[GAMBIT] Mesh quality

prince_pahariaa · May 2, 2012, 03:58

Hey friends

Sorry for big explanation but i cant figure out in short.

I am dealing with very complicated geometry for which i need to perform heat transfer analysis. Detail of geometry is like this.

1) Their is 3 cylindrical capsule inside which heat generation occurs of height 40mm. It is partly filled with powder substance. Heat is mainly generated from this powder solid.

2) To contain these three cylinder, there are 3 holders of height 20mm. Each holders has 4 holes from where coolant can pass through. Holders are solid made up of Aluminium.

3) There is solid Aluminium base which support these three holder. And to support base, there is one guide tube (Al) which run from top to bottom of flow area. There is a plug to tighten the capsule at its place.

I have not considered the thickness of capsule in my analysis. But every other component thickness is considered. All Aluminium part has a heat generation of 0.5W/gm.

Then i made a cylindrical flow area which cover all components. I use split operation to split all solid component from flow area.

Used T/Grid mesh..
I am getting one highly skewed element (<0.97).

My question is am i good to go for fluent part ??
I have an impression that mesh quality should not be more than 0.9.
But i am not able to generate such mesh.

For all the component i can generate hexa mesh. But for cylindrical flow geometry i have to use split operation and i can not use hexa there.

Need any suggestion to improve my geometry
or with this complicated geometry this quality is good..

I am enclosing diagram of my geometry.

I appreciate any help.

-mAx- · May 2, 2012, 05:07

I don't know TGrid, but in Gambit you can check your quality mesh and see where are skewed cells >> http://my.fit.edu/itresources/manual....htm#gc_exmesh

Did you try to mesh your domain on the fly? I mean you only have one volume?

Far · May 2, 2012, 06:29

You can use cut cell meshing method in Tgrid. The mesh quality of 0.97 can be ok or bad, depending on if it in important flow region or not.

Are you making surface mesh in gambit and if yes, whats the quality you get for surface mesh?

prince_pahariaa · May 2, 2012, 08:31

Hello Max
Thanks for replying.

I did try to mesh the volume in single shot but in vain. I have examined the mesh using examine tool only keeping quality type "equisize skew" and switching on "show worst element button". There i was getting worst quality of about 0.99 which i guess is not acceptable.

I have single flow volume but in my problem i have many other components. I meshed them too and export it to fluent. For heat transfer problem i guess i need to mesh that too.. needn't I ?? As heat is generated in wall ??

I excluded the plug part from my geometry and tried mesh. In this case i am getting worst mesh quality of about 0.9 which might be okay. I will try with plug after this attempt.

Let me know if anything you can suggest to make better grid.

Once again
Thanks

-mAx- · May 3, 2012, 00:59

in Gambit go to examine mesh, set the lower limit at 0.9 and let the upper at 1.
When you apply, then Gambit will show you all cells which are skewed.
In 95% of cases those cells are generated because of your geometry.
Just check if in the neighbourhood of those cells you have any sharp angle, small edge, etc...
For heat transfer, I don't have any experience, but I believe you can use Shell conduction in Fluent (so no need to modelize any "depth" of your wall). But as I said, I am not experienced in this field

prince_pahariaa · May 4, 2012, 07:47

Hello Far

Sorry for replying late.
I did not make any surface mesh.. I directly used volume mesh tool.
I did get bad quality mesh in area of interest but there is only one grid which is more than 0.97. Will that be a problem ??

In mean time, i have simplified my geometry a little and get acceptable quality (worst element 0.86). Right now i am trying with it.

How the surface mesh help in improving quality ??
Should i mesh the surface first and then go for volume mesh ??

From my experience (although its very little), i never observe any good improvement in mesh quality by this method. May be i am doing it in wrong way.. I will appreciate if u let me know how it actually help ??

May 2, 2012, 05:07		#2
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	I don't know TGrid, but in Gambit you can check your quality mesh and see where are skewed cells >> http://my.fit.edu/itresources/manual....htm#gc_exmesh Did you try to mesh your domain on the fly? I mean you only have one volume? __________________ In memory of my friend Hervé: CFD engineer & freerider

May 3, 2012, 00:59		#5
-mAx- Super Moderator Maxime Perelli Join Date: Mar 2009 Location: Switzerland Posts: 3,297 Rep Power: 41	in Gambit go to examine mesh, set the lower limit at 0.9 and let the upper at 1. When you apply, then Gambit will show you all cells which are skewed. In 95% of cases those cells are generated because of your geometry. Just check if in the neighbourhood of those cells you have any sharp angle, small edge, etc... For heat transfer, I don't have any experience, but I believe you can use Shell conduction in Fluent (so no need to modelize any "depth" of your wall). But as I said, I am not experienced in this field __________________ In memory of my friend Hervé: CFD engineer & freerider

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
[snappyHexMesh] No layers in a small gap	bobburnquist	OpenFOAM Meshing & Mesh Conversion	6	August 26, 2015 09:38
Polyhedral Mesh Quality in Star-CCM+	niazaliahmed	STAR-CCM+	3	March 8, 2012 13:51
[ICEM] Tetra mesh quality before and after prism layer	Chander	ANSYS Meshing & Geometry	0	December 25, 2011 22:04
fluent add additional zones for the mesh file	SSL	FLUENT	2	January 26, 2008 11:55
Icemcfd 11: Loss of mesh from surface mesh option?	Joe	CFX	2	March 26, 2007 18:10

May 2, 2012, 06:29		#3
Far Super Moderator Sijal Join Date: Mar 2009 Location: Islamabad Posts: 4,553 Blog Entries: 6 Rep Power: 54	You can use cut cell meshing method in Tgrid. The mesh quality of 0.97 can be ok or bad, depending on if it in important flow region or not. Are you making surface mesh in gambit and if yes, whats the quality you get for surface mesh?

May 2, 2012, 08:31		#4
prince_pahariaa Member prince Join Date: Jun 2011 Posts: 56 Rep Power: 14	Hello Max Thanks for replying. I did try to mesh the volume in single shot but in vain. I have examined the mesh using examine tool only keeping quality type "equisize skew" and switching on "show worst element button". There i was getting worst quality of about 0.99 which i guess is not acceptable. I have single flow volume but in my problem i have many other components. I meshed them too and export it to fluent. For heat transfer problem i guess i need to mesh that too.. needn't I ?? As heat is generated in wall ?? I excluded the plug part from my geometry and tried mesh. In this case i am getting worst mesh quality of about 0.9 which might be okay. I will try with plug after this attempt. Let me know if anything you can suggest to make better grid. Once again Thanks

May 4, 2012, 07:47		#6
prince_pahariaa Member prince Join Date: Jun 2011 Posts: 56 Rep Power: 14	Hello Far Sorry for replying late. I did not make any surface mesh.. I directly used volume mesh tool. I did get bad quality mesh in area of interest but there is only one grid which is more than 0.97. Will that be a problem ?? In mean time, i have simplified my geometry a little and get acceptable quality (worst element 0.86). Right now i am trying with it. How the surface mesh help in improving quality ?? Should i mesh the surface first and then go for volume mesh ?? From my experience (although its very little), i never observe any good improvement in mesh quality by this method. May be i am doing it in wrong way.. I will appreciate if u let me know how it actually help ??