[instanton]

Hi everyone:


I've modified the externalSolarLoad (eSL) viewFactor tutorial to incorporate a new geometry, having both air on the inside of the solid region (so that the solid region is a shell) and reducing the problem to two dimensions in the typical OpenFOAM way. The mesh is made using gmsh and imported via gmshToFoam. I've attached an image of the mesh.



I now have regions I've called interior, exterior, envelope, and ground. For interior and exterior, I've copied the air files of eSL; for envelope, solid; and for ground, floor. I added a pRefCell and pRefValue to the fvSolution for the interior, as it doesn't have fixed value BCs.


Meshing seems to work just fine, but I am getting an error, seemingly with the radiation model. From the stack trace:

[stack trace]
=============
#1 Foam::sigSegv::sigHandler(int)sh: 1: addr2line: not found
addr2line failed
#2 __kernel_rt_sigreturn
#3 Foam::radiation::solarLoad::updateSkyDiffusiveRadi ation(Foam::HashSet<int, Foam::Hash<int> > const&, Foam::HashSet<int, Foam::Hash<int> > const&)sh: 1: addr2line: not found
addr2line failed
#4 Foam::radiation::solarLoad::calculate()sh: 1: addr2line: not found
addr2line failed
#5 Foam::radiation::viewFactor::calculate()sh: 1: addr2line: not found
addr2line failed
#6 Foam::radiation::radiationModel::correct()sh: 1: addr2line: not found
addr2line failed
#7 ?sh: 1: addr2line: not found
addr2line failed
#8 ?sh: 1: addr2line: not found
addr2line failed
#9 __libc_start_mainsh: 1: addr2line: not found
addr2line failed
#10 ?sh: 1: addr2line: not found
addr2line failed
=============
Segmentation fault


I have no clue how to debug this error. I can post whatever information you need to help with this error. Thank you so much!

[Hawxliquid]

Weird timing, but I actually tried to do almost the exact same thing. I was getting crazy with the error.
After reading your post, the only common point we had was that we did it in 2D. So I took my case, and just made it 3D while retaining everything, and voilà!
It seems the 2D model is an issue here

[instanton]

Originally Posted by Hawxliquid

Weird timing, but I actually tried to do almost the exact same thing. I was getting crazy with the error.
After reading your post, the only common point we had was that we did it in 2D. So I took my case, and just made it 3D while retaining everything, and voilà!
It seems the 2D model is an issue here

Ah, I had considered that this was the issue, but hadn't worked up the nerve to test it! Thank you! I suppose the fundamental issue must be the use of empty patches, wouldn't you think?



Now, the modeling question is: Is there a way to simulate in 2d with a different set of boundary conditions, or is it just kind of meaningless to pose a solar load model in two dimensions, and three dimensions are really required? I had hoped to run my model in 2d first before taking on the greater computational burden... this is a question I'll have to think about, but maybe you have some thoughts.


Cheers.