|
[Sponsors] |
January 23, 2010, 21:32 |
assemble the mesh
|
#1 |
New Member
liuhuan
Join Date: Dec 2009
Posts: 13
Rep Power: 17 |
Hi,everyone.
when i use the Icem_cfd, i can assemble the mesh. Or i can open the other mesh in my project, i can choose the merger option. So i can mesh the complex geometry by use of the merger option. But how to merger the part-mesh and what is the rule ? i think the rule is the global coordinate system . And how can i make sure of the sufficient precision in the interface when i assemble the mesh so that i can obtain a good result in the cfx-post ? |
|
March 15, 2010, 21:34 |
Merge or Concatonate?
|
#2 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Yes, the merge assumes that the location will be based on the global coordinate system.
However, you can move the mesh around if you need to (find the transform option). You can translate, scale, rotate, mirror, etc. Selecting the mesh is easier if everything is nicely broken up into parts. However, the "Merge" option when you load a mesh (as opposed to replace), is really just merging the files. It doesn't merge anything at a node for node level. If you want to do that, you need to go into "Edit Mesh => Merge => Merge Meshes. You would then select an interface part and the merging would happen. It is a very robust merge tool, but their are certain rules which I have gone thru several times on CFD-Online. You can find an older post. As for tolerance, that depends. If you want to actively merge the mesh node for node, then that probably has a tolerance around 1/5th the element size... but it is really based on the element part names and is pretty robust if the perimeter is well projected. I find the easiest way is to start from the same geometry file and break it up... Then everything is using the same coordinate system and the interface surfaces are in exactly the same location on both sides... I never have tolerance problems when I start from a single model. If you are not intending to actually merge the nodes, then the tolerance is really a solver concern... If the nodes already line up and you just want to merge them automatically, you can do a "Merge with a tolerance"... |
|
March 20, 2010, 03:39 |
|
#3 | |
New Member
liuhuan
Join Date: Dec 2009
Posts: 13
Rep Power: 17 |
Quote:
Thank you very much ! i will study it ! |
||
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Dynamic Mesh Problem. | Tom Clark | FLUENT | 10 | June 21, 2021 05:27 |
Meshing aifoil in ICEM | student123a | ANSYS Meshing & Geometry | 13 | December 8, 2010 11:40 |
problem when converting mesh (made by ICEM) using fluentMeshToFoam | Forrest_Lei | OpenFOAM | 11 | October 16, 2009 07:28 |
2d irregular grid | Remy | Main CFD Forum | 1 | December 22, 2008 05:49 |
General questions on grid-based computing | Adrin Gharakhani | Main CFD Forum | 21 | June 5, 2000 14:47 |