CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Pre-Processing (http://www.cfd-online.com/Forums/openfoam-pre-processing/)
-   -   Axisymmetric Whirl BCs; Wedge or Cyclic? (http://www.cfd-online.com/Forums/openfoam-pre-processing/107132-axisymmetric-whirl-bcs-wedge-cyclic.html)

c_dowd September 18, 2012 09:50

Axisymmetric Whirl BCs; Wedge or Cyclic?
 
Hi all
I'm attempting to model a buoyancy driven whirling flow (about the z axis) using an axisymmetric mesh. To do this I have generated a 5 degree wedge straddling the x axis using block mesh, and defined the wedge walls as "wedge" types. My inlet conditions are fixed velocities at the outer edge of this wedge (the outside of the cylinder), specifying an x and y component.

My issue is that when I view the results, my y velocity is not being generated at the boundary, rather a much smaller value going in opposite directions on either face (instead of the same direction). Is this because the wedge patch type is intended only for axisymmetric problems with no flow through the wedge faces?

To fix this I've also tried using cyclic patches, which gives me the correct y component velocity however makes the x velocity behave strangely, specifically having slightly different (yet significant) values on each wedge face. My question is, is there a better way of specifying the velocity inlet on the outer edge of a wedge shaped mesh to avoid this?

Also, I was under the impression that the axisymmetric mesh would be solved similar to a 2d problem, if this is true though why do I get slightly different velocity values on each wedge face?

I'm still new to this so I feel it's likely to be something obvious i've overlooked.

linnemann September 18, 2012 13:10

Quote:

My issue is that when I view the results, my y velocity is not being generated at the boundary, rather a much smaller value going in opposite directions on either face (instead of the same direction). Is this because the wedge patch type is intended only for axisymmetric problems with no flow through the wedge faces?
Yes that is true.

As you say you need to use cyclics and I would recommend that you use at leas 10 cells in the circumferential direction.

c_dowd September 18, 2012 19:32

Ahh thanks a lot, what kind of angle should my mesh describe? At the moment i've got it at 22.5 degrees and it seems to be running okay. Thanks again for the help.

linnemann September 19, 2012 08:48

Hi

22.5 is a fine number in my opinion, but this is of course something you can play with to see if you get consistent results.

hiasl December 5, 2012 11:38

Hey Cameron,

I have a similar problem as you had - flow through the wedgefaces should be possible... Which changes did you have to do for the switch from wedge to cyclic? (I'm a OF-beginner)
Thank you so much!

Matthias


All times are GMT -4. The time now is 20:12.