CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   ANSYS Meshing & Geometry (http://www.cfd-online.com/Forums/ansys-meshing/)
-   -   [ICEM] mesh of a spinnaker (http://www.cfd-online.com/Forums/ansys-meshing/100404-mesh-spinnaker.html)

michele ambrosio April 25, 2012 05:27

mesh of a spinnaker
 
1 Attachment(s)
Any idea about meshing this geometry?
I am not looking for an accurate mesh for the moment.. just need some guidelines to proceed.
The idea is of course to have the mesh more fine close to the sail and coarser away from it. I was thinking about an unstructured mesh, but if anybody knows about a structured one is well accepted.
thanks in advance

michele

diamondx April 25, 2012 17:05

I took a look at you geometry. problem i find is Thickness. it's completely ignoring the geometry inside. This is an interesting case, first time i deal with it. I'm quite sure thickness is the problem. Where did you draw your geometry ? if in Ansys modeler, can you attach an IGES file of your geomtry from design modeler

michele ambrosio May 7, 2012 09:49

1 Attachment(s)
Sorry for replying after so many days, but I was busy. Any suggestion is really appreciated

Ralen May 11, 2012 08:50

TRM_SRF has a thickness or it is a surface?

Far May 11, 2012 10:56

it is a surface?


ok it ishttp://www.google.com.pk/imgres?imgu...PUBMAg&dur=350

http://en.wikipedia.org/wiki/Spinnaker

michele ambrosio May 11, 2012 11:07

yes it is only a surface without thickness

Ralen May 11, 2012 13:58

I built the mesh, but how to make fluent or cfx will "see" an impenetrable surface without thickness.

michele ambrosio May 11, 2012 14:15

@ Ralen:
Just give the boundary conditions of the sail as a wall

michele ambrosio May 11, 2012 14:16

Which software did you use for creating the mesh?
thank you a lot in advance for your help

Ralen May 11, 2012 14:48

1 Attachment(s)
Quote:

Originally Posted by michele ambrosio (Post 360638)
@ Ralen:
Just give the boundary conditions of the sail as a wall

This for Fluent. And for CFX a message "Select a BC type." Wall is not acceptable.

Far May 11, 2012 15:26

1 Attachment(s)
Here are the files.
Need some modifications to improve the quality.

Far May 11, 2012 15:29

pics
 
pics :

http://img36.imageshack.us/img36/616/66819953.png

http://img832.imageshack.us/img832/111/58871591.png

Far May 11, 2012 15:36

More pics
 
http://img692.imageshack.us/img692/1593/84815376.png

http://img706.imageshack.us/img706/9286/31458680.png

Far May 11, 2012 18:36

Quote:

"Select a BC type." Wall is not acceptable.
Do not specify bc in ICEM. CFX will automatically get the boundary condition using the named selection.

Also use import mesh option in CFX-pre to read the mesh in .cfx or .cfx5

michele ambrosio May 11, 2012 19:09

Thank you a lot it has been very useful!

Far May 12, 2012 00:16

Ralan's topology has the better quality than mine.

Ralan: Could you please check both and indicate what's the difference?


Since your case has surface in 3d , therefore in ICEM one has to explicitly project the faces to the wall of surface geometry.

Far May 12, 2012 00:22

Named selection for BC for CFX
 
Quote:

Do not specify bc in ICEM. CFX will automatically get the boundary condition using the named selection.
For example, inlet/in will have the inlet boundary and outlet/out will have the outlet boundary. similarly wall has the wall condition.

If your naming scheme does not fulfil these requirments still, specifying the different boundaries will give you option to select them later in CFX pre.

Just right click on the component(your geometry) in CFX-Pre and insert new boundary. In this way you can specify wall, symmetry etc. BCs for the remaining boundaries.

Ralen May 12, 2012 01:49

Quote:

Originally Posted by Far (Post 360673)
Since your case has surface in 3d , therefore in ICEM one has to explicitly project the faces to the wall of surface geometry.

Yes, I associate the faces to the sail surface.
I built o-grid mesh only for the sail. You built for area around sail. I also used the internal o-grid.

Far May 12, 2012 02:11

Like this
 
1 Attachment(s)
Like this ? But still quality is low!!!!

http://img692.imageshack.us/img692/9412/90916891mh.png

Far May 12, 2012 02:18

Quote:

Any idea about meshing this geometry?
The mesh extent is not enough in the y- and Z-direction. It is a low velocity simulation and you need at-least 10-15 lengths upstream and 20-25 lengths downstream.


All times are GMT -4. The time now is 14:20.