BWB_CFD Tutorial OpenVSP geometry generation issue

[stanka...@gmail.com]

Dear SUAVE Team,

I have a problem with the BWB_CFD.py case. I am running SUAVE on Linux and successfully compiled OpenVSP there. I am able to create a wing by myself using simple python commands, so the Python API works well. Hosever, when I am trying to run the BWB case, the generated geometry looks wrong.

One question I had regarding this is:
In vsp_write.py case, you work with Sect_0 which should be invalid for OpenVSP as it is a hiddent one (as Rob McDonald said). If I try to play with the Sect_0, I generate something similar to my SUAVE geometry result. But if I use Sect_1 as the root, my script works.
Please, let me know if there are any tricks with the VSP link and if you hade this problem before?

I have attached the BWB base vsp geometry and my own .py script with which I played with the sections. I have also attached my results.

Script:
import vsp as vsp

span = 7
root_chord = 5
tip_chord  = 2
sweep = -10
dihedral = -5

wing_id = vsp.AddGeom( "WING","" )
vsp.SetGeomName(wing_id, 'test_wing')

vsp.InsertXSec(wing_id,1,2)

vsp.SetDriverGroup( wing_id, 1, vsp.SPAN_WSECT_DRIVER, vsp.ROOTC_WSECT_DRIVER, vsp.TIPC_WSECT_DRIVER )

vsp.SetParmVal(wing_id, "X_Rel_Location", "XForm", 0)              
vsp.SetParmVal(wing_id, "Y_Rel_Location", "XForm", 0)              
vsp.SetParmVal(wing_id, "Z_Rel_Location", "XForm", 0)

vsp.SetParmVal( wing_id,'Root_Chord','XSec_1',root_chord)
vsp.SetParmVal( wing_id,'Tip_Chord','XSec_1',tip_chord)
vsp.SetParmVal( wing_id,'Span','XSec_1',span)
vsp.SetParmVal( wing_id,'Sweep','XSec_1',sweep)
vsp.Update()

vsp.SetParmVal( wing_id,'Tip_Chord','XSec_2',1)
vsp.SetParmVal( wing_id,'Span','XSec_2',2)
vsp.SetParmVal( wing_id,'Sweep','XSec_2',0)
vsp.Update()

vsp.WriteVSPFile("testtestwing.vsp3")
print('done')

 

                                                                                                                                                     SUAVE + VSP Results:

My Results

Thank you,

Stan

[Tim MacDonald]

Hi Stan,

It's possible that the API representation has changed recently. We haven't tested this method with the OpenVSP versions that have been released in recent months. Modifying segment 0 was necessary at the time we implemented these functions, and we still regularly use them with older versions of OpenVSP. Rob has mentioned the possibility of modifying this segment in older posts such as this one.

If you've found something that works for you I'd suggest you go ahead with that and post an issue about this on our Github page so that we can investigate when we have the bandwidth. It's always good to hear feedback about what might no longer be working.

Tim

[ephio...@gmail.com]

Hello, Tim,

Actually, the code should be fine and my question with the sections was wrong. However, on Ubuntu, have you ever experienced that fact that SUAVE generates a model like the black one above while a separate self-made python script uses API correctly?

Stan

[Tim MacDonald]

I don't recall any issues like that. I did notice that we have a note on a missing tip segment, but that shouldn't cause the problem you have in the first image. If it's plausible for you to do I'd suggest seeing what happens if you run with an older version of OpenVSP.

Tim

[ephio...@gmail.com]

I have found the issue. It had nothing to to do with neither SUAVE, nor OpenVSP. The error was in the python itself. Maybe, the installation was wrong, but if I include any other module apart from the vsp, the vsp model starts looking like the one above.

Stan

[Gazi Yavuz]

Hi,

I have the same problem in SUAVE 2.4.0 with OpenVSP 3.25.0. Have you fixed the issue about "missing tip segment"? Is that a bug about such specific tutorial or should the way how SUAVE communicates with VSP change?

Thanks,

Gazi

13 Kasım 2019 Çarşamba tarihinde saat 16:25:42 UTC+3 itibarıyla ephio...@gmail.com şunları yazdı:

[ebot...@gmail.com]

Hi,

I'm not sure how Stan fixed the issue he had.

However, if you look at the tutorial it does now have a tip segment.

-Emilio