[Mentalo]

Hi Folks,

I'm solving an axisymmetric Pipe-Flow with wedge boundary conditions. On watching the residuals, I was wondering why why all 3 velocity components are solved (see Picture below). Due to the U_y velocity residual my simulation never converges.

The wedge angle is 2deg on both sides, so less than 5deg. My blockMesh looks like this:

convertToMeters 0.001;

vertices
(
    (0 0 -315)
    (5.25 -0.18333404 -315)
    (5.25 0.18333404 -315)
    (0 0 -315)
    (0 0 0)
    (5.25 -0.18333404 0)
    (5.25 0.18333404 0)
    (0 0 0)

    (0 0 1000)
    (5.25 -0.18333404 1000)
    (5.25 0.18333404 1000)
    (0 0 1000)

    (12.5 -0.43650962 0)
    (12.5 0.43650962 0)
    (12.5 -0.43650962 1000)
    (12.5 0.43650962 1000)

    (100 -3.4920769 0)
    (100 3.4920769  0)
    (100 -3.4920769 1000)
    (100 3.4920769  1000)

);

blocks
(
    hex (0 1 2 0 4 5 6 4) (3 1 158) simpleGrading (0.6 1 0.8)
    hex (4 5 6 4 8 9 10 8) (3 1 500) simpleGrading (0.6 1 1.5)
    hex (5 12 13 6 9 14 15 10) (3 1 500) simpleGrading (1.2 1 1.5)
    hex (12 16 17 13 14 18 19 15) (40 1 500) simpleGrading (1.5 1 1.5)

);

edges
(
);

boundary
(
    symmetry
    {
      type symmetryPlane;
      faces
    (

     (0 4 4 0)
     (4 8 8 4)
     
     );
    }

    inlet
    {
        type inlet;
        faces
        (           

     (2 0 0 1)
        
        );
    }

    pilot
    {
        type inlet;
        faces
        (           

     (13 6 5 12)
        
        );
    }


    coflow
    {
        type inlet;
        faces
        (           

     (17 13 12 16)
        
        );
    }

   surrounding
    {
      type patch;
      faces
    (

     (18 19 17 16)

     );

    }

    wall
    {
        type wall;
        faces
        (           

     (2 1 5 6)
        
        );
    }

    outlet
    {
        type outlet;
        faces
        (           

     (8 8 10 9)
     (10 15 14 9)
     (15 19 18 14)
        
        );
    }

    rotper1
    {
      type wedge;
      faces
    (

     (1 5 4 0)
     (5 9 8 4)
     (12 14 9 5)
     (16 18 14 12)

     );
    }
     
    rotper2
    {
      type wedge;
      faces
    (

     (6 2 0 4)
     (10 6 4 8)
     (15 13 6 10)
     (19 17 13 15)

     );
    }



 
);

mergePatchPairs
(
);

Regards
Philipp

[kishpishar]

Hi Philipp,

I have noticed the same thing too and I think this behavior is very common for Axisymmetric cases with simpleFoam type solvers. I think (although I am not 100% sure) that the third velocity component (Uz) is solved to account for an axisymmetric swirl component and to include the possibility of giving a swirl component to the inlet velocity. Of course, most simple cases like pipe flows, we specify the Uz component at the inlet zero and the swirl actually doesn't develop.

I think in your case, we don't need to look at the U_y residual development - we can safely conclude that the solution is converged after 600 iter.

regds
kumar

[glypo]

I agree with Kumar here. I've recently compared simpleFoam axisymmetric against simpleFoam with the full three-dimensional geometry. Though lacking convergence in Y, the other two components converged and matched up with the full model quite nicely. I believe there is nothing to worry about.