For wing only, varying incidence and alpha give different results (VSPAERO)

[asdFletcher]

Hi all,

I am very new to Open VSP , so far the program seems great , and is relatively easy to use. All the work is much appreciated!

I was trying a simple test in VSPAERO and getting some odd results , I believe the problem may tie into what was raised in this previous post (https://groups.google.com/forum/#!topic/openvsp/Qx7Ok-QRbs4) but may be a little different:

As a basic test I ran just a wing (rectangular , no twist, no dihedral, NACA4415) at incidence = 40 deg and then the same wing at alpha = 40 deg (no incidence). In theory ( I think ) the results should always be the same , whether right or wrong. However, for wings with "low" (~6) spanwise "U" sections the results were approximately the same, but this did have an affect on wings with "high" (~20) spanwise "U" sections. The prior issue of a CL peak at the tips (see above link to prior post) seems to be what I am experiencing here, and if the spanwise sections are low enough the affect becomes a low portion of the total lift. However, I found it interesting that varying incidence instead of alpha actually reduced the tip CL peaks.

Additionally , the same thing seems to be happening for drag (or CDo): that there is a drag peak at the wing tip which is accentuated when increasing the number of panels, and that varying incidence instead of alpha reduces the peak. Also, this affect seemed to disappear at low wing angles. Please see attached graphs for a better explanation.

OpenVSP version 3.5.1

[asdFletcher]

Hi again guys,

I hate to bump my own post, but ... Has anyone had a chance to take a look at this or know what is going on?  I may be missing something.  I re-read that last post and realized it isn't very clear, so to re-phrase for clarity, the main question is:

Why , when running 2 identical geometries does VSPAERO give different results?

Case 1: wing only alpha = X , incidence = 0
Case 2: wing only alpha = 0 , incidence = X

Kind regards,

asdFletcher

OpenVSP version 3.5.1

[Mark Moore]

You're trying to run vspaero at 40 drag angle of attack.  Why bother, any result is meaningless above 12 or 15 degrees and you're beyond the linear CL alpha region...

--
You received this message because you are subscribed to the Google Groups "OpenVSP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openvsp+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Rob McDonald]

Do you see a substantial difference for smaller angles?

I suspect you've found a small angle issue somewhere in the
formulation. It may be able to be fixed, but you also have to be
aware of what you're doing when you run a potential theory code at
40deg.

The other potential difference is in the farfield behavior of the
wake. The wake is relaxed, but the wake 'ends' continue off to
infinity. I suspect they convect in the freestream direction, but if
they convect in the 'X' direction, that would explain a difference.

You can show the wakes extended back in the viewer, it might be worth
comparing those side-by-side.

Rob

[Liam Collins]

I can confirm that there is still a difference even at small angles.

[Liam Collins]

For example, if you run a basic 1 x 1 ft. section (symmetry on) with a NACA 12 airfoil, first setting root incidence and tip twist to 4 deg. then setting those to zero and setting angle of attack to 4 deg. you get nearly the exact same thing towards the root, but a very different story at the tip.

I've attached model and screenshots. I realize that the tessalation on my section is probably somewhat overkill, but I think it makes it a bit easier to see the problem.

It is true, as a baseline, that a wing with an incidence of X at every section is mathematically equivalent to a wing with an angle of attack of X, right? In which case one or both of these cases is diverging from the accurate solution? I presume the divergent one is the one with the spikes at the tips?

[Liam Collins]

Thanks for the fix in 3.13.2!

[Rob McDonald]

Thanks for the bug report. Glad that worked for you.

Rob