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

[ICEM] Meshing problem in Formula one

Register Blogs Community New Posts Updated Threads Search

Like Tree3Likes

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   January 16, 2013, 15:38
Default
  #81
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
What is the purpose of these parameters? And are they important only for the volume meshing? and how to calculate them?

1. Height

2. Height ratio

3. Num layers

3. tetra size ratio

4. Tetra witdth

Also I assume following definition of these parameters:

1. Max size : ICEM always uses this size

2. Min size : This is the min size allowed if required by some other parameters. Mesh will not be refined below this limit.

I know that Int wall will define the baffles (zero thickness wall) in the mesh. But what is the difference between int wall and split wall?

Far is offline   Reply With Quote

Old   January 17, 2013, 03:41
Default
  #82
Senior Member
 
AB
Join Date: Sep 2009
Location: France
Posts: 323
Rep Power: 21
BrolY will become famous soon enough
1. Height
2. Height ratio
3. Num layers
Those are for the prism. You first select the intial height of the first layer of prism, then the growth ratio, then the number of layer.

3. tetra size ratio
4. Tetra witdth
Those are for tetra inflation. First the growth ratio and the number of tetra layer.

Good question about split wall and int wall ..
BrolY is offline   Reply With Quote

Old   January 17, 2013, 03:52
Default
  #83
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Quote:
Originally Posted by BrolY View Post
1. Height
2. Height ratio
3. Num layers
Those are for the prism. You first select the intial height of the first layer of prism, then the growth ratio, then the number of layer.
Will these parameters override the prism parameters set in the global mesh setting > Prism
Far is offline   Reply With Quote

Old   January 17, 2013, 03:56
Default
  #84
Senior Member
 
AB
Join Date: Sep 2009
Location: France
Posts: 323
Rep Power: 21
BrolY will become famous soon enough
They are the same, I think the prism parameters set in the global mesh setting > Prism are just a shortcut.
BrolY is offline   Reply With Quote

Old   January 17, 2013, 04:13
Default
  #85
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Quote:
Originally Posted by BrolY View Post
3. tetra size ratio
4. Tetra witdth
Those are for tetra inflation. First the growth ratio and the number of tetra layer.
Good info. But instead of tetra width, the better name is number of tetra layers as you have just mentioned.

Can I apply tetra inflation on the surface?
Far is offline   Reply With Quote

Old   January 17, 2013, 04:19
Default
  #86
Senior Member
 
AB
Join Date: Sep 2009
Location: France
Posts: 323
Rep Power: 21
BrolY will become famous soon enough
What do you mean ?
Specify those informations directly on the surface mesh menu (Mesh -> Surface Mesh Setup) ? The answer is yes.
But if you want to inflate the 2D mesh on the surface, that won't work. You have to use the edge parameter for that I think.
Tetra width inflates the mesh on the volume mesh around the surface, not on the surface mesh ... I hope my explanations are clear :$
BrolY is offline   Reply With Quote

Old   January 17, 2013, 09:26
Default Surface mesh and Free nodes
  #87
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Surface mesh completed in ICEM. Trying to make volume mesh in either ICEM or TGrid.

I am getting free nodes in Tgrid. Any one idea why free nodes are there?








Far is offline   Reply With Quote

Old   January 17, 2013, 10:33
Default
  #88
Senior Member
 
scipy's Avatar
 
Alex Pasic
Join Date: Aug 2011
Location: Croatia
Posts: 199
Rep Power: 15
scipy is on a distinguished road
Send a message via Skype™ to scipy
Split wall and Internal wall are used mainly for preventing too high a degree of proximity refinement if you have 2 walls that are very close together (for example an endplate of a wing that consists of 2 parallel surfaces and is closed by some radius sweep or similar) and to determine prism growth on both sides of a zero-thickness surface.

Don't remember which one is for which, I think internal wall is for proximity and split for both-sided prisms, but don't hold me to this.

BTW Far, that is a rather poor quality mesh on the wings themselves, very high aspect ratio and low angle elements. I wouldn't try to force a "mapped face" mesh there.
scipy is offline   Reply With Quote

Old   January 17, 2013, 11:56
Default
  #89
Super Moderator
 
diamondx's Avatar
 
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 28
diamondx will become famous soon enough
you have a good geometry, the autoblock can do a good job
__________________
Regards,
New to ICEM CFD, try this document --> https://goo.gl/KAOIwm
Ali
diamondx is offline   Reply With Quote

Old   January 17, 2013, 12:02
Default
  #90
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Quote:
Originally Posted by scipy View Post
Split wall and Internal wall are used mainly for preventing too high a degree of proximity refinement if you have 2 walls that are very close together (for example an endplate of a wing that consists of 2 parallel surfaces and is closed by some radius sweep or similar) and to determine prism growth on both sides of a zero-thickness surface.

Don't remember which one is for which, I think internal wall is for proximity and split for both-sided prisms, but don't hold me to this.

BTW Far, that is a rather poor quality mesh on the wings themselves, very high aspect ratio and low angle elements. I wouldn't try to force a "mapped face" mesh there.

int wall is for the baffles aka zero thickness plate which is normally overlooked by the ICEM. But I feel you may be correct for the split wall, but need to check it more closely. Cant we grow the prism on the both sides of the internal wall?

I did use the autoblock there, as other options were not working such as patch dependent, PI etc. Any idea to overcome low quality problem there?

Also any hint why I am getting free nodes in Tgrid and I believe mesh is nicely done on all surfaces and I am only exporting surface mesh.

Quote:
Originally Posted by diamondx View Post
you have a good geometry, the autoblock can do a good job
I have used autoblock on wings and three walls perpendicular to the end plate...
Far is offline   Reply With Quote

Old   January 17, 2013, 12:05
Default
  #91
Super Moderator
 
diamondx's Avatar
 
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 28
diamondx will become famous soon enough
Quote:
I have used autoblock on wings....
then there is an option in icem to generate prism without turning those shell into tri
__________________
Regards,
New to ICEM CFD, try this document --> https://goo.gl/KAOIwm
Ali
diamondx is offline   Reply With Quote

Old   January 17, 2013, 12:09
Default
  #92
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Quote:
Originally Posted by diamondx View Post
then there is an option in icem to generate prism without turning those shell into tri
I didn't get this point. These shells are already tri!!!


or you mean I should use the quad elements on the wing?


This thread is really becoming interesting due to high level participation from my friends on cfd-online. Thanks for your contribution.
Far is offline   Reply With Quote

Old   January 17, 2013, 12:10
Default
  #93
Super Moderator
 
diamondx's Avatar
 
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 28
diamondx will become famous soon enough
yes, by autoblock i meant going to blocking sections and generating a 2d surface block
__________________
Regards,
New to ICEM CFD, try this document --> https://goo.gl/KAOIwm
Ali
diamondx is offline   Reply With Quote

Old   January 17, 2013, 12:22
Default
  #94
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Project files are shared here.

https://dl.dropbox.com/u/68746918/Formula1_Far.zip
Far is offline   Reply With Quote

Old   January 17, 2013, 13:05
Default
  #95
Super Moderator
 
diamondx's Avatar
 
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 28
diamondx will become famous soon enough
did you create that small box for a later merge of tri with hexa...
__________________
Regards,
New to ICEM CFD, try this document --> https://goo.gl/KAOIwm
Ali
diamondx is offline   Reply With Quote

Old   January 17, 2013, 13:09
Default
  #96
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Probably I do that. But intention is to get the refined mesh in the vicinity of wheel, wings and end plate. So this small box is really helping to hold the fine mesh in the important region while the outer domain will be filled with bigger elements (it may or may not be Hexa)
Far is offline   Reply With Quote

Old   January 20, 2013, 09:19
Default Update
  #97
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
An update on the meshing.

I have tried following options

1. Tetra inside and tetra outside. I am not able to make the volume mesh through denauly. Always get the one bigger cell. Although I am following tutorial of mesh merge in circular pipe (three section with two portions with tetra and one with hexa).

2. Then I tried in Tgrid. Volume mesh is easy to do here. There exported and imported volume mesh in ICEM for inner domain. Make the geometry with option Mesh >>> facets and build topology. Every thing is fine. There made the mesh in outer domain. Again same problem

3. Made the octree mesh in the outer domain (deleted every thing in the inner domain). Deleted volume mesh and smoothed and exported and imported in Tgrid. Filled the volume with advacing front with ratio 1.25. Saved file

4. Open both files in Fluent and created non conformal interface. (this is working and orthogonal quality is 0.2)

Problems/Future goals

1. How to make the conformal mesh at the interface in ICEM? I know the mesh merge option, but I am not able to make the volume mesh in inner domain or outer domain when both domains are present.

2. How to make the Hexa tunnel and then merge the tetra mesh? Which step should come first?

3. Can I merge both volume meshes in ICEM which are made in T.grid?

PS. 1 What are the best practices for the similar problems in ICEM?
2. If some one need any file (or all), I am more than happy to this including the help on overall process in ICEM, Tgrid and Fluent.

Progress:

Complete mesh in Fluent:





Outer domain:




Last edited by Far; January 20, 2013 at 09:50.
Far is offline   Reply With Quote

Old   January 20, 2013, 09:46
Default Remaining Pics
  #98
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
Inner domain:






Tgrid:
Far is offline   Reply With Quote

Old   January 20, 2013, 10:05
Default
  #99
Super Moderator
 
diamondx's Avatar
 
Ghazlani M. Ali
Join Date: May 2011
Location: Tokyo, Japan
Posts: 1,385
Blog Entries: 23
Rep Power: 28
diamondx will become famous soon enough
My method to the conformal mesh:

1- generate the volume mesh for the inner domain (T-grid or ICEM), whatever is fine for you.
2- save the mesh file, as "inner-domain"
3- delete the volume mesh of the inner-domain, and the shell in the ground, leave only the shell element in the interface.
4- from those shell elements in the interface, you will generate a volume mesh for the outer domain (t-grid or icem), i would have done it in tgrid, then i will save the mesh file for the outer domain. Now we have two msh files, but the advantage is that the nodes in the interface are in the exact same position. which makes it easy to merge !!!

Yet i have to try all this, but i'm sure it will work, i did some test with some simple cubes recently at work
__________________
Regards,
New to ICEM CFD, try this document --> https://goo.gl/KAOIwm
Ali
diamondx is offline   Reply With Quote

Old   January 20, 2013, 10:13
Default
  #100
Far
Super Moderator
 
Sijal
Join Date: Mar 2009
Location: Islamabad
Posts: 4,553
Blog Entries: 6
Rep Power: 54
Far has a spectacular aura aboutFar has a spectacular aura about
Send a message via Skype™ to Far
i tried. But no success.

This method is working perfectly fine on the simple cylinders or cubes.
Far is offline   Reply With Quote

Reply


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
Problem with querying the meshing database in CFX-Pre ashtonJ CFX 18 April 18, 2023 23:49
[Other] Hex meshing problem DM12 patrick ANSYS Meshing & Geometry 7 January 9, 2015 07:21
[GAMBIT] 3D boundary layer and meshing problem in GAMBIT 2.4.6 prashanthreddyh ANSYS Meshing & Geometry 1 December 20, 2011 00:35
[ICEM] Hexa meshing blocking problem fergal ANSYS Meshing & Geometry 0 June 28, 2011 10:01
Strange ANSYS Meshing 13.0 Problem brunob ANSYS Meshing & Geometry 3 June 21, 2011 20:03


All times are GMT -4. The time now is 19:53.