|
[Sponsors] |
December 2, 2005, 06:46 |
2D Modelling using unstructured mesh
|
#1 |
Guest
Posts: n/a
|
Hi,
I used to run 2D simulation using structured mesh, when I turn to unstructured mesh for some curve shapes, I am bit confused. Using structured mesh, the third direction only has one cell and there is no gradient at that direction for any parameters. I am not sure how to embed only 'one cell' in the third direction using unstructured mesh. If the dimension on the third direction is as small as one side of a tetrahedron cell, I am having a problem like 'xxx is not a strict plane' when setting up boundary conditions. If I add more cells in the third direction there will be gradients for parameters. Anyone comes cross this problem? Thanks a lot Regards Yingchun |
|
December 2, 2005, 09:30 |
Re: 2D Modelling using unstructured mesh
|
#2 |
Guest
Posts: n/a
|
Best practice is simply to mesh your surface using standard methods - surface mesh only - then extrude that surface mesh one element thick. So you'll end up with prisms/wedges (extruded triangles) for your third dimension.
|
|
December 2, 2005, 09:50 |
Re: 2D Modelling using unstructured mesh
|
#3 |
Guest
Posts: n/a
|
Thanks Myron,
I will try your approach and see what I can get from it. Cheers Regards, Yingchun |
|
December 6, 2005, 13:22 |
Re: 2D Modelling using unstructured mesh
|
#4 |
Guest
Posts: n/a
|
Hi Myron,
I could use ICEM CFD to generate a large surface with the small curve surface inside it, mesh it and then extrude it as you suggested. My expection was that after the extrusion the surrounding surfaces would be generated automatically then I can define names for the surfaces, and then later for boundary conditions. But it seems no new surfaces are generated, only a mesh there. Any comments? By the way, i don't seem to have 'surface mesh only' option and not sure if i can put inflation in. Thank you Yingchun |
|
December 8, 2005, 10:11 |
Re: 2D Modelling using unstructured mesh
|
#5 |
Guest
Posts: n/a
|
You don't need surfaces to define parts - then to boundary conditions. All you need to do is put the surface mesh elements into appropriate parts. (Right-click on Parts > Create Part) Create the parts and add the appropriate elements to the parts. Those parts will then show up for applying boundary conditions. You probably won't be able to use the Prism process after extrusion - but you can extrude with a function so you can start with a thin layer then expand as you extrude more layers.
|
|
December 8, 2005, 12:07 |
Re: 2D Modelling using unstructured mesh
|
#6 |
Guest
Posts: n/a
|
Hi Yingchun,
I had the same problem when setting up symmetry planes in CFX-Pre. The reason was that the geometry and mesh imported from ICEM was (invisibly for my eyes) not correct. Before I created surfaces by using the embedding curves. I had to create the model on an alternative way again by using the extrusion option from a curve to create the surface. Afterwards it worked fine. I believe the reason was the topology command in ICEM where planes were broken into parts at some locations where their vertices met because of the Software inaccuratness. Strange, I know! Hope this helps a bit further. Regards Tobi |
|
December 8, 2005, 13:39 |
Re: 2D Modelling using unstructured mesh
|
#7 |
Guest
Posts: n/a
|
Hi Myron,
Thank you for your reply. Sorry, I still couldn't get it somehow. I don't have the 'mesh only option' because a separate license is needed for 'Mesh shell' which I don't have at the moment. If your explanation is based on this, I am not able to try it. For example, I draw a surface in ICEM, then give it a part name â€" say 'surface1'. It's only a surface, I am not able to create Body. I can mesh it after giving it a surface mesh size, but it seems all the elements are included in 'surface1' and I am not able to select elements as you suggested. Not sure where went wrong with me. Am I right if I just generate a surface, and try to extrude mesh, and it turns out to be a 'body'? Thank you for your help very much (the help document doesn't really help me to get through this) Regards Yingchun |
|
December 9, 2005, 10:11 |
Re: 2D Modelling using unstructured mesh
|
#8 |
Guest
Posts: n/a
|
Yes - take that surface mesh you have - and extrude it. Mesh > Extrude Mesh. Indicate the volume part (for the new 3-D elements), the top part (the surface elements opposite your starting surface mesh) and the side part (the surface elements through the "thickness" of the extrusion). Once you have that - you can move the new surface elements into parts of your choice.
|
|
December 12, 2005, 07:05 |
Re: 2D Modelling using unstructured mesh
|
#9 |
Guest
Posts: n/a
|
Thank you very much, Myron. I will have a go with this soon and let you know the outcomes.
Regards Yinngchun |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Gambit problems | Althea | FLUENT | 22 | January 4, 2017 04:19 |
ICEM unstructured mesh | peterputer1 | ANSYS | 0 | October 7, 2009 05:40 |
[Other] 2D unstructured mesh | majidhojjat | OpenFOAM Meshing & Mesh Conversion | 1 | May 14, 2009 18:09 |
Problem with solving for unstructured mesh | evgenii | OpenFOAM Running, Solving & CFD | 5 | November 27, 2006 05:37 |
[Commercial meshers] Unstructured hex mesh | lr103476 | OpenFOAM Meshing & Mesh Conversion | 12 | November 24, 2006 06:56 |