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

[ANSYS Meshing] Tetrahedral or Hexahedral

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

Reply
 
LinkBack Thread Tools Display Modes
Old   December 21, 2011, 10:02
Default Tetrahedral or Hexahedral
  #1
Member
 
Mina
Join Date: Apr 2011
Posts: 88
Rep Power: 6
Mina_Shahi is on a distinguished road
Dear Ansys users

I am going to mesh a 3D geometry, this geometry is rather simple one, and consisting 2 rectangulars part, 31 tubes with 1mm diameter are coneected to them acting as inlet boundaries, (see the attached file)

Geometry.png

Now i am wondering that with these small tube which mesh method is better Tetrahedral or Hexahedral ? i am going to mesh it in Workbench and then do CFD analysis,
Mina_Shahi is offline   Reply With Quote

Old   December 21, 2011, 15:22
Default
  #2
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,969
Blog Entries: 6
Rep Power: 39
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
mix of both
Far is offline   Reply With Quote

Old   December 22, 2011, 06:52
Default
  #3
Member
 
Mina
Join Date: Apr 2011
Posts: 88
Rep Power: 6
Mina_Shahi is on a distinguished road
Quote:
Originally Posted by Far View Post
mix of both
Thank you for the answer, so Hex DOMINANT should be Ok in ansys workbecnh, true??
Mina_Shahi is offline   Reply With Quote

Old   December 22, 2011, 12:34
Default
  #4
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 36
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
I agree with "Far", but don't recommend the Hexadominant method for CFD.

The Hexadominant method starts by paving quads at the walls and then marches inward with isotropic hexas which can crash somewhat badly in the middle. This is good for FEA structural analysis where most of the interesting stuff happens near the surface and uniform elements are sufficient to capture it all, but it is not good for CFD.

However, if you tried Multizone (or sweep) with an inflation layer, you could get a nice combination of swept hexas that would produce a good mesh for CFD.
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   December 22, 2011, 12:42
Default
  #5
Far
Super Moderator
 
Far's Avatar
 
Sijal Ahmed Memon (turboenginner@gmail.com)
Join Date: Mar 2009
Location: Islamabad Pakistan
Posts: 3,969
Blog Entries: 6
Rep Power: 39
Far will become famous soon enoughFar will become famous soon enough
Send a message via Skype™ to Far
What I wanted to say is "Always prefer the hexa mesh but when the quality cannot be maintained e.g. aspect ratio, skewness then it is better to switch to unstructured mesh".
Also in unimportant areas such as far-field use the unstructured mesh while using the structured mesh in boundary layer region, thereby inner mesh does not propagate in outer region.
Far is offline   Reply With Quote

Old   December 22, 2011, 17:36
Default
  #6
Member
 
Mina
Join Date: Apr 2011
Posts: 88
Rep Power: 6
Mina_Shahi is on a distinguished road
Quote:
Originally Posted by PSYMN View Post
I agree with "Far", but don't recommend the Hexadominant method for CFD.

The Hexadominant method starts by paving quads at the walls and then marches inward with isotropic hexas which can crash somewhat badly in the middle. This is good for FEA structural analysis where most of the interesting stuff happens near the surface and uniform elements are sufficient to capture it all, but it is not good for CFD.

However, if you tried Multizone (or sweep) with an inflation layer, you could get a nice combination of swept hexas that would produce a good mesh for CFD.
Thanks for your suggestion, i tried Multizone or sweep but it gives errors even when i changed the grid size ... also Hexadominant is not good it gives a warrning. a warning message states that a low percentage of hex elements or poorly shaped hex elements may result. the only method which worked was Tetrahedrons. what do you think?
Mina_Shahi is offline   Reply With Quote

Old   December 22, 2011, 17:51
Default
  #7
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 36
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Tetra is, by far, the easiest way to go and often a good way to start.

Your model looked simple enough to me that I thought MultiZone would have done nicely. Hexa is slightly better for quality and speed of convergence, but if the model is giving you trouble, go for Tetra/Prism...

One other thing you could try is "CutCel" hexa. It is as automatic as Tetra, but gives majority Hexa. It does use hanging nodes though so it really only works well with Fluent. Get 14.0 though, since it was improved a lot over 13.0 both in speed and robustness, especially when used with Prism.

Best regards,

Simon
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   December 22, 2011, 17:58
Default
  #8
Member
 
Mina
Join Date: Apr 2011
Posts: 88
Rep Power: 6
Mina_Shahi is on a distinguished road
Quote:
Originally Posted by PSYMN View Post
Tetra is, by far, the easiest way to go and often a good way to start.

Your model looked simple enough to me that I thought MultiZone would have done nicely. Hexa is slightly better for quality and speed of convergence, but if the model is giving you trouble, go for Tetra/Prism...

One other thing you could try is "CutCel" hexa. It is as automatic as Tetra, but gives majority Hexa. It does use hanging nodes though so it really only works well with Fluent. Get 14.0 though, since it was improved a lot over 13.0 both in speed and robustness, especially when used with Prism.

Best regards,

Simon
Did you use ANSYS workbench for meshing? there are 6 methods to make a mesh, CFX mesh, MultiZone, Tetra, Hexa dominant, ...

There is no ""CutCel" hexa" on that? By the way how Tetra works from this point :numerical error or dissipation error ?

Thank you again
Regards
Mina_Shahi is offline   Reply With Quote

Old   December 23, 2011, 16:45
Default
  #9
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 36
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Cutcel is not quite like all the others. It is a global method, so you don't set a method on a part or body, instead just left click on "mesh" in the model tree and then look down in the details panel. You will find a pull down that lets you turn on Cutcel.

I don't know what you are asking here
Quote:
By the way how Tetra works from this point :numerical error or dissipation error ?
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   December 20, 2012, 04:52
Question Warning in the Ansys Meshing
  #10
New Member
 
Join Date: Feb 2012
Posts: 6
Rep Power: 5
toda is on a distinguished road
Hello everybody,

I`m new in ANSYS, when I mesh some geometry in the Ansys Meshing, there is a warning: "Inflation created stairstep mesh at the some locations". What the meaning of this warning? If I use this mesh on Fluent simulation, is there any effect about this warning?

Thank you very much for the reply.
toda is offline   Reply With Quote

Old   December 20, 2012, 13:48
Default
  #11
Senior Member
 
PSYMN's Avatar
 
Simon Pereira
Join Date: Mar 2009
Location: Ann Arbor, MI
Posts: 2,662
Blog Entries: 1
Rep Power: 36
PSYMN has a spectacular aura aboutPSYMN has a spectacular aura about
Quote:
Originally Posted by toda View Post
Hello everybody,

I`m new in ANSYS, when I mesh some geometry in the Ansys Meshing, there is a warning: "Inflation created stairstep mesh at the some locations". What the meaning of this warning? If I use this mesh on Fluent simulation, is there any effect about this warning?

Thank you very much for the reply.
It just means that your prism layers are not all fully developed. So if you wanted 15 layers everywhere, but there wasn't room for 15 layers in some places, it will "step down" to 14 or less. Some solvers don't mind stair stepping at all (CFX), Fluent prefers full layers, but it really depends on where the stair stepping is.
__________________
-----------------------------------------
Please help guide development at ANSYS by filling in these surveys

Public ANSYS ICEM CFD Users Survey

This second one is more general (Gambit, TGrid and ANSYS Meshing users welcome)...

CFD Online Users Survey
PSYMN is offline   Reply With Quote

Old   January 15, 2013, 00:55
Default VOF models on venturi scrubber simulation
  #12
New Member
 
Join Date: Feb 2012
Posts: 6
Rep Power: 5
toda is on a distinguished road
Dear all.


My simulation is about venturi scrubber. A venturi scrubber is designed to effectively use the energy from the inlet gas stream to atomize the liquid being used to scrub the gas stream, so the gas stream will more clean. The end of the inlet water pipe there is some small nozzles. This venturi have 2 inlet and 2 outlet. Two inlet are inlet water and inlet gas. Because flow in venturi scrubber is turbulent, I make inflation in region near the venturi scrubber wall. The boundary conditions are inlet water=velocity inlet, inlet gas=velocity inlet, outlet1 and outlet2 = pressure outlet. In Fluent, I use models: Multiphase VOF, Energy Equation, and Viscous is k-epsilon. Multiphase have 2 eulerian phase, that is air and water liquid. The solution method is PISO. The absolute criteria of convergence is 1e-4 except energy (energy is 1e-6). After 7903 iteration, the simulation still not convergen. The simulation also have reversed flow in outlet1 and outlet2. Anyone can help me to solve this problem?? Is the multiphase VOF true for this problem?? Any help is highly appreciated. Thanks.

I`m sorry if the discussion is out of ansys meshing topic.

Best Regards
Toda
Attached Images
File Type: jpg Venturi Scrubber_1.jpg (35.6 KB, 30 views)
File Type: jpg Nozzle_2.JPG (19.2 KB, 27 views)
File Type: jpg Nozzle.JPG (18.2 KB, 22 views)
File Type: jpg Iteration.jpg (83.3 KB, 31 views)
File Type: jpg Mass Flow Rate.jpg (33.1 KB, 21 views)
toda is offline   Reply With Quote

Old   January 15, 2013, 00:59
Default VOF models on venturi scrubber simulation
  #13
New Member
 
Join Date: Feb 2012
Posts: 6
Rep Power: 5
toda is on a distinguished road
The more attachments about my simulation.

Best Regards
Toda
Attached Images
File Type: jpg Contours Total Temperature Magnitude.jpg (89.3 KB, 29 views)
File Type: jpg Contours Velocity Magnitude.jpg (94.5 KB, 21 views)
toda is offline   Reply With Quote

Reply

Thread Tools
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 On
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
tetrahedral or hexahedral mesh? srr ANSYS Meshing & Geometry 4 September 23, 2010 12:56
Tetrahedral and hexahedral mesh based on FVM kuang Main CFD Forum 0 July 20, 2008 04:52
Hexahedral or Tetrahedral Maria Angelica NUMECA 3 February 19, 2007 01:10
hexahedral vs. tetrahedral cells Christian Main CFD Forum 3 March 21, 2006 07:41
Tetrahedral vs. Hexahedral Meshing for CFD Andy Bartels Main CFD Forum 14 August 22, 2000 07:35


All times are GMT -4. The time now is 22:50.