[hiaselhans]

Hi,


I want to run a k-omega-sst case similar to the "motorbike tutorial".
It was successful with a closed patch surface and i was able to retrieve pressure distribution across the patch surface by sampling discrete points to the surface patch.

Now i would like to do the same for an opened patch surface. I encountered that SnappyHexMesh creates the mesh on both sides of the patch as intended, but they tend to intersect into each other. As a result i retrieve funny pictures of inversed pressure coming through the surface.
Also Cp values retrieved from sampling a patchcloud are affected by this and at random.
I tried to dig into tools like extrudeMesh but it does not feel like the right way.


How can i achieve a discrete surface and extract pressure values? Which way to follow?

* CreateBaffles?
* extrudeToRegion?
* Fine-Tune SnappyHM?


Attached are pictures of my case with the "weird" surface. My geometry is a *.obj file with (quad) panels.


thanks! best regards,
Simon

[hiaselhans]

Hi,
I managed to achieve almost the results i need.


What i did was to change my patch to faceType baffle in SHM-Dict:

patchInfo
     {
          type wall;
          inGroups (gliderGroup);
     }
     cellZone glider;
     faceZone glider;
     faceType baffle;
     }

It seems to do the job - almost.
The thing is that i have some "boxes" where the master- and slave-patch get flipped. I checked my *.obj file and all the face normals are pointing in the right direction. after running my case the normals are flipped for some panels and patches are mixed up. How can i tell SHM to set baffles according to face normals?

[schuyler]

I could be wrong, but since both sides are meshed individually, I don't think baffles are correct. Usually a baffles is one group of faces with a slave and master side. You have two groups of faces.

From you first post, I would have suspected that this was just a rendering problem in paraview because the two faces are so close. If you refine the mesh it could get better? Do the results make sense? One way to get around it would be to split the surface into the top and bottom and visualize them independently.

Again, I could be wrong, and I'm not familiar with using sHM to mesh an open surface like that.

Schuyler

[hiaselhans]

Hi Schuyler,

Thanks a lot for your message.
Actually it is just one single face, not an extruded body. What i want to simulate is a paraglider. It is just a very thin patch of boxes with an entry-hole in the front. The result i want to achieve is a pressure distribution of inside and outside pressure.

However, i digged a little in the source of SHM and it seems not too simple to retain the direction of the face.
In the end it's not a too serious problem as all i need is a pressure distribution for the surface, so i can flip the values afterwards in my scripts.

But...
Next troubles i had was that on further refinement of the mesh, the solver went diverging and giving me floating point exceptions. i checked the mesh visually and with checkMesh but couldn't find an obvious reason...

If someone is willing to help me get it running, i can offer Tandem paragliding flights in return
I did a lot of calculations in the past with closed boxes but would really like to see this with opened ones.

simon