Weird Wing Tip Phenomenon

170 views
Skip to first unread message

Oliver Shih

unread,
Oct 4, 2016, 9:04:07 PM10/4/16
to OpenVSP
Hi all,

So I've been running VSPAero with a wing using a customized airfoil designed by one of my teammates. What I would expect from the Load Distribution is that Cl would be approximately continuous throughout the span. However it appears to be discontinuous at the tip at a high AoA (in the simulation we used 15 degrees). Here's the plot out of Load Distribution:


At first we thought it was because of the customized airfoil, so we redid the analysis using a regular NACA four series, and it generates a pretty decent distribution as expected. Then I went on and added a small section to the wing tip that transforms from the customized airfoil to the NACA four series, but the distribution we get out of VSPAero is even weirder:

washout.vsp3

Rob McDonald

unread,
Oct 4, 2016, 10:16:09 PM10/4/16
to ope...@googlegroups.com
Oliver,

I'm not sure how much help this is, but it appears to be related to the camber combined with high alpha.

If you change to a NACA symmetrical airfoil, the behavior goes away.  If you make it a 4412, the behavior is still there.

If you look at the delta Cp plot in the VSPAERO Viewer, you will see a 'hot spot' of activity on the trailing edge of the wingtip.

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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Oliver Shih

unread,
Oct 4, 2016, 10:36:23 PM10/4/16
to ope...@googlegroups.com
Rob,

Thanks for replying so soon! Isn't it night time in the States right now? Hope I haven't been interrupting your sleep or anything.

We thought about it, and introduced a small section on the wing that transitions from the high camber airfoil to a symmetrical NACA four series. The problem is still there, and the spike in Cl no longer appears at a location with high camber, but instead still resides on the tips where the airfoil is symmetrical. In fact the spike doesn't begin at such a high AoA; it can also be observed at relatively low AoA like 5 degrees. As to the hotspot on the trailing edge, I will try to modify the geometry to see if the problem persists. Thanks again for your help!

Oliver

--
You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/XZC4OWQ0iuo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+unsubscribe@googlegroups.com.

Oliver Shih

unread,
Oct 4, 2016, 10:52:20 PM10/4/16
to ope...@googlegroups.com
I found that trimming and capping the tip trailing edge improves the situation by a little, but not much. Also increasing taper ratio worsens the situation, and decreasing taper improves the situation a little as well.

2016-10-05 13:16 GMT+11:00 Rob McDonald <rob.a.m...@gmail.com>:

--
You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/XZC4OWQ0iuo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+unsubscribe@googlegroups.com.

Rob McDonald

unread,
Oct 4, 2016, 11:02:34 PM10/4/16
to ope...@googlegroups.com
I noticed that your airfoil had a very slightly open trailing edge at the root that caused the geometry generation to crash on my Mac.

It is a bug I should chase, but forcing the root airfoil TE to be fully closed helped get past that issue.

Rob

Oliver Shih

unread,
Oct 6, 2016, 4:13:29 AM10/6/16
to ope...@googlegroups.com
I haven't noticed that bit, and it hasn't been causing me problems like crashing; would it be fixed if I round off the root cap?

My teammate has shared our problem with my professor, and he simply said ignore the tip phenomenon for now and cut it off when screenshoting. Other TE treatment I tried in VSP has only been slightly improving the situation but none of them actually fixes the weird spike. 

One thing I haven't tried though is to modify the airfoil geometry. If I have time I will modify the TE bit to see what it does to the problem. In the meantime thanks for everything!

bsa...@terrafugia.com

unread,
Nov 14, 2016, 12:31:41 PM11/14/16
to OpenVSP
Hello Rob,

I am noticing a similar phenomenon when I am analyzing my models as well.  I've tried to simplify and I think the only configuration that manages to give a sensical load distribution is a symmetric wing section.  For instance, if I use a 2314 wing section, hershey-bar planform, no dihedral or anything, I get this semi-span loading:


From this mesh:


Any thoughts on what I may be doing wrong?  I will run AVL and see if it comes up with similar results and post a comparison. 

Thanks,
Bryan



On Tuesday, October 4, 2016 at 11:02:34 PM UTC-4, Rob McDonald wrote:
I noticed that your airfoil had a very slightly open trailing edge at the root that caused the geometry generation to crash on my Mac.

It is a bug I should chase, but forcing the root airfoil TE to be fully closed helped get past that issue.

Rob

On Tue, Oct 4, 2016 at 10:52 PM, Oliver Shih <olive...@gmail.com> wrote:
I found that trimming and capping the tip trailing edge improves the situation by a little, but not much. Also increasing taper ratio worsens the situation, and decreasing taper improves the situation a little as well.
2016-10-05 13:16 GMT+11:00 Rob McDonald <rob.a.m...@gmail.com>:
Oliver,

I'm not sure how much help this is, but it appears to be related to the camber combined with high alpha.

If you change to a NACA symmetrical airfoil, the behavior goes away.  If you make it a 4412, the behavior is still there.

If you look at the delta Cp plot in the VSPAERO Viewer, you will see a 'hot spot' of activity on the trailing edge of the wingtip.

Rob

On Tue, Oct 4, 2016 at 9:04 PM, Oliver Shih <olive...@gmail.com> wrote:
Hi all,

So I've been running VSPAero with a wing using a customized airfoil designed by one of my teammates. What I would expect from the Load Distribution is that Cl would be approximately continuous throughout the span. However it appears to be discontinuous at the tip at a high AoA (in the simulation we used 15 degrees). Here's the plot out of Load Distribution:


At first we thought it was because of the customized airfoil, so we redid the analysis using a regular NACA four series, and it generates a pretty decent distribution as expected. Then I went on and added a small section to the wing tip that transforms from the customized airfoil to the NACA four series, but the distribution we get out of VSPAero is even weirder:

--
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.

--
You received this message because you are subscribed to a topic in the Google Groups "OpenVSP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/openvsp/XZC4OWQ0iuo/unsubscribe.
To unsubscribe from this group and all its topics, send an email to openvsp+u...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
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.

Nick Brake

unread,
Nov 15, 2016, 2:05:19 PM11/15/16
to OpenVSP
I ran a comparison to AVL with a rectangular wing and a NACA 2210 airfoil and this is what I got:


I've also attached my vsp3 and avl files in the *.rar archive attached here if anyone wants to see if they can recreate, hopefully I didn't make any mistakes.  I used AVL v3.35 compiled windows and OpenVSP v3.9.1.



n
Weird Wing Tip Phenomenon.rar
Reply all
Reply to author
Forward
0 new messages