CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   Simple Frustrating Meshing Issue in Gambit (w/pics) (http://www.cfd-online.com/Forums/ansys-meshing/62901-simple-frustrating-meshing-issue-gambit-w-pics.html)

Dylan March 23, 2009 12:34

Simple Frustrating Meshing Issue in Gambit (w/pics)
 
2 Attachment(s)
Hello everyone, Thanks in advance for any help or insight you can provide to the issue I'm having..

Attached are 2 photos of a simple, 2D geometry I am attempting to mesh in Gambit. The first shows the geometry, meant to study submerged jet impingement on a surface. (The jet is fashioned as the long, thin segment of the face. It impinges on the surface, which is represented by the right-most vertical edge. The minuscule edge at the exit of the jet and resulting cutout of the face serves to separate the jet from the other fluid. This edge is 0.001 in length, compatible with all mesh intervals used.

As you can see, the two horizontal edges connecting to the impingement surfaces have been split so that I can construct a finer mesh (w/ regards to x dir.) in close proximity to the surface. When I then mesh the edges with the interval sizes I want, and then the face I keep running into the same issue; The face mesh becomes skewed at the wall instead of simply being finer with regards to the x-direction. Please see second image, which is zoomed in at the right-bottom corner of the geometry. The skewing from both ends of the surface edge seems to converge at the y-location synonymous with the jet exit..

Again, I meshed the edges first. Then the face with quad/submap. I have tried telling it to ignore size functions and many other things, but can't seem to figure out exactly what is going on. Thank you so much for any help you can give, this is really slowing me down! I'll be happy to provide any other necessary info.

Thanks,
Dylan

-mAx- March 23, 2009 14:59

A sketch of with BC may be helpfull for understanding your issue.
Regarding the gambit side, I would split your face in basic surfaces (square-rectangle) for an optimal control of your mesh

Freeman March 23, 2009 15:09

Hi Dylan,

I would try to do a vertical line between this two vertexes, split the domain's 'big face' and then mesh these resulting 2 faces separately.

Hope this helps a bit, regards

Freeman

Dylan March 23, 2009 15:15

1 Attachment(s)
Thanks for the suggestion. If I were to divide the existing face into 3 separate smaller rectangular faces, it may solve the issue.. But, what boundary condition would I put at the interface of the faces to ensure uninterrupted fluid flow between the separate faces? I always try to keep a singular face for a continuum of fluid.

Regards,
Dylan

P.S. I've attached a labeled figure that may clarify the wordy description.

Freeman March 23, 2009 15:21

Quote:

Originally Posted by Dylan (Post 210461)
Thanks for the suggestion. If I were to divide the existing face into 3 separate smaller rectangular faces, it may solve the issue.. But, what boundary condition would I put at the interface of the faces to ensure uninterrupted fluid flow between the separate faces? I always try to keep a singular face for a continuum of fluid.

Regards,
Dylan

P.S. I've attached a labeled figure that may clarify the wordy description.

Don't worry about the lines existing in the middle of your domain. If you don't specify nothing to them (when defining your boundary types), then Fluent will interpret them as interior elements. Even if you export your mesh into Fluent and it says this line is a wall, you can change it into an interior element manually... and then you can even merge this line into the rest of the interior elements by going to Grid->Mesh and in "Interior", select the interior elements and your line: Fluent will merge them all into one interior zone.

Regards,
Freeman

Dylan March 23, 2009 16:36

Thanks a ton, guys! That worked like a charm!

Regards,
Dylan

Freeman March 23, 2009 17:14

And we celebrate it :)

Cheers,

Freeman


All times are GMT -4. The time now is 09:47.