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/)
-   -   [ICEM] Problem in selecting faces for O-grid creation (http://www.cfd-online.com/Forums/ansys-meshing/87911-problem-selecting-faces-o-grid-creation.html)

saisanthoshm88 May 2, 2011 15:10

Problem in selecting faces for O-grid creation
 
While I was Hex meshing a geometry. I wasn't able select an inner face for O-grid creation.

Following link has a PPt that has images to throw a better illustration of the problem. Some one please look into it and suggest a solution.

http://www.2shared.com/document/Lh7d...x_problem.html

By the way my geometry mostly resembles the one discussed with in the thread:

http://www.cfd-online.com/Forums/ans...blem-dm12.html

P.S: Simon, please let me know if this is a bug with the tool and please provide your comments and suggestions

PSYMN May 4, 2011 15:19

Two things...

First, it is very strange that it would behave that way. If you sent me the files I would take a closer look. I don't think I have ever seen this behavior before...

Second, For OGrid, selecting a face between two blocks just selects the block on the other side of face so that the OGrid can pass through it. When you select a face on the boundary, it really selects a VORFN block that you can't see and lets the OGrid flow through that.

So instead of selecting that face to let the OGrid pass into the main cylinder, you should just select blocks inside the cylinder.

Best regards,

Simon

saisanthoshm88 May 10, 2011 05:19

1 Attachment(s)
Hi Simon,

Here are the files please take a look into them and comment upon the issue:

Geometry: http://www.2shared.com/file/GBn0phQZ/project1.html

Blocking :
http://www.2shared.com/file/QGE8Hlm4/project1.html

These files show the blocking with out any O- grid and the problem can be seen when it is attempted to select faces for O-grid creation.


Also can you please tell me if two different O - grids ( one through the main tube and the other through the inlet tube) are required (or) can I have a single O - grid at a stretch running through the main tube and inlet tube( Please refer to the image attached to see what I meant).

jeevankumarb May 12, 2011 04:47

1 Attachment(s)
Hello santosh,

For your geometry single O-grid will be good, passing through both the cylinders, saperate O-grid wotn be a good idea it might lead to poor quality. As simon mentioned selecting an intermediate face will only select the adgecent block and will propogate the O-grid. I have made a O-grid, i hope it might be usefull fro you.

regards
Jeevan

saisanthoshm88 May 12, 2011 09:26

1 Attachment(s)
No it is better to have separate O - grids for the cylinders. You can find deformed blocks adjacent to the curve at which the two cylinders meet when you proceed with a single O - grid.

I've made two separate O - grids to the geometry. Please check out the attachment. Download the geometry file and then load the blocking into it . You'll find the premesh quality much better now.

PSYMN May 18, 2011 13:40

Its a feature...
 
I finally got to this one (thanks to the persistence of Santhosh),

Now that I have played with the model, I remember this better. Previously we got complaints from users that there were too many internal faces visible when selecting for Ogrid and this sometimes made it difficult for them to see what they were picking. Since you shouldn't be picking internal faces anyway, we just turned off the display of those. Instead of picking an internal face, please pick the next block over. That is all that picking the face would have done anyway...

As for the one ogrid or two debate, either way works... I usually look at what I expect the flow to be doing... If this is a slow moving thick-viscous fluid with a lot of back pressure, then the new fluid will likely enter along the wall and I would go with one Ogrid. If it is a faster moving fluid that injects its self through the boundary layer and into the cyclone, then the two Ogrid method is better.

If you go purely by the mesh quality numbers, I think you can probably get better quality with the two Ogrid method.

saisanthoshm88 May 19, 2011 00:13

Thanks for the clarification Simon.


All times are GMT -4. The time now is 14:05.