nonManifoldPoints when using fluent3DMeshToFoam

Jasper Z · November 21, 2020, 00:44

Hi all,

I was wondering what does nonManifoldPoints mean? the checkMesh writes:

"<<Writing 503452 conflicting points to set nonManifoldPoints"
But I am running my case normally. Can anyone please explain what's the meaning of nonManifoldPoints? Will it affect the results?
Thank you in advance!

geth03 · November 23, 2020, 03:31

did you define a boundary condition for the part of the internal mesh?

jorkolino · August 9, 2021, 05:43

Quote:

Originally Posted by geth03

did you define a boundary condition for the part of the internal mesh?

What do you mean by boundary condition on internal mesh? I thought boundary conditions are only specified on boundaries

geth03 · August 10, 2021, 04:08

sorry, i should have explained it, sounds weird i know.

i think the op did create a geometry and when he tried to create the mesh
it did not work out the way he originally intended.

by boundary condition in the internal mesh i wanted to know if he declared an internal face as wall. bc this wall would be recognised from one side, the other side could lead to problems, i.e. non manifold points.

when somebody asks a question, he/she should provide some detail, more data. otherwise i need to ask for it or just speculate.

it would be nice if he/she could export the points to VTK and show an image from the points and the mesh, so we know what he/she did.

jorkolino · August 10, 2021, 11:37

Well, I have the same problem, without any baffles (internal surfaces) in the domain. It is a simple box exported by a mesh generator into mesh format (.msh). A box with 6 external walls. However, the mesh generator also exports the internal points as default-interior. They are not part of any boundary patch or continuum patch (at least not user-defined), but for some reason the mesher does some definition of them in the mesh file. After import in OF with fluent3DMeshToFoam, OF complains to find non-manifold points and also puts them into a separate set. OF doesn't seem to make much use of this set, so I wonder, if I should be worried by the presence of non-manifold points from a computational point of view.

geth03 · August 11, 2021, 03:08

when i had that problem, OF complained also, the number of iterations was high but the calculation time was similar to the simulation with corrected mesh.

how did you create your geometry with ansys spaceclaim?
did you create multiple components and then shared the faces?

could you show the mesh and the non-manifold points?

November 21, 2020, 00:44	nonManifoldPoints when using fluent3DMeshToFoam	#1
Jasper Z Member Huan Zhang Join Date: Nov 2020 Posts: 55 Rep Power: 5	Hi all, I was wondering what does nonManifoldPoints mean? the checkMesh writes: "<<Writing 503452 conflicting points to set nonManifoldPoints" But I am running my case normally. Can anyone please explain what's the meaning of nonManifoldPoints? Will it affect the results? Thank you in advance!

August 10, 2021, 11:37	Manifold points even without internal surfaces	#5
jorkolino Member George Pichurov Join Date: Jul 2010 Posts: 52 Rep Power: 15	Well, I have the same problem, without any baffles (internal surfaces) in the domain. It is a simple box exported by a mesh generator into mesh format (.msh). A box with 6 external walls. However, the mesh generator also exports the internal points as default-interior. They are not part of any boundary patch or continuum patch (at least not user-defined), but for some reason the mesher does some definition of them in the mesh file. After import in OF with fluent3DMeshToFoam, OF complains to find non-manifold points and also puts them into a separate set. OF doesn't seem to make much use of this set, so I wonder, if I should be worried by the presence of non-manifold points from a computational point of view. wht likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
fluent3DMeshToFoam and mergeMeshes crash with large (around 170 mio cells) meshes	maxdre91	OpenFOAM Pre-Processing	2	April 27, 2022 08:44
fluent3DMeshToFoam - (periodic) cyclic boundary	cyln	OpenFOAM Running, Solving & CFD	0	October 1, 2017 23:19
[Commercial meshers] fluent3DMeshToFoam conversion problem	CFDnewbie147	OpenFOAM Meshing & Mesh Conversion	14	March 12, 2014 05:16
Possible Bug in pimpleFoam (or createPatch) (or fluent3DMeshToFoam)	cfdonline2mohsen	OpenFOAM	3	October 21, 2013 09:28
OpenFOAM command from inside MATLAB	sega	OpenFOAM Post-Processing	18	September 25, 2012 07:35

November 23, 2020, 03:31		#2
geth03 Senior Member Join Date: Dec 2019 Location: Cologne, Germany Posts: 355 Rep Power: 8	did you define a boundary condition for the part of the internal mesh?

August 10, 2021, 04:08		#4
geth03 Senior Member Join Date: Dec 2019 Location: Cologne, Germany Posts: 355 Rep Power: 8	sorry, i should have explained it, sounds weird i know. i think the op did create a geometry and when he tried to create the mesh it did not work out the way he originally intended. by boundary condition in the internal mesh i wanted to know if he declared an internal face as wall. bc this wall would be recognised from one side, the other side could lead to problems, i.e. non manifold points. when somebody asks a question, he/she should provide some detail, more data. otherwise i need to ask for it or just speculate. it would be nice if he/she could export the points to VTK and show an image from the points and the mesh, so we know what he/she did.

August 11, 2021, 03:08		#6
geth03 Senior Member Join Date: Dec 2019 Location: Cologne, Germany Posts: 355 Rep Power: 8	when i had that problem, OF complained also, the number of iterations was high but the calculation time was similar to the simulation with corrected mesh. how did you create your geometry with ansys spaceclaim? did you create multiple components and then shared the faces? could you show the mesh and the non-manifold points?