Complex Interface Mesh

Jonny6001 · February 12, 2011, 07:05

Hello,

I have a case which has a fairly complex interface between solid-fluid and solid-gas.

I have a nice surface mesh on all parts. I then create the contact in-place interface between the relevant faces. I then re-run the surface re-mesher because the interface need to be conformal. I then run my volume mesh.

The overall quality of the mesh is good but on some of the interface faces, I have small patches with what looks like holes. They are not holes in the actual surface because the leak detector doesn't find any leaks plus the volume mesher wouldn't run if they were holes.

I am sure that they are not treated as holes. My feeling is they are cells that could have been outside the interface tolerance and therefore are not treated as being part of the interface.

Have you guys come accross this issue? Could it be solved by refining the surface meshes?
I have attached a close-up of an example in the link below.

Thanks a lot for your time.

http://img153.imageshack.us/i/64085950.png/

xamo · February 14, 2011, 03:27

Hello Jonny6001,

check the mesh at the interface after running the surface remesher. I made the experience, that the remesher sometimes builds two different meshes on the contact interface.

If your interface 2d mesh is ok, you do not need to run the surface remesher after creating the in-place interface. Just build the volume mesh and everything should be ok.

Regards
xamo

and_ · February 15, 2011, 10:27

Are you showing just the interface or also the corresponding wall in the picture?
This looks like the interface was not created in all areas. Some cells are still part of the wall.
It is advisable to use the mult-region imprint tool, that is part of the surface repair tool, before remeshing.
This tool converts to seperate boundaries, that are part of different regions into one interface. That way mismatch between interface boundaries during the meshing process is avoided.
In that tool you can specify a tolerance and automatically browse through contacting boundary faces

Jonny6001 · February 16, 2011, 05:59

Thanks a lot for your help.
I have sorted it out. The surfaces are very high quality so I didn't have to perform any repair. I changed some of the surface mesh in the regions where there was tight curvature and both sides of the interface slowly converge.
I guess if you can afford to use exactly the same mesh specs on both regions, the interface would naturally be close.

It was exactly as you described, some of the faces outside the interface tolerance were left in the original wall.

The multi-region imprint and surface repair toolbox is something I will certainly be interested in learning. I usually spend all my time using Catia to prep the surfaces perfectly, re-import in to Catia as an IGES then re-check the surfaces after the IGES conversion.
I have extensively used the surface wrapper but would prefer if I didn't have to.

Again, thanks a lot to both of you.

February 12, 2011, 07:05	Complex Interface Mesh	#1
Jonny6001 Member Jonny Join Date: Aug 2009 Posts: 72 Rep Power: 17	Hello, I have a case which has a fairly complex interface between solid-fluid and solid-gas. I have a nice surface mesh on all parts. I then create the contact in-place interface between the relevant faces. I then re-run the surface re-mesher because the interface need to be conformal. I then run my volume mesh. The overall quality of the mesh is good but on some of the interface faces, I have small patches with what looks like holes. They are not holes in the actual surface because the leak detector doesn't find any leaks plus the volume mesher wouldn't run if they were holes. I am sure that they are not treated as holes. My feeling is they are cells that could have been outside the interface tolerance and therefore are not treated as being part of the interface. Have you guys come accross this issue? Could it be solved by refining the surface meshes? I have attached a close-up of an example in the link below. Thanks a lot for your time. http://img153.imageshack.us/i/64085950.png/

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Gambit problems	Althea	FLUENT	22	January 4, 2017 04:19
[Gmsh] 2D Mesh Generation Tutorial for GMSH	aeroslacker	OpenFOAM Meshing & Mesh Conversion	12	January 19, 2012 04:52
VOF interface compression and mesh size	idrama	OpenFOAM	1	September 26, 2010 07:36
Dynamic Mesh moving interface help	akash.iitb	FLUENT	0	August 24, 2010 00:53
Mesh Quality at the Interface	Can Atli	FLUENT	0	November 23, 2005 11:32

February 14, 2011, 03:27		#2
xamo New Member Join Date: Mar 2009 Posts: 20 Rep Power: 17	Hello Jonny6001, check the mesh at the interface after running the surface remesher. I made the experience, that the remesher sometimes builds two different meshes on the contact interface. If your interface 2d mesh is ok, you do not need to run the surface remesher after creating the in-place interface. Just build the volume mesh and everything should be ok. Regards xamo

February 15, 2011, 10:27		#3
and_ New Member Join Date: Jul 2010 Posts: 11 Rep Power: 16	Are you showing just the interface or also the corresponding wall in the picture? This looks like the interface was not created in all areas. Some cells are still part of the wall. It is advisable to use the mult-region imprint tool, that is part of the surface repair tool, before remeshing. This tool converts to seperate boundaries, that are part of different regions into one interface. That way mismatch between interface boundaries during the meshing process is avoided. In that tool you can specify a tolerance and automatically browse through contacting boundary faces

February 16, 2011, 05:59		#4
Jonny6001 Member Jonny Join Date: Aug 2009 Posts: 72 Rep Power: 17	Thanks a lot for your help. I have sorted it out. The surfaces are very high quality so I didn't have to perform any repair. I changed some of the surface mesh in the regions where there was tight curvature and both sides of the interface slowly converge. I guess if you can afford to use exactly the same mesh specs on both regions, the interface would naturally be close. It was exactly as you described, some of the faces outside the interface tolerance were left in the original wall. The multi-region imprint and surface repair toolbox is something I will certainly be interested in learning. I usually spend all my time using Catia to prep the surfaces perfectly, re-import in to Catia as an IGES then re-check the surfaces after the IGES conversion. I have extensively used the surface wrapper but would prefer if I didn't have to. Again, thanks a lot to both of you.