|
[Sponsors] |
October 27, 2005, 09:37 |
Meshing with Hexagonal Elements.
|
#1 |
Guest
Posts: n/a
|
Hai Everyone,
I have been trying to mesh a 3-dimensional geometry with hexagonal mesh using 'GAMBIT'.This particular planar geometry has curves on the two sides which is distributed asymmetrically which acts as 'RIB'.I couldn't mesh it with 'HEX' (hexagonal mesh which i prefer to 'T GRID') both using 'Map' and 'copper' because of the uneven distribution of the curves on the surface. I could mesh it with T-Grid (Hex Core).But I am curious to know whether I can use Hexagonal mesh whole through out the Geometry . Is there any other option that i can try out to Mesh this kind of geometry with hexagonal cells???? Thanks a lot. |
|
October 27, 2005, 10:29 |
Re: Meshing with Hexagonal Elements.
|
#2 |
Guest
Posts: n/a
|
Hex meshing in Gambit is a tricky thing. You either need to have pretty simple geometry, or you need to spend the time to break your geometry down into a bunch of simple parts (this is called volume decomposition). You also need to understand how Gambit treats different vertex types, and how those vertex types can be modified. Almost any geometry can be meshed with a hex mesh in Gambit, given an infinite amount of time and patience (patience probably being more important than time when talking about Gambit). Basically, in volume decomposition, what you're trying to do is turn complicated geometry into a collection of boxes (and these boxes can be warped and twisted, within reason, around your geometry). Once you have these boxes, then you can manually go in and mesh each box with a hex mesh. It's not a simple process (like changing a default value in Gambit), and requires practice, patience, and a lot of time to be able to do with complicated geometry. You have to be able to visualize how a hex mesh would look in each box, and how the hex mesh in each box interacts with the hex mesh in the boxes around it. In my experience, I work with a lot of "production" hardware... and this geometry is pretty complex, and I could lose weeks or months trying to just create a hex mesh, when I can spend a day getting a tet mesh. You need to balance the benefits of using a hex mesh, to the effort needed to create this mesh. A sufficiently refined tet mesh can be just as accurate as a hex mesh anyway (yes, it'll take more memory and longer to converge, but if it would take 6 weeks to mesh with a hex mesh, but only an extra day to run with a tet mesh, then is it worth it?).
Jason |
|
October 27, 2005, 10:45 |
Re: Meshing with Hexagonal Elements.
|
#3 |
Guest
Posts: n/a
|
Thanks a lot for your reply Geo.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
[ANSYS Meshing] Hex meshing producing poor quality elements on multibody part | aarvay | ANSYS Meshing & Geometry | 8 | March 16, 2012 16:29 |
2D Meshing - No matter how thin my model is, it is still 2 elements thick | RossFS | ANSYS Meshing & Geometry | 6 | April 14, 2010 01:13 |
How to introduce meshing elements at desired locations | enr_venkat | ANSYS Meshing & Geometry | 5 | March 9, 2010 06:39 |
Penetrating elements in extruded mesh | Michael P | CFX | 2 | May 20, 2005 09:06 |
CFX4.3 -build analysis form | Chie Min | CFX | 5 | July 13, 2001 00:19 |