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

[ICEM] Failed Tet Mesh on Surface

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

Like Tree1Likes
  • 1 Post By kad

Reply
 
LinkBack Thread Tools Display Modes
Old   March 5, 2017, 01:20
Default Failed Tet Mesh on Surface
  #1
New Member
 
SakshaatDevata
Join Date: Dec 2016
Posts: 22
Rep Power: 2
anand32 is on a distinguished road
Hi everyone

I am trying to mesh some tubes contained in a cuboid . The tubes are closed and hollow while the cuboid is filled with air. The trouble is that while the surfaces of the cuboid gets meshed , most of the surface of the tubes is not meshed. They get enclosed as subset named "Failed Surfaces" under Geoemetry in the display tree area.

I am using the Quick mesh algorithm, combined with all Tri Patch dependent surface mesh.

COuld you please give some advice?
anand32 is offline   Reply With Quote

Old   March 6, 2017, 10:15
Default
  #2
Senior Member
 
Kushal Puri
Join Date: Nov 2013
Posts: 103
Rep Power: 5
Kushal Puri is on a distinguished road
Quote:
Originally Posted by anand32 View Post
Hi everyone

I am trying to mesh some tubes contained in a cuboid . The tubes are closed and hollow while the cuboid is filled with air. The trouble is that while the surfaces of the cuboid gets meshed , most of the surface of the tubes is not meshed. They get enclosed as subset named "Failed Surfaces" under Geoemetry in the display tree area.

I am using the Quick mesh algorithm, combined with all Tri Patch dependent surface mesh.

COuld you please give some advice?
Use patch independent mesher
Kushal Puri is offline   Reply With Quote

Old   March 6, 2017, 11:20
Default
  #3
kad
Senior Member
 
Join Date: Feb 2013
Location: Germany
Posts: 181
Rep Power: 16
kad will become famous soon enoughkad will become famous soon enough
For patch dependent meshing the topology of your geometry must be intact. This means curves have to be attached to surfaces and so on. There is a description for this in the manual. For building topology you can use "Build Diagnostic Topology" under "Repair Geometry". Further choose appropriate mesh sizes on surfaces and curves.
anand32 likes this.
kad is offline   Reply With Quote

Old   March 12, 2017, 20:13
Default
  #4
New Member
 
SakshaatDevata
Join Date: Dec 2016
Posts: 22
Rep Power: 2
anand32 is on a distinguished road
Thank you Kad.

I am learning ICEM right now and what I have learnt and understood about Build Topology option is written below. Could you let me know whether I am correct:

Build Topology App generally does these things:

1. It extracts curves from surfaces:
a. If the curves are not there on the edges of the surfaces , they are created.
b. If there are curves on the surface but are dormant (deleted not permanently), they are activated.
c. If there are curves on the surface that are already active, they are retained.
2. If two surfaces are separated by a distance < tolerance , then they are sort of joined and a red curve is created between them. This makes Build topology option critical for patch dependent meshing..
3. If the Filter Angles And Filter Curves option is switched on, then those points (curves) are made dormant , where two curves (surfaces) meet at an angle less than the specified angle.
4. It color codes the curves; the curves which are attached to two surfaces are red , to one surface is yellow, and to more than 2 surfaces is blue. A yellow curve after build topology means that volume meshing will most probably not work, displaying the error "there are holes in your geometry".
Filter Points and Filter Curves option is switched on only in the case when we need to create/extract curves. If the aim is to only check connectivity, then this option is switched off.


Thank you
Vishal
anand32 is offline   Reply With Quote

Old   March 14, 2017, 06:42
Default
  #5
kad
Senior Member
 
Join Date: Feb 2013
Location: Germany
Posts: 181
Rep Power: 16
kad will become famous soon enoughkad will become famous soon enough
I think you got most of the points correct. I never really investigated in this, but Build Topology should not be critical for patch dependent meshing even if there is e.g. a minor gap between surfaces. As long as surfaces are correctly connected to curves (and the actual mesh size is larger than the gap size) patch dependent surface meshing should work. As I stated before a working topolgy is a must for some patch dependent meshing.

Point 4: As long as you are using Octree for creating surface mesh it should somehow be able to deal with a "dirty" geometry. This means Octree might work without a proper topolgy. It is capable of meshing even across small gaps. That it why it is also called "Robust" in ICEM.
kad is offline   Reply With Quote

Old   March 14, 2017, 11:08
Default
  #6
New Member
 
SakshaatDevata
Join Date: Dec 2016
Posts: 22
Rep Power: 2
anand32 is on a distinguished road
Quote:
Originally Posted by kad View Post
I think you got most of the points correct. I never really investigated in this, but Build Topology should not be critical for patch dependent meshing even if there is e.g. a minor gap between surfaces. As long as surfaces are correctly connected to curves (and the actual mesh size is larger than the gap size) patch dependent surface meshing should work. As I stated before a working topolgy is a must for some patch dependent meshing.

Point 4: As long as you are using Octree for creating surface mesh it should somehow be able to deal with a "dirty" geometry. This means Octree might work without a proper topolgy. It is capable of meshing even across small gaps. That it why it is also called "Robust" in ICEM.

Thank you Kad for your reply.

To clarify things further, does that mean

1. For Octree mesh:

Since Octree is a geometry mesher, it does not really care about the surface. So whatever parameter we specify in the "Shell Meshing parameters", like "All Tri" Or "Patch Dependent/Patch Independent" are really irrelevant to Octree meshing.

2. For Patch Dependent Mesh:
This meshing type requires that surfaces should have curves associated with them. "Build Topology" is only one way to ensure that. If surfaces are already attached to curves, build topology may not be required at all; It is a good practice and safe practice to Build Topology before Patch Dependent Meshing, but not really a required practice.

Let me know if I am correct or is there any scope for refining my knowledge in this area.

Thank you

Last edited by anand32; March 14, 2017 at 11:09. Reason: add some material
anand32 is offline   Reply With Quote

Old   March 14, 2017, 11:48
Default
  #7
kad
Senior Member
 
Join Date: Feb 2013
Location: Germany
Posts: 181
Rep Power: 16
kad will become famous soon enoughkad will become famous soon enough
Quote:
Originally Posted by anand32 View Post
Thank you Kad for your reply.

To clarify things further, does that mean

1. For Octree mesh:

Since Octree is a geometry mesher, it does not really care about the surface. So whatever parameter we specify in the "Shell Meshing parameters", like "All Tri" Or "Patch Dependent/Patch Independent" are really irrelevant to Octree meshing.
Octree does care about some of the parameters set under "Shell Meshing parameters" e.g. maxsizes. However you more control over the mesh by using patch dependent method because it respects more of those parameters. But patch dependent is also harder to handle. See the various reasons stated before.

You have to be aware that patch dependent and patch independent are two totally different algorithms. The first one is a "real" 2D algorithm that uses cloosed loops that are defined by surfaces and curves. Patch independent method always uses the 3D Octree algorithm and throws the volume mesh automatically away for you. You can check this by surface meshing a (closed) volume with patch independent surface mesher and with octree volume mesher. You can see that the result and even the output in the terminal is almost identical. A detailed description of the Octree algorithm can be found in the manual. Further internally it always produces tri elements.

For 3D Volume Meshing with Octree the entries for "All Tri" Or "Patch Dependent/Patch Independent" are irrelevant, as they only apply to 2D Meshing. For 2D Meshing the entries for "Mesh type" are relevant in both cases. But for patch dependent e.g. quad elements are created from tri elements as the octree algorithm produces pure tri elements.

Quote:
Originally Posted by anand32 View Post

2. For Patch Dependent Mesh:
This meshing type requires that surfaces should have curves associated with them. "Build Topology" is only one way to ensure that. If surfaces are already attached to curves, build topology may not be required at all; It is a good practice and safe practice to Build Topology before Patch Dependent Meshing, but not really a required practice.

Let me know if I am correct or is there any scope for refining my knowledge in this area.

Thank you
This is correct. You are always on the safe side if you check topology first e.g. by coloring curves by count (RMB on curves in Display tree).
kad 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
SHM problem : KVLCC2 with appendage mesh sc.park OpenFOAM Native Meshers: snappyHexMesh and Others 1 March 13, 2016 14:28
No layers in a small gap bobburnquist OpenFOAM Native Meshers: snappyHexMesh and Others 6 August 26, 2015 09:38
Layers:problem with curvature giulio.topazio OpenFOAM Native Meshers: snappyHexMesh and Others 10 August 22, 2012 09:03
2D Mesh Generation Tutorial for GMSH aeroslacker Open Source Meshers: Gmsh, Netgen, CGNS, ... 12 January 19, 2012 04:52
external flow with snappyHexMesh chelvistero OpenFOAM 11 January 15, 2010 20:43


All times are GMT -4. The time now is 00:21.