Problems with flaps using thick analysis

[Mike Kelly]

I am trying to analyze the attached model using OpenVSP 3.46.0. I want to produce polars for an alpha sweep with flaps deployed a 0, 30, and 60 degrees. I get smooth results with flaps at 0 degrees but diverging results with the flaps at 30 or 60 degrees. My initial results showed negative values for Cdtot, which I overcame by increasing the Num U parameter on the wing. However, even setting it to 100 did not smooth out the results for Cltot, Cdtot, and Cmtot. Looking at other messages here I tried to boost the wake iterations to 10 and ensured that my wing had sharp trailing edges set. I also isolated just the wing for the analysis but saw no change. I also found it strange that the load distribution curves were different for the left vs. right wing. Putting the wing and tail in a thin set produced excellent results at all flap angles. Since I have multiple airfoils on the wing, I wanted to use a thick analysis on the aerodynamic twist effects if possible. Is there something I can change on the VSPAero settings or on my model design to get a smoother result?

[Mike Kelly]

Actually, flap deflections of 0, 15, and 30 are more realistic. I still have the same issue with the thick analysis though.

[Rob McDonald]

I can try to look into it.

However, having different airfoils spanwise along a wing is no reason to prefer thick instead of thin representations.  Go with thin.

The reasons to use thick are:

 - The blockage effects of a component are of primary significance.

 - You need local information on surface Cp (not just delta Cp across a lifting surface)

 - You have an intersection problem that is improved by using a thick body.

Rob

--
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.
To view this discussion visit https://groups.google.com/d/msgid/openvsp/fc13215f-9ecb-44e0-a6f0-5dfb85f1d230n%40googlegroups.com.

[Mike Kelly]

Thanks, Rob. I'll work with thin wings. I didn't realize that VLM would account for camber and thickness on my wing. I am finding that even at 0 degrees alpha and no control surface deflections the solution does not converge with an all-thick analysis.

[Rob McDonald]

It will account for camber, but not thickness.  You can see the camber in the thin-surface mesh of the wing.

However, for typical wings, the thickness variation across the span doesn't change the flow significantly.

Perhaps if you had a blended wing body with huge / rapid thickness changes that induced a significant spanwise flow (diverted from chordwise), but that would be an unusual situation.

Rob

To view this discussion visit https://groups.google.com/d/msgid/openvsp/5586f175-4642-4bc3-8eb7-454652f06f35n%40googlegroups.com.