CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > ANSYS > ANSYS Meshing & Geometry

[ICEM] Disable Matching of Nodes on Parallel Edges

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree3Likes
  • 1 Post By bluebase
  • 1 Post By AtoHM
  • 1 Post By bluebase

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   January 2, 2021, 23:42
Question Disable Matching of Nodes on Parallel Edges
  #1
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Hello All,

I am currently creating a structured mesh for my geometry in ICEM.

From trying and as stated in the ICEM Tutorial Manual, when the number of nodes is changed on one edge, all parallel edges will automatically have the same number of nodes.

Is there a way to disable this from occurring for a structured mesh?

The the ICEM Help Manual, p.272 (the attached image), depicts what I want, but this mentions it has to do with patch dependent options (?), and I believe merging two different meshes, so I do not think it applies for my case.

Thanks in advance.
Attached Images
File Type: png Image_1.PNG (34.6 KB, 17 views)
aero_head is offline   Reply With Quote

Old   January 3, 2021, 10:06
Default
  #2
Senior Member
 
Sebastian Engel
Join Date: Jun 2011
Location: Germany
Posts: 561
Rep Power: 18
bluebase will become famous soon enough
Hi Kira,

for a fully structured(mapped) mesh it is not possible to have different element counts on parallel edges.

However, you can integrate (partly-) unstructured blocks within a blocking topology.

In the following menu you can switch block types from mapped to free: Blocking tab > Edit Block > Convert Block Type.
Such a free block uses the unstructured-mesh settings to fill it with elements. You could use for example a hexa/quad-dominant mesh to fill the block.
In 3D you also have the option of a so-called swept block. It is a prismatic-elements block where only top and bottom faces are free/unstructured meshes.

If hybrid meshes (structured + unstructured) are not an option, you would have to deal with more complex strategies In order to replace the impossible block with a blocking structure which matches your given edge counts.
You can find the so-called nesting strategy.

As far as i know, ICEM does not have a feature to create it automatically on the blocking level.

However, if you use different refinement levels on adjacent blocks, you can choose to resolve the difference in edge count. This only works if the edge counts differs by a factor of three. In other cases you end up having so-called hanging nodes - practically an internal interface.

Many solvers nowadays can deal with hanging nodes, though.


Best,
Sebastian
aero_head likes this.
bluebase is offline   Reply With Quote

Old   January 3, 2021, 10:39
Default
  #3
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Hello Sebastian,

Thank you so much for your very detailed reply.

You have confirmed my suspicion that it is not possible the way I have been trying it. The geometry is 3D, so I would have to look into using these swept blocks, or use nesting as you have mentioned.
aero_head is offline   Reply With Quote

Old   January 15, 2021, 00:19
Default
  #4
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Quote:
Originally Posted by bluebase View Post
Hi Kira,

If hybrid meshes (structured + unstructured) are not an option, you would have to deal with more complex strategies In order to replace the impossible block with a blocking structure which matches your given edge counts.
You can find the so-called nesting strategy.

As far as i know, ICEM does not have a feature to create it automatically on the blocking level.

However, if you use different refinement levels on adjacent blocks, you can choose to resolve the difference in edge count. This only works if the edge counts differs by a factor of three. In other cases you end up having so-called hanging nodes - practically an internal interface.

Many solvers nowadays can deal with hanging nodes, though.


Best,
Sebastian
Hello Sebastian,

Do you have any tutorials/examples of the nesting strategy you were talking about, as well as refinement levels on adjacent blocks?

Thanks in advance
aero_head is offline   Reply With Quote

Old   January 15, 2021, 03:59
Default
  #5
Senior Member
 
M
Join Date: Dec 2017
Posts: 414
Rep Power: 8
AtoHM is on a distinguished road
Kira, have a look around page 177 of this slide show: https://manualzz.com/doc/7414055/ice...r-cfd-analysis
aero_head likes this.
AtoHM is offline   Reply With Quote

Old   January 15, 2021, 11:00
Default
  #6
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Quote:
Originally Posted by AtoHM View Post
Kira, have a look around page 177 of this slide show: https://manualzz.com/doc/7414055/ice...r-cfd-analysis
Thanks AtoHM,

It states that this is used for unstructured mesh only? So there would be no way to do this for my structured mesh?
aero_head is offline   Reply With Quote

Old   January 15, 2021, 14:39
Default
  #7
Senior Member
 
Sebastian Engel
Join Date: Jun 2011
Location: Germany
Posts: 561
Rep Power: 18
bluebase will become famous soon enough
Resolving a refinement is indeed an operation which can only be run on an unstructured mesh.
However, this might be misleading a little bit since every premesh is converted to unstructured before exporting.
Hence, you can consider this feature a post processing step of your premesh.
Although it's possible to use this feature on any other unstructured mesh as well.

You just cannot run it in the premeshing phase.
aero_head likes this.
bluebase is offline   Reply With Quote

Old   January 19, 2021, 23:33
Default
  #8
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Thank you again Sebastian and AtoHM,

I am hoping I could get some last points of clarification;

If I were to implement this separate refinement in my mesh, it would introduce some skew to my cells, right? Right now, I have a perfectly orthogonal mesh (100% quality and min. angle of 90 degrees), and I wish to keep this.

EDIT: I did the block refinement and it tells me there is no skew, even when I use the 'resolve refinements' option. Interesting!

Along that line of thinking, if I were to have one mesh that was orthogonal and another that wasn't, I would need more nodes for the orthogonal case to 'make up' for the orthogonality, right?

Hope what I am asking is clear.

Last edited by aero_head; January 20, 2021 at 23:32. Reason: Did refinement
aero_head is offline   Reply With Quote

Old   January 21, 2021, 00:10
Thumbs up
  #9
Senior Member
 
Join Date: Nov 2020
Location: Canada
Posts: 377
Rep Power: 5
aero_head is on a distinguished road
Sebastian and AtoHM,

Thank you so much for your replies. The block refinement was exactly what I was looking for!

I'll be sure to make the transition from the other blocks that I did not refine as smooth as I can.
aero_head is offline   Reply With Quote

Reply

Tags
icem, meshing, structured mesh

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
[snappyHexMesh] SnappyHexMesh for internal Flow vishwa OpenFOAM Meshing & Mesh Conversion 24 June 27, 2016 09:54
Something weird encountered when running OpenFOAM in parallel on multiple nodes xpqiu OpenFOAM Running, Solving & CFD 2 May 2, 2013 05:59
[snappyHexMesh] external flow with snappyHexMesh chelvistero OpenFOAM Meshing & Mesh Conversion 11 January 15, 2010 20:43
parallel computing with non-domain member nodes luke.christ FLUENT 0 June 27, 2009 07:12
CFX4.3 -build analysis form Chie Min CFX 5 July 13, 2001 00:19


All times are GMT -4. The time now is 19:58.