Hello all!
I have been using OpenVSP recently to compare results from CFD and VSPAero for an aircraft. In VSPAERO, I have been testing both VLM and panel Method.
The process I have followed as suggested by your ground school is to build slowly the aircraft, first the wing, the wing-fuselage, wing-fuselage-tail and I am now at the point where I am including the control surfaces on the wing (starting for the moment with just one flap element near the root). VLM works fine in all cases. For the panel method however, I've managed to make it work in all the previous cases with the exception of the one with flaps up to the point where I don't know what else I can try to make it work.
The issue I am seeing is that the whole wing has a logical load distribution but the inclusion of the subsurface creates a discontinuity at the spanwise start and end of the flap element (image attached, there is no flap deflection in this run). I'll be posting as well the panel distribution of the wing (notice that the root panels do not follow the same distribution (that is a result of the changes made for the wing-fuselage junction panel method analysis).
I've tried to place the start and end of the flap at the middle of an existing panel, modify chordwise panel distribution. I've run it as well on the recently released version of OpenVSP. I have seen in some posts here that in some cases there is no need to use panel method, but it is my intention to have a model of the aircraft that works for both VLM and panel method.
I would appreciate a lot if you could suggest any advice! I could share the .vsp3 file as well if needed or any other information that you require.
Thank you in advance,
David