vspAERO: alpha_max way too high

[lena.flinga]

Hi everyone,

I'm a newbie with openVSP and as a start I tried getting some aerodynamic data for a regular wing (span 1m, no sweep, tip chord = root chord = 1) with a NACA 0010 profile. However, in vspAERO, I'm getting very strange C_L over alpha curves. The C_L curve only seems to gradually decrease around 30-35 degrees alpha. This seems very high, for a NACA0010, shouldn't it be somewhere around 15 degrees? I must have some faulty settings, but I can't figure out what it is. Can any of you help me out?

I have attached some screenshots of the geometry, vspaero-settings and my results.

On the screenshots I have used the Panel Method, I'm getting similar results for VLM though. I have deliberately let it calculate up to 90 degrees to show how weird the results are.

Thanks so much in advance,

Lena

[lena.flinga]

One more thing: I figured it might have to do with the low aspect ratio, that's why I tried just adding a wing in openVSP, not changing any geometry and then tried it again. The curve is a lot smoother but alpha_max is still way too high (see screenshots attached)

Thanks for the help!

[cibinj...@gmail.com]

Both the vortex lattice method and the panel method are potential flow based methods. They provide linear flow solutions and cannot model stall (which is a viscous phenomenon).

Also, if you're comparing against 2d airfoil results, keep the wing aspect ratio as high as possible (typically around 20 or more) to avoid finite wing effects.

[lena.flinga]

oh, so I can't figure out where my wing will start to stall at all?

[cibinj...@gmail.com]

With just a potential flow based method, no you can't.

However, most solvers get around this using lookup tables, correction factors or limiters.

VSP Aero has a CL limiter implemented, I believe.

In VSPAero > Advanced > Other, there's a 'CLmax' field available. If you input the maximum airfoil CL (generally the CL at stall angle), it limits the CL of the wing sections to that value.

This should give an approximate idea of the point at which the wing stalls. As with any viscous [3D] phenomenon, it has limitations when complex wing planforms or low-aspect ratio wings are used. You will not be able to model post-stall lift coefficients either.

[cibinj...@gmail.com]

This should give an approximate idea of the point at which the wing stalls. 

... by 'point' I mean the angle of attack. Not the point of separation on the wing surface.

It's not possible to predict the point of separation over the wing using these methods. 

[lena.flinga]

perfect, thanks so much!

[zach....@fhengineering.com]

Potential flow methods cannot model stall, but they will report local lift coefficients required by the airfoils to *not* stall.  So if the VLM or panel code shows you need a local (2D) cl of 1.2 at some span station on your wing/surface, and you go look at your XFOIL results for that airfoil and see the viscous 2D cl_max is a cl of 1.6, then you have some margin from stall at that wing span location.  This margin will vary along the wing based on its twist, and the VLM solution should be close when margin is sufficient.  XFOIL is typically optimistic compared to wind tunnel and flight test results I have been privy to, so some engineering judgement comes in when determining how much stall margin to carry on a given 3D wing design referencing 2D XFOIL stall predictions.

VLM and Panel methods can be very useful for predicting stall AOA and stall progression along the wing with the right approach and judgement.  NS CFD models often have a really hard time with modeling stall, so simple tools + experience with the airfoils is often the way to go.  YMMV!  Good luck!