CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > ANSYS > ANSYS Meshing & Geometry

[ICEM] problem to merge HEXA/TETRA grid

Register Blogs Members List Search Today's Posts Mark Forums Read

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   January 22, 2011, 06:27
Default problem to merge HEXA/TETRA grid
  #1
New Member
 
Matteo
Join Date: Oct 2010
Posts: 5
Rep Power: 16
bikino is on a distinguished road
Hi I am a new user of the package ANSYS for solving fluid dynamics problems, I have to mesh a burner, so I was advised to use a hybrid mesh HEXA / TETRA, within a complex domain go to directly create a geometry consists in ICEM stuck on a cylinder with a truncated cone, which will generate the HEXA mesh, and the rest of the domain will use a mesh TETRA, the problem is created when the two met through the mesh and mesh merge command, on the surface of union which remains a shell-type quad / tri once imported into CFX or FLUENT remains, but in theory there should not be because the fluid passes.
-the gate before exporting, create errors in the mesh
if imposed as a surface-interface in CFX I do not have good risultai
Does anyone have advice?? Or tell me what is wrong in the process of generating a hybrid mesh??
Thank you very much
bikino is offline   Reply With Quote

Old   January 24, 2011, 04:42
Default
  #2
Senior Member
 
AB
Join Date: Sep 2009
Location: France
Posts: 323
Rep Power: 22
BrolY will become famous soon enough
You should have created two interfaces into ICEM (the surface where you connected your 2 mesh should be meshed in Tetra and in Hexa, so 2 interfaces).
So, once you merged your two interfaces, you should delete both of them (if the fluid is supposed to pass thru there). It would delete the surface elements, but keep the volume elements.
BrolY is offline   Reply With Quote

Old   January 24, 2011, 16:48
Default
  #3
New Member
 
Matteo
Join Date: Oct 2010
Posts: 5
Rep Power: 16
bikino is on a distinguished road
I create a conical cylinder-shaped surface, I create the blocking splits into 5 parts, I associate the blocking parts of the curves, and splits to go to ogrid, the blocking is created in the same part that contains the body, while the surface interface is located in another part.
dell'ogrid assign to each element, a number of nodes, move the body in the domino hexa and do the command to convert unstructured mesh, carry out the body count the mesh, I'm going to do the merge of meshs, but every time I error vertex incorrect, what wrong?
bikino is offline   Reply With Quote

Old   January 25, 2011, 04:33
Default
  #4
Senior Member
 
AB
Join Date: Sep 2009
Location: France
Posts: 323
Rep Power: 22
BrolY will become famous soon enough
Here is the procedure :

1) Go to Merge Nodes/Merge Tolerance. Then select a tolerance lower than your small mesh size for example. Select the option "Ignore Projection". And after that, select the the two interfaces which should be in contact. Click Apply.
2) Then go to the "Delete Elements mesh" tab, select "Delete by parts" and select your interface parts.

Anyway, is there a big difference between the cell sizes of your both meshes ?
Maybe you could try to play with the tolerance.
Hope it helps !
BrolY is offline   Reply With Quote

Old   January 29, 2011, 21:00
Default My 2 cents...
  #5
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Bikino,

After the merge, all the triangles on the interface part (you can use two parts or just one, it doesn't matter but some people like them to be separate so they can check them separately) should be gone and only quads should remain. If you have any triangles left, then the merge didn't work properly...

I have lots of posts about how to do this correctly, but the main thing is that the interface part(s) should only be about the interface (never used elsewhere) and that the entire perimeter should be shared with curve projected nodes (hexa side needs to be associated to the perimeter curves).

Anyway, I suspect that you did do it right and just have a layer of quads between your two connected volumes. If you want it to be one domain, make sure to move all the volume elements into a single part (right click on the one volume part => Add to part => use the selection tool bar to select all the volume elements.) Then you can delete the shells in the interface part... (Edit Mesh (tab) => Delete => Use the selection toolbar to select by parts, then select the interface parts to delete all their shells.

Best regards,

Simon
PSYMN is offline   Reply With Quote

Old   January 30, 2011, 05:31
Default
  #6
New Member
 
Matteo
Join Date: Oct 2010
Posts: 5
Rep Power: 16
bikino is on a distinguished road
Grazie per la risposta Simon,
non so cosa ci sia di sbagliato, quando creo la mesh di volume sia la parte tetra che la parte hexa sono nella stessa part, che è la stessa che contiene il BODY e il BLOCKING, mentre la superficie di interfaccia si trova in un altra part e la sua mesh di tipo shell è costituita sia da elementi tri sia da elementi quad.
E fino a qui tutto ok. adesso cerco di unire la mesh con il comando merge mesh, ma ogni volta ho dei vertici che non vengono uniti, come posso fare a unire questi elementi manualmente? quando creo il blocking devo mantenere in 2 part diverse le superfici e le curve? nel mesh part parameters devo impostare un parametro dimensionale per la superficie di interfaccia??
Thank you very much
Matteo
bikino is offline   Reply With Quote

Old   January 30, 2011, 13:20
Default
  #7
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Spiacente, il mio italiano non è abbastanza buono per capire questo messaggio.

I tried translation software, but it didn't do a very good job.
PSYMN is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Display Grid problem Mohsin FLUENT 1 March 12, 2010 04:45
Problem with BC on non-overlapping grid interface Ori Levin FLUENT 1 August 29, 2006 19:32
Grid adaption problem William FLUENT 0 May 25, 2006 08:14
grid generation problem vineet FLUENT 3 April 3, 2002 23:33
nonstaggered grid problem Zhi-Xing Yu Main CFD Forum 7 August 26, 1998 14:22


All times are GMT -4. The time now is 11:36.