|
[Sponsors] |
May 12, 2010, 13:06 |
ICEM Ogrid
|
#1 |
New Member
Brian Polly
Join Date: Feb 2010
Posts: 26
Rep Power: 16 |
Hello all,
I am trying to figure out if it is better use Ogrid early in blocking process, then split or at end after splitting? Thanks Brian |
|
May 13, 2010, 07:59 |
|
#2 |
Member
james britton
Join Date: Jan 2010
Posts: 38
Rep Power: 16 |
Im just in the process of teaching myself icem but from all the tutorials ive been doing regarding blocking it has always been a matter or blocking first then implementing ogrid, so I would assume this is the correct way?
|
|
May 14, 2010, 09:32 |
2 types of Ogrid
|
#3 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
There are 2 reasons for Ogrid...
If you are just adding an Ogrid for boundary layer purposes, then you would add that much later so that the rest of the blocking is in place and you won't have to adjust the ogrid verts again when there are twice as many verts to move. However, I often use Ogrids for capturing topology... Such as in the Sphere Cube tutorial or the UPipe tutorial or with various wings, etc. You will find many of my threads on CFD online where creating an Ogrid happens very early in the process... often the first step. |
|
May 14, 2010, 13:33 |
|
#4 |
New Member
Brian Polly
Join Date: Feb 2010
Posts: 26
Rep Power: 16 |
Simon that is why I was questioning. The tutorials create it at the end of the blocking process but you seem to create it in the beginning. I have tried to read many of your posts and watched your videos and did many tutorials, but ICEM has been a nightmare for me working on my own projects.
Questions I would greatly appreciate help with: Creating periodic BC for modeling 120degrees of cylindric flow field? Fluent gives me a violation when I try to specify it in ICEM. If I try to specify in fluent via TUI I get errors about faces or something. What would be proper blocking method for this type of flow field where I am modeling a turbine blade located nearer the inlet. I have seem nice meshes done in Gambit where: structure mesh on blade, then structure mesh in near volume to capture boundary layer effects, then tetra in rest. I have done some prism mesh tutorials, but think I would rather have the stuctured turn into tetra if possible. Any help would be greatly appreciated. I have spent so much time on this with little result wise to show for it. Very frustrating. Thank you Brian |
|
May 14, 2010, 13:55 |
Pics...
|
#5 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Pictures would help... I saw your other post, but not sure why you can't make them small enough. I use MWSnap (freeware) and just save the image as a jpeg or gif and they are small enough...
In ICEM CFD, you can set up periodicity in global params (under the mesh tab). I show this in one of my movies on the youtube site... I even show a 120 degree wedge. After setting up the periodicity globally, you must also apply it to the blocking. You can use Blocking Tab => Edit Blocks => Make verts periodic. If you have single vertex at the axis, you can just click it twice. It needs to be done in the mesh before going to Fluent. Fluent just applies a boco, but it requires periodic mesh to work correctly. Lets get some pics up here and I can help you more specifically... |
|
May 14, 2010, 16:44 |
|
#6 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
Hey Brian,
I just helped someone else with a very similar model to yours... Maybe you can find the post... THis model starts with a quarter ogrid along the axis just to get the right starting shape. Then a few splits to rough out the hub and three splits around the airfoil just to create an area to put the CGrid in. Next create the CGrid around the roughed out wing, then collapse the trailing edge block. Your model is not as twisted as this, but you will still need one quarter ogrid to shift the periodic to compensate for your extra index on the trailing side. Then associate, setup periodicity, etc. When I did the earlier model, I found it easier to block a small region near the hub, then I stretched out the blocking during the association phase after I had the topology all in place. Don't worry about the particular curvature of your trailing edge, etc. The blocking is elastic and will fit to it. in your mind, you can block it schematically exactly the same as this other example. To create a boundary layer along the cone, you can just split that quarter Ogrid index that we started with. Here are some pics that may get you started... |
|
May 14, 2010, 16:46 |
Round at the root, sharp along the blade...
|
#7 |
Senior Member
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,663
Blog Entries: 1
Rep Power: 47 |
One other thought. Your model has a round cross section at the blade root... Don't worry about it. We will still put a collapsed CGrid around it. The blocking will just shrink-wrap around the transition and it will be fine.
|
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
ICEM BC to FLUENT | Simon | CFX | 3 | September 9, 2008 18:08 |
Design Modeler to ICEM..... | Joe | CFX | 0 | January 24, 2008 04:39 |
ICEM: 2D->3D (Ogrid around faces) ? | Alexey | CFX | 0 | May 29, 2007 11:04 |
ICEM 10 mesh scale foe CFX | Korsh Mik | CFX | 2 | October 7, 2005 10:07 |
ICEM CFD 5.1 Hex-Tet mesh merging failure | bogesz | CFX | 1 | January 29, 2005 07:46 |