CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Native Meshers: snappyHexMesh and Others (http://www.cfd-online.com/Forums/openfoam-meshing-snappyhexmesh/)
-   -   SHM not snapping (http://www.cfd-online.com/Forums/openfoam-meshing-snappyhexmesh/112029-shm-not-snapping.html)

marupio January 20, 2013 18:50

SHM not snapping
 
2 Attachment(s)
snappyHexMesh is castellating (refining) but it is not snapping to the surface, no matter what I try to do.

When my own geometries didn't work, I tried the pipeBend tutorial suggested on the SwiftSnap page ( http://openfoamwiki.net/index.php/Contrib/SwiftSnap ). The result is shown in the attachment.

I also tried the "cylinderCase" described on this wiki ( https://sites.google.com/site/snappy.../cylinder-case ), and the result is also attached.

This has been tried in OF2.1-x and OF1.6-ext.

To me, this tells me the case setups are correct, but there's something not working on my machine. Perhaps some geometric-based dependency is improperly installed or something... I am working on Linux Mint XFCE.

That is my tale of woe.

Does anyone have any suggestions?

-Dave

alberto January 20, 2013 19:41

2 Attachment(s)
Hi Dave,

I have modified the cylinder case because it did not take advantage of the surface features, and changed a few settings in snappyHexMeshDict (taken from the flange tutorial). Could you please run the case with ./Allrun and see what happens, so we have a common reference? You find a screenshot of the result I obtained attached.

Best,

marupio January 20, 2013 23:06

Thanks Alberto,

I ran your case, and the result did snap. It looks like the image you attached. So that raises the question: what am I doing wrong?

I did a diff between the two case directories, and I think my mistake is exactly what you tweeted me about:
Quote:

Alberto Passalacqua ‏@albertop78 @DavidGaden Did you forget to use surfaceFeatureExtract or to include the emesh? :-)
Mind you, the pipe bend tutorial was using the emesh...

So I will use your case as a template and report back tomorrow with the results!

Thanks again!

-Dave

alberto January 20, 2013 23:19

OK. Let me know :-)

marupio January 22, 2013 21:37

2 Attachment(s)
Thanks for the assistance, Alberto.

So I got your case working, but I still can't seem to get mine working. Maybe you can tell me what I'm doing wrong (if you have the time)?

I've attached the case file, and the blender file is available here for the next few weeks:

https://dl.dropbox.com/u/18048766/Un...Reactor2.blend

The result I get is shown in the attached picture.

alberto January 23, 2013 00:02

1 Attachment(s)
Hi Dave,

I attach a case that seems to work. Major changes:

- added all the stl files to the eMesh creation (there were typos and not removing the log the last 4 files were not generated by the ./Allrun script)

- refined blockMesh box

- changed surfaceFeatureExtract angle to 150

Check the mesh quality to be safe, and in case it's not good enough, I'd refine the blockMesh box.

Best,

marupio January 23, 2013 22:03

Wow! Thanks so much, Alberto! That is an awesome mesh.

It would be embarrassing to admit how long I've been struggling with this. This also was standing in the way of my final simulation for my PhD. (All the other meshes were simple cubes, so it was easy to just use blockMesh.)

As far as I can see, my mistakes were little typos with the feature extract command, and I didn't include all the eMesh files.

Thank-you so much!

-Dave

alberto January 23, 2013 22:28

Quote:

Originally Posted by marupio (Post 403671)
Wow! Thanks so much, Alberto! That is an awesome mesh.

Welcome.

Quote:

As far as I can see, my mistakes were little typos with the feature extract command, and I didn't include all the eMesh files.
Yes, that was it, and some tweaking of the angle.

Best,


All times are GMT -4. The time now is 16:45.