[snappyHexMesh] different sHM results on same geometry when everythin is in one or several stl-files.

Laika · September 6, 2016, 15:46

Dear forum,

I'm working towards a multiregion case but I keep hitting problems for the first region...

I have a geometry defined as a set of stl-files that combined define a closed region. When thrown at sHM with surfaceFeatureRefinement and proper settings, this produces a nice mesh of the inside of my volume. A finer mesh version of it gave me nice results in an actual simulation.

When I combine all these stl-files in one single stl file, sHM does not produce the desired mesh: it meshes both the inside and the outside of my mesh as if the volume wasn't closed.
That's strange, as the surfaces in the combined stl are exactly the same as the surfaces in the original set of separate one-subsurface-per-file set of stl-files.

No surfaces are missing in the combined file: I extracted the subsurfaces from the combined stl, and with this set of files, sHM produces the desired mesh easily.

Clearly there are some fundamental differences in how an stl with multiple regions is treated compared to a set of separate stl files.
Can someone shed a light on these differences please?
Is this problem familiar to someone?
How can I mesh the inside of the domain in my merged stl-files?

(My final goal is to have several unconnected similar flow domain in one simulation. As the multiregion sHM procedure needs stl-files of closed volumes only, it would be a good start to get one such geometry definition running...)

thanks in advance for any help or useful hints,

kind regards,
Laika,
still orbiting

Laika · September 8, 2016, 04:09

To make my question a bit clearer:

My geometry is defined by a set of stl surfaces that do not form the watertight closed surface one would like to have, as there are edges that are only connected to one triangle. Aha, one would say, you just need to make the good stl file with all connected facets!

OK, but then why can I make a nice mesh when I put these same faceted surfaces in separate stl-files? It are still the exact same facets with the open edges. This time sHM recognises they form a closed volume, nicely refines near these faces, and correctly keeps the inside and throws out the outside.

Why does it matter whether the surfaces are in one file, or in separate files?

On the other hand, this might be a good way to proceed when you have a geometrically well defined volume, but with edgewise-unconnected patches: simply split the file in separate one-patch-per-file stl-files.

I hope this triggers some thinking...

best regards,
Laika,
still orbiting
laika@inorbit.eu

September 6, 2016, 15:46	different sHM results on same geometry when everythin is in one or several stl-files.	#1
Laika New Member Join Date: Jun 2009 Location: Belgium Posts: 11 Rep Power: 16	Dear forum, I'm working towards a multiregion case but I keep hitting problems for the first region... I have a geometry defined as a set of stl-files that combined define a closed region. When thrown at sHM with surfaceFeatureRefinement and proper settings, this produces a nice mesh of the inside of my volume. A finer mesh version of it gave me nice results in an actual simulation. When I combine all these stl-files in one single stl file, sHM does not produce the desired mesh: it meshes both the inside and the outside of my mesh as if the volume wasn't closed. That's strange, as the surfaces in the combined stl are exactly the same as the surfaces in the original set of separate one-subsurface-per-file set of stl-files. No surfaces are missing in the combined file: I extracted the subsurfaces from the combined stl, and with this set of files, sHM produces the desired mesh easily. Clearly there are some fundamental differences in how an stl with multiple regions is treated compared to a set of separate stl files. Can someone shed a light on these differences please? Is this problem familiar to someone? How can I mesh the inside of the domain in my merged stl-files? (My final goal is to have several unconnected similar flow domain in one simulation. As the multiregion sHM procedure needs stl-files of closed volumes only, it would be a good start to get one such geometry definition running...) thanks in advance for any help or useful hints, kind regards, Laika, still orbiting

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
UDF issue	MASOUD	Fluent UDF and Scheme Programming	14	December 6, 2012 13:39
[snappyHexMesh] STL - File and sHM	Tobi	OpenFOAM Meshing & Mesh Conversion	0	November 14, 2010 08:12
critical error during installation of openfoam	Fabio88	OpenFOAM Installation	21	June 2, 2010 03:01
OpenFOAM15 paraFoam bug	koen	OpenFOAM Bugs	19	June 30, 2009 10:46
Results saving in CFD	hawk	Main CFD Forum	16	July 21, 2005 20:51

September 8, 2016, 04:09		#2
Laika New Member Join Date: Jun 2009 Location: Belgium Posts: 11 Rep Power: 16	To make my question a bit clearer: My geometry is defined by a set of stl surfaces that do not form the watertight closed surface one would like to have, as there are edges that are only connected to one triangle. Aha, one would say, you just need to make the good stl file with all connected facets! OK, but then why can I make a nice mesh when I put these same faceted surfaces in separate stl-files? It are still the exact same facets with the open edges. This time sHM recognises they form a closed volume, nicely refines near these faces, and correctly keeps the inside and throws out the outside. Why does it matter whether the surfaces are in one file, or in separate files? On the other hand, this might be a good way to proceed when you have a geometrically well defined volume, but with edgewise-unconnected patches: simply split the file in separate one-patch-per-file stl-files. I hope this triggers some thinking... best regards, Laika, still orbiting laika@inorbit.eu