CFD Online Discussion Forums

CFD Online Discussion Forums (
-   ANSYS Meshing & Geometry (
-   -   [ICEM] Quarter O-grid from vertices and !!!!!! Why? (

Far March 8, 2012 07:36

Quarter O-grid from vertices and !!!!!! Why?
2 Attachment(s)
Dear All

I am trying to create the Quarter O-grid for very simple geometry given in ANSYS ICEM 13.0 training lectures downloaded from ANSYS website. I am also using version 13 for this exercise. Although ICEM Ver 14 is also installed on some pcs and same problem exist on V 14 as well.

Problem description:

1a. Refer to Fig 1. I am trying to reproduce this topology i.e. quarter O-grid for wedge shape type block.
1b. I have done all steps like creating block, split into different blocks i.e. 3 length wise and 3 width wise shown in Fig. 2
1c. Then used collapse block to reproduce the wedge shape block (similar to Fig 1) and then created the Quarter O-grid from vertices by selecting the appropriate vertices as suggested in Fig. 1. Final blocking is shown in Fig 4.
1d. Blocking and tin files before Quarter O-grid and after Quarter O-grid are also attached with this post.
2. I have also tried Quarter O grid through O-blocking panel in split block panel by selecting the two edges, but this does not work at all.

My questions are
1. For steps 1a-1d I am able to create the quarter O-gird but ICEM does not like to mesh it and goes into infinite loop for ever.
2. For step 2 I am simply not able to create the quarter O-grid at all

Any solution,as this problem seems to be very simple but I am not able to figure it out.

Far March 8, 2012 08:36

1 Attachment(s)
Found solution, but not satisfied with the procedure.

In brief there were some duplicate blocks in quarter O-grid region, deleted them. Shall post details soon. Any other solution to this problem i.e. I want the clean blocking from starting? Any suggestion/comment please!!!

energy382 March 8, 2012 08:42

5 Attachment(s)
there were problems with your blocks (quarter o-grid). you can see it in vorfn panel. two or three blocks were nested.

on the pics you can see two simple ways to create quarter o-grids

energy382 March 8, 2012 08:44

3 Attachment(s)
the way I usually did

step1: merge verticies
step 2: transform block -> y-grid
step 3: merge verticies
(step 4: adjust)

Far March 8, 2012 08:50

I have also adopted two procedures. One similar to yours and other was to delete the nested or duplicate blocks.

For first method I have deleted y-block and blocks besides in width wise direction. Then created the blocks trough extrude along curve command and merged vertices. Then created Quarter o-grid though 6 vertices.

For second method I have simply deleted duplicate blocks and it does not require any extra work.

But your method in last post seems to be interesting. Can you explain it bit more especially Fig 1?

Far March 8, 2012 09:01

But my question remained unanswered " Why I am not able to create the quarter o-grid in the first attempt as suggested in Ansys ICEM training (Fig 1, Post #1) material?.
Why I need to follow the long route to get the very simple topology?

energy382 March 8, 2012 09:15

4 Attachment(s)
don't know this ansys training topology.

but the easiest way to create this topology is shown on the attached pics (at least in my opinion)

blocking1 -> blocking 2 i deleted the middle block

cfd seeker March 9, 2012 02:00

Far and Energy...can we generate the structured hexa mesh on the wing with circular or rectangular domain using this topology method? I am looking for this for many days but output is still zero...can you please help me?

Far March 9, 2012 04:17

Sure. Please start new thread, we shall try to help you from our limited knowledge.

cfd seeker March 9, 2012 11:38

Thanks I am starting it just now but your knowledge seems to be quiet big than mine :)

PSYMN March 9, 2012 17:28

Far, that "training material" is really just to give a hint. It is not one of the tutorials or anything like that... It is just part of a collection of slides that we show because we don't have time in the training to get into more details...

What you are missing is the first step. :o

You had probably started with a top down blocking, collapsed a block to get the wedge and then "deleted" the wedge.

This is different from the bottom up assembly shown in the "traning material". Energy382 also apears to have done a bottom up assembly.

The problem with your top down approach is that when you "delete" a block in ICEM CFD Hexa, it really just puts it into the VORFN part. The connectivity is still there, etc. It is just in a dead material that you have turned off. You might have better luck if you actually deleted the block (delete block with the "permanently" option on).

Personally, I usually create these with a top down method. You split the block in each direction and then merge one corner away (two merge operations in each index.)

Far March 9, 2012 23:40


Personally, I usually create these with a top down method.
First I tried this method, but it seems to not working at all (Refer to point # 2 Post #1). Any hint please?

Far March 10, 2012 06:03


Personally, I usually create these with a top down method. You split the block in each direction and then merge one corner away (two merge operations in each index.)
Used this method and merged vertices and then deleted wedge block. Created quarter o-grid by quarter-ogrid from vertices. I need the procedure which does not use the single step of bottom-up.

Can I use ogrid panel to create the quarter o-grid

Far March 13, 2012 08:53

1 Attachment(s)
Here is youtube clip for creating Y-block for top down approach.

All times are GMT -4. The time now is 22:10.