detached faces on inclined planes
 

Greetings,

While trying to set up external aerodynamic simulations using snappy
I noticed that my wing surfaces were turning out bumpy.

After playing with simpler geometries a bit, I isolated this very
simple case that exercises such bad behavior: a tilted box inside
a blockMesh-ed wind tunnel. If I don't add boundary layers the
bumps are stuck to the box. Here with three boundary layers you
can see them flying above the surface of the box on the left edge
(also along the front edge, but those are more difficult to spot).

http://www.cfd-online.com/Forums/mem...ngingchads.png

I'm wondering if I'm doing something obviously wrong, or is
this likely to be a bug in snappy?

Best,

--diego

Update: false alarm, I am now fairly certain this is actually a bug in paraview.
When I run the flow, the detached patches don't seem to act like a wall, so the
parsimonious explanation would be that paraview erroneously assigns some of
the internal mesh faces to the patch.

Greetings Diego and welcome to the forum!

Just a quick question/answer - Try following the instructions given here: http://openfoamwiki.net/index.php/FA...is_in_ParaView - if following those instructions you still see these floating triangles, then you might want to check with checkMesh to see if the mesh is sane or not:
Best regards,
Bruno

Thanks for the tips Bruno. You were right of course, the mesh ended up broken
after layer addition.

I have since turned my attention to cfMesh and am getting beautiful boundary layers
this way but the simpleFoam solver tends to blow up with these meshes even with no
boundary layers. If you can offer any clues to why this might be, please do!

Thanks,
Diego

Hi Diego,

Uh... OK, here's a clue: 42

Yes, it's a clue :D
But a better answer is this - Please follow these instructions: http://www.cfd-online.com/Forums/ope...-get-help.html ;)

Best regards,
Bruno

Yup, after some more wrestling with it I finally broke down and uploaded entire case directories to this thread: :)

http://www.cfd-online.com/Forums/ope...tml#post523737