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

[ICEM] Help with fixing imported IGES model

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

Like Tree1Likes

Reply
 
LinkBack Thread Tools Display Modes
Old   August 9, 2010, 04:49
Default Final CAD fix (hopefully)
  #21
siw
Senior Member
 
Join Date: Jul 2009
Posts: 444
Rep Power: 14
siw will become famous soon enough
Finally, I got a smoothed surface mesh from the Octree mesher of the required element quantity and distribution and without any messages about holes . I haven't gone on and generated the TGlib AF Delaunay and Prisms just yet as the Octree was left to run overnight and I ant to get this issue fixed too re-run the Octree.

On inspection of the surface mesh at the fuselage/wing trailing edge interface I noticed 2 small curves which ICEM has followed with mesh elements (only about half a dozen elements). These 2 curves are in the original IGES file but should not be there. I've tried deleting them and re-running Build Topo but ICEM is finding them again. So they must be due to the surface.

The surface image just shows the perspective that the zoomed-in wireframe image is viewed from and the yellow lines are due to me trying different things with the Build Topo and not in the final mesh. The wireframe image shows the wing trailing edge with the 2 curves (bold red) and the trailing edge surface (thin orange). So how can I get the 2 curves to follow the straight trailing edge surface? I think this needs to be done with one of the Geometry tools rather than a tolerance setting, but I've had no success with deleting/merging/creating surfaces.

Thanks.
Attached Images
File Type: jpg Snap3.jpg (53.8 KB, 29 views)
File Type: jpg Snap1.jpg (57.9 KB, 39 views)
siw is offline   Reply With Quote

Old   August 9, 2010, 14:02
Default No need for a geometry based solution.
  #22
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 35
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
This is a case of how you think about it... Geometry is a problem for you, but the Octree doesn't care about the geometry as much as you do.

So, next time, you can just delete the curves and not bother to re-run build topology. It just means you won't have nodes lining up along that surface edge... do you actually need the elements to line up there? Lining up elements forces them into certain shapes based on the geometric constraints. This limits quality, etc.

I would say Octree only needs curves at the boundaries of sharp features and the perimeter of boundary conditions...

So, just delete the curves (and points for that matter) that do not mark the edges (and corners) of sharp features and boundary conditions...

Now, as for you existing mesh, I guess you won't want to remesh it... Just go into Edit Mesh => Move Nodes, and change the association of those curve projected nodes to "surface projected". This will free them of the silly curves. You could delete line elements also. Then run the smoother again and let it do its thing.

Sorry, I couldn't provide a geometry based solution. Hopefully this is better. If this really is the boundary of an important boco, let me know and we can work thru that.

Best regards,

Simon
happydogy likes this.
PSYMN is offline   Reply With Quote

Old   August 12, 2010, 04:39
Default Why now a Delaunay fault?
  #23
siw
Senior Member
 
Join Date: Jul 2009
Posts: 444
Rep Power: 14
siw will become famous soon enough
Okay, I've now managed to get a checked, smoothed surface mesh from the Octree mesher with no holes or strange elements as posted above at about the required element quantity and distribution. The octree mesh took +8 hrs to run (not sure total time as I left the PC running overnight from a 7am start) and ended with a mesh of about 23 million elements (2 million surface elements).

I then ran the Delaunay mesher (with TGlib and Use AF) and it finished in a matter of minutes with a volume element count of about 5 million as shown in the image. Clearly it didn't work. Why? I don't know what or why stuff has gone outside of the domain when I started with a checked, smoothed surface mesh. Next is how can this be fixed? I was expecting this part to be straight forward.

Thanks.
Attached Images
File Type: jpg Snap1.jpg (71.7 KB, 34 views)

Last edited by siw; August 12, 2010 at 04:46. Reason: Typo
siw is offline   Reply With Quote

Old   August 24, 2010, 11:06
Default Hmmm...
  #24
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 35
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
I have no idea why it would do that... Did you run all the checks before hand (no single edges, overlapping faces, etc.)? Have you tried any other methods (Standard Delaunay or Advancing Front?)
PSYMN is offline   Reply With Quote

Old   August 24, 2010, 11:22
Default
  #25
Member
 
Ludvik Lanicek
Join Date: Apr 2009
Location: Czech Republic
Posts: 52
Rep Power: 9
Ludvik is on a distinguished road
Define a Material Point and used it for mesher.
Ludvik is offline   Reply With Quote

Reply

Thread Tools
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 On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
help for different between les model (subgrid-scale model) liuyuxuan FLUENT 1 October 2, 2009 15:25
Grid resolution for full-scale and down scaled model gravis Main CFD Forum 0 October 2, 2009 10:27
2 stage axial turbine model convergence issues sherifkadry CFX 2 September 7, 2009 20:51
model han Main CFD Forum 0 September 27, 2005 08:35
Why Turbulence models are not universal. Senthil Main CFD Forum 4 July 5, 2000 04:34


All times are GMT -4. The time now is 20:11.