[cfMesh] feature edge error

zordiack · September 26, 2016, 07:13

Hello dear fellow foamers. I've been struggling with cfMesh again with a fairly complex geometry, more specifically with the feature edges. I'm using cfMesh v1.1.1 with OpenFOAM-dev. My workflow is currently the following:

1. Create surface stl meshes for all patches
2. Combile stl-files to create a single surface file
3. surfaceFeatureEdges -angle (something) whole.stl whole.fms
4. FMSToSurface -exportFeatureEdges whole.fms temp.fms

What you see in the following picture is the original whole.stl file, which btw has good resolution and correct edges, and the result of step 4 which is the file temp_featureEdges.vtk.

As you can see, at some point the feature edges are moved out of place and it can only happen in steps 3 or 4 which use cfMesh tools. I'm currently suspecting that this is some kind of numerical error or possibly an error related to .vtk format and Paraview? In any case, the feature edges really are wrong, since cfMesh doesn't use them and the resulting mesh can be seen in the next picture.

This isn't even the worst part in the geometry, but the others result in similar bad cells when the edge is not detected correctly. If someone has some insight on this I'd be happy to get some help.

franjo_j · October 11, 2016, 05:43

Hello,

I do not see nothing wrong in the procedure for preparing the geometry. You may need to play with the angle tolerance to select the edges at all relevant locations.

It is common to see the feature edges out of place, and it is a consequence of the ascii vtk that loses precision when it is written on disk.

cfMesh does not capture each feature edge explicitly. It searches for faces in the surface mesh that are bounded by a closed loop of feature edges. I somehow suspect that you do not have a closed loop at the location where the feature edges are not captured.

zordiack · October 11, 2016, 05:46

So it's not possible to use feature edges that do not form a closed loop? Because that is exactly the case regarding the area in the picture. Some feature edges just don't form a closed loop due to for example roundings.

franjo_j · October 11, 2016, 06:09

Correct, the edges shall be closed and form a face.

September 26, 2016, 07:13	feature edge error	#1
zordiack Member Pekka Pasanen Join Date: Feb 2012 Location: Finland Posts: 87 Rep Power: 14	Hello dear fellow foamers. I've been struggling with cfMesh again with a fairly complex geometry, more specifically with the feature edges. I'm using cfMesh v1.1.1 with OpenFOAM-dev. My workflow is currently the following: 1. Create surface stl meshes for all patches 2. Combile stl-files to create a single surface file 3. surfaceFeatureEdges -angle (something) whole.stl whole.fms 4. FMSToSurface -exportFeatureEdges whole.fms temp.fms What you see in the following picture is the original whole.stl file, which btw has good resolution and correct edges, and the result of step 4 which is the file temp_featureEdges.vtk. As you can see, at some point the feature edges are moved out of place and it can only happen in steps 3 or 4 which use cfMesh tools. I'm currently suspecting that this is some kind of numerical error or possibly an error related to .vtk format and Paraview? In any case, the feature edges really are wrong, since cfMesh doesn't use them and the resulting mesh can be seen in the next picture. This isn't even the worst part in the geometry, but the others result in similar bad cells when the edge is not detected correctly. If someone has some insight on this I'd be happy to get some help.

October 11, 2016, 05:43		#2
franjo_j Senior Member Franjo Juretic Join Date: Aug 2011 Location: Velika Gorica, Croatia Posts: 124 Rep Power: 16	Hello, I do not see nothing wrong in the procedure for preparing the geometry. You may need to play with the angle tolerance to select the edges at all relevant locations. It is common to see the feature edges out of place, and it is a consequence of the ascii vtk that loses precision when it is written on disk. cfMesh does not capture each feature edge explicitly. It searches for faces in the surface mesh that are bounded by a closed loop of feature edges. I somehow suspect that you do not have a closed loop at the location where the feature edges are not captured. schiffbauer likes this. __________________ Principal Developer of cfMesh and CF-MESH+ www.cfmesh.com Social media: LinkedIn, Twitter, YouTube, Facebook, Pinterest, Instagram

October 11, 2016, 06:09		#4
franjo_j Senior Member Franjo Juretic Join Date: Aug 2011 Location: Velika Gorica, Croatia Posts: 124 Rep Power: 16	Correct, the edges shall be closed and form a face. __________________ Principal Developer of cfMesh and CF-MESH+ www.cfmesh.com Social media: LinkedIn, Twitter, YouTube, Facebook, Pinterest, Instagram

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Building OpenFOAM1.7.0 from source	ata	OpenFOAM Installation	46	March 6, 2022 13:21
[swak4Foam] installing funkySetFields	igo	OpenFOAM Community Contributions	1	November 20, 2012 20:16
Compile problem	ivanyao	OpenFOAM Running, Solving & CFD	1	October 12, 2012 09:31
[swak4Foam] groovyBC: problems compiling: "flex: not found" and "undefined reference to ..."	sega	OpenFOAM Community Contributions	12	February 17, 2010 09:30
How to get the max value of the whole field	waynezw0618	OpenFOAM Running, Solving & CFD	4	June 17, 2008 05:07

October 11, 2016, 05:46		#3
zordiack Member Pekka Pasanen Join Date: Feb 2012 Location: Finland Posts: 87 Rep Power: 14	So it's not possible to use feature edges that do not form a closed loop? Because that is exactly the case regarding the area in the picture. Some feature edges just don't form a closed loop due to for example roundings.