VSPAERO Fuselage Issue

[Hans]

Hi, I use the latest VSPAERO/OpenVSP and trying to analyse the fuselage with wings, but I have faced the problem with the fuselage only. The solver gives the results but it takes time more than usually due to this:

These should be the same value if we assume a sensical mesh... but we problably coarsened too much!
The solver will likely be fine... but you might want check the coarse meshes if this mesh is actually used...
Edge: 1 is global edge: 3 with nodes Node1,2: 1 3 at xyz: 0.000000 0.000000 0.000000
Edge: 2 is global edge: 5 with nodes Node1,2: 3 5 at xyz: 0.000000 0.000000 0.000000
Edge: 3 is global edge: 7 with nodes Node1,2: 5 7 at xyz: 0.000000 0.000000 0.000000
Edge: 4 is global edge: 9 with nodes Node1,2: 7 10 at xyz: 0.000000 0.000000 0.000000                  

The analysis of wing does not yeld in this, only the attempt to analyse separetly the fuselage.

If somebody solved this problem, please, help me too! Or maybe you know what it means, it would be helpful.

PS. The old version does not encounter this problem. And the new file with ONLY fuselage part yelds in this problem too.

Sincerely.

[Rob McDonald]

You'll have to post your file for anyone to be able to help.

Rob

[Hans]

Thanks for the answer! I've attached the .vsp3 and screenshot for more clarity. There is no issue few days ago, it arrised yesterday and I don't know the reasons... For me it looks like the bug, because as I said in the older versions there are no problems. The main problem that the solve time is larger than usually due to this.

Sincerely.

[Rob McDonald]

I don't think anything is wrong.  I think that is just how VSPAERO reacts to using a thin cruciform fuselage with a particularly coarse mesh.

Two things, it is recommended that you bury the wing root inside the fuselage.  Don't leave a gap.  Don't try to match it perfectly.  Bury the wing, let OpenVSP take care of trimming.

Then, use the mixed thick / thin capability.  Set the fuselage to the Thick set and the wing to the Thin set.

Rob

[Brandon Litherland]

Something else to watch out for is VSPAERO intentionally taking additional iterative steps to work on the multi-pole method.  You'll notice for each iteration that VSPAERO works through each level. This does take longer but the solution is typically better and more converged.  VSPAERO is so fast relative to solvers of similar capability that the time difference isn't a large cost.  

[Hans]

Rob and Brandon thanks you all

Got it. These "rows" is dissapeared when the geometry for the Thin Set does not contain the generated vertical panels (for the ZX symmetry option in the analysis). So for the Thick Set analysis there is no such "rows". It's what I noticed, but the analysis of more complex geometry for the Thick Set (a combination of few wings within fuselage with PODs) gives the same rows. I'm not scared by them anymore, thanks. It works how it works, ok.