CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (https://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] Cylinder Meshing (https://www.cfd-online.com/Forums/ansys-meshing/98652-cylinder-meshing.html)

mll3261 March 15, 2012 11:10

Cylinder Meshing
 
I was wondering what the best process for using ICEM would be to mesh a cylinder. I wanted the circular surface to have a mesh that radially grows from the center. I have tried using ANSYS meshing but I could not get the circular surface to complete the radially mesh. I had gotten around this by using a small square in the middle of this surface and divided the rest of the surface into four pieces. I am now trying to find a way to get a purely radially mesh. I am new to ICEM and was just wondering what the process would be to complete this mesh, if possible, in ICEM.

Far March 15, 2012 11:36

this may be helpful http://www.cfd-online.com/Forums/ans...cylinders.html

CRT March 15, 2012 11:38

I think that the easy way is to make a o-grid and use the bunching laws: geometric 1 or 2. donīt forget to associate the edges to the curves at the top and the bottom of your cylinder.
it could be good that u check the tutorials.

mll3261 March 15, 2012 12:28

2 Attachment(s)
I tried following the tutorial for the 3D pipe, but I end up with just a rectangular mesh inside of my cylinder. I associated the edges of the squares at the end of the block to the circular curve. I have attached the image of the block and geometry curves, as well as the incorrect mesh.

Far March 15, 2012 12:46

There may be three possible problems
1. Edge association is not correct.
2. If step 1 is done, use edit edge >>> spline
3. Create four points on each circle and associate vertex to point.

To improve quality, create O-block as well.

Far March 15, 2012 16:54

1 Attachment(s)
I suspect you have made a very basic mistake about association. Check attached tin and blk files

http://img805.imageshack.us/img805/5365/cylindert.png

Utzname April 20, 2014 13:42

Hi, I am struggling with the same issue. I want to create a mesh, radially growing from the center. I am using KIVA and hence not allowed to use o-grid mesh.
Here a picture:

http://docs.autodesk.com/ACD/2010/EN...h_cylinder.png

I am fairly new to ICEM and really dont understand why it is so hard.
When I am doing a partially circle ( i.e. 90 degrees ) the mesh is automatically starting from the edge.
Thanks a lot for your help!

Far April 20, 2014 13:49

1 Attachment(s)
Quote:

Originally Posted by Utzname (Post 487166)
Hi, I am struggling with the same issue. I want to create a mesh, radially growing from the center. I am using KIVA and hence not allowed to use o-grid mesh.
Here a picture:

http://docs.autodesk.com/ACD/2010/EN...h_cylinder.png

I am fairly new to ICEM and really dont understand why it is so hard.
When I am doing a partially circle ( i.e. 90 degrees ) the mesh is automatically starting from the edge.
Thanks a lot for your help!

Create ogrid first. then merge vertices of inner ogrid.


http://www.cfd-online.com/Forums/att...1&d=1398016643

sirX April 4, 2017 19:09

Far could you please extend a little bit more the explanation about how to obtain your fully radial resultant mesh with ICEM? I am struggled with the same problem and find no solution... Hope you can help me! Thank you in advance.

kad April 5, 2017 06:31

5 Attachment(s)
Start with normal ogrid for a cylinder. See the first attached picture attached pictures. Then merge the inner vertices of the ogrid with the merge vertices tool (picture2). Do this for both sides of your cylinder and you get a structure like shown in picture 3.

You can also use an alternative approach by extrusion. Here I started with a 2D surface planar blocking (picture 4). Convert it to unstructered mesh the usual way and then do a full 360 degree extrusion by rotation with the 2D grid. Both methods end up with a grid shown in picture 5.

Far April 5, 2017 13:52

kad explained very well.


All times are GMT -4. The time now is 15:10.