 5 4) b2 (25 25 1) simpleGrading (1 1 1) 
);

boundary
(
    axis
    {
        type empty;
        faces
        (
            (2 3 3 2)  
        );
     }
);

everything is ok, and the axis is created correctly;

Patches
----------------
  patch 0 (start: 1200 size: 0) name: axis
  patch 1 (start: 1200 size: 1325) name: defaultFaces

BUT if I change only the x-coordinate of the first axis vertex:

vertices
(
    (              49   1.37369130467505 -0.059976657627337 ) 
    (              57   1.37369130467505 -0.059976657627337 )  
    (              49.91                 0                  0 ) 
    (              57                  0                  0 )  
    (              49   1.37369130467505  0.059976657627337 ) 
    (              57   1.37369130467505  0.059976657627337 )  
);

blockMesh creates 2 faces in the axis patch.

Patches
----------------
  patch 0 (start: 1200 size: 2) name: axis
  patch 1 (start: 1202 size: 1325) name: defaultFaces

These faces are, approx, in the middle of the axis and have zero area. And, of course, it avoids to run any simulation.


Also, it depends of discretization as well. If the 25 cells are replaced by 26, the mesh is correct again.


Does have anybody any clue about the reason of this behavior? How to avoid it? Is it a bug of blockMesh?


Looking forward for your comments.


Best regards


Robert

[rcastilla]

Hi,

I have seen now that there is a similar issue, 3 years old, in the OF issue tracking system


https://bugs.openfoam.org/view.php?id=1443#c3970


And it is marked by solved, but I cannot find the solution...


Best regards


Robert