Comments below the NASA Armstrong Flight Research Center video Proving Prandtl- With A Twist! incclude:

  • This video is entirely a hidden jewel in youtube, it deserves much more attention than what it has had until now. It is both educational and inspirational!

  • It's a clever, non-obvious idea, using the washout to eliminate adverse yaw...

What, and where is the Horizontal wiglet discussed in the video, or in the Horten design discussed there as well?

When I look at the various images in the video, I just see a flat wing as far as the shape is concerned; what distinguishes the winglet from the rest of the wing?

edit: I see a break in the wing near the end, but as far as I can see, the shape or orientation doesn't change from what the wing would look like anyway. What is it that makes the end of this wing a winglet?

enter image description here

  • 4
    $\begingroup$ I just watched the second video. This NASA academy brazenly claims to have newly discovered what we have known for 80 years (well, if we looked in the right places, that is). Another instance of NASA marketing overselling trivial "discoveries". And, of course, the same falsehood about tip vortices creating induced drag is repeated again. $\endgroup$ – Peter Kämpf Jan 2 at 17:56

You can't from the planform alone.

First, winglets are no magical device. Calling the wing with bell-shaped lift distribution one with horizontal winglets tries to free-ride on the mystique that NASA marketing has created around the winglet. But the physics behind it are rather mundane and the thrust which is created by the outer wing is a bit of payback for the higher losses at mid-span from a steep lift gradient over span.

Next, all wings create thrust if you define it narrowly enough. This comes from the suction force on the forward upper side of the airfoil and is called leading edge thrust.

Induced drag is the backward tilt of the aerodynamic forces and is caused by lift creation. The least amount of drag for a given amount of lift and wingspan can be achieved with the elliptic distribution over span. The bell-shaped distribution creates more drag for the same lift and wingspan, since it has higher spanwise lift gradients at mid-span.

What is it that makes the end of this wing a winglet?

This is a matter of definition. The so-called winglet area is where the wingtips carry only little positive or even negative lift. Like in a winglet, this gives the local lift a forward component which works like the opposite of induced drag. Call it induced thrust, if you want: This is what is common to winglets and the negatively loaded wingtip. That in turn is caused by wing twist and local control surface deflection. You cannot see from the top view how lift is distributed over span.

But the bell-shaped lift distribution has some interesting advantages:

  • Since most lift is created near the wing root, the spar bending moment can be kept low for a given amount of lift. This allows for a lightweight wing structure and is especially important for large aircraft.
  • With aileron deflection, the lift distribution on the up-going wing becomes nearly elliptical while the one on the down-going wing becomes even worse, increasing induced drag there significantly. This reduces adverse yaw such that no vertical tail is needed.

Sounds great, doesn't it?

Actually no, it doesn't when you take a closer look:

  • Due to the low maximum lift coefficient of flying wings, the wing surface of a flying wing needs to be much higher than that of a conventional configuration of the same landing speed where a tail surface allows the use of powerful trailing edge flaps, raising wing weight and drag substantially.
  • The bell-shaped lift distribution is like flying all the time with spoilers half deployed. Aileron deflection retracts the spoiler on the up-going wing and extends it fully on the down-going wing. Kind of like the split ailerons of the B-2. I think it is better to only use spoilers during manoeuvring. Also, the Horten flying wings were all known for marginal directional stability, especially at high speed when sweep did not help much. It was too little to even compensate for unsymmetric thrust. A fin or added artificial stabilization would be highly advisable.
  • $\begingroup$ This is more than I bargained for, but I will hunker down and try to understand it all now. Thank you for taking the time to include so much into one post! $\endgroup$ – uhoh Jan 1 at 14:19
  • $\begingroup$ So what is your view on that they just lengthened the span and called this a "horizontal winglet"? $\endgroup$ – jjack Jan 1 at 16:36
  • $\begingroup$ @jjack: No, also the right twist is needed, correctly called washout. That's why they make the pun on "Prandtl with a twist". $\endgroup$ – Peter Kämpf Jan 1 at 16:45
  • $\begingroup$ And you say that winglets generate thrust? What about the drag that they also generate? Do we have a net thrust? $\endgroup$ – jjack Jan 1 at 16:47
  • $\begingroup$ @PeterKämpf I get that "Prandtl with a twist". $\endgroup$ – jjack Jan 1 at 16:48

I have several posts and snarky comments on here that describe the function of winglets precisely as outlined in this video; that is, they exploit the circulation around the tip to generate thrust (like sails on a boat, which is why they were originally called "tip sails"). Almost all descriptions talk vaguely about how they give a reduction in induced drag. This is the first time in a long time I've seen it explained so clearly, and is great to see.

Anyway, remembering that a winglet is a flying surface that generates thrust from tip circulation, what they have done here is simply a flat tip extension with its incidence set (more nose down than a normal wing tip) to exploit the same circulation, kind of earlier in the circular movement of the flow (at 9 o'clock instead of 12 you might say). This placement seems to generate a much stronger thrust component from the vortice than a vertical winglet, so strong that it is enough to completely cancel out the increased drag from the nearby down aileron.

This means that the elimination of adverse yaw this way, along with the sweep back that provides a natural weathervaning tendency, allows you to completely do away with rudders.

Coping with asymmetric engine thrust is another job of rudders not addressed here and a multi-engine aircraft would still need some kind of asymmetric thrust compensation device, but besides that, it seems brilliant.

  • 1
    $\begingroup$ Thanks for your answer! I'm not very familliar with winglets; all I see is a wing in the image. What makes the winglet different from just the end of the wing? What delineates the end of the wing and the beginning of the winglet? Would it be articulated forward or backward when in flight? $\endgroup$ – uhoh Dec 3 '18 at 12:56
  • 1
    $\begingroup$ Is it possible then to add a short explanation to "how can it be recognized" in the context of this test glider? I just see a wing with a gap in it, but I don't understand what makes the end into a winglet, and not just the end of the wing. $\endgroup$ – uhoh Dec 7 '18 at 3:06

Looking at the picture of the glider, the orientation of the hinges indicates that the winglets should provide directional control, replacing the rudder as well as the ailerons. The original Horten aircraft had spoilers to perform the same role as the rudder in a conventional design.

  • 1
    $\begingroup$ Thanks, but I still can not visualize the difference between "flat winglets" as discussed in the video say between04:30 and 05:45, and a wing of the same length without "flat winglets". $\endgroup$ – uhoh Dec 31 '18 at 12:37

Great work from the NASA team and an interesting way of thinking in 3 dimensions.

A look at bird wing anatomy shows how they decrease lift and increase drag on the same side: by using their "wrist" to pivot their wing tip leading edge down. You can do this by sticking your arm out and rolling your wrist. Conventional aircraft need rudders to counteract "adverse yaw" created by downward pointing aileron (higher AOA, higher lift) on opposite side of turn. This is the "coordinated" turn. A spoiler on the same side is more "bird like" and can be found on the venerable B52.

But before we go throwing away our vertical stabilizers and rudders it is very important to study cross wind performance. A dihedral aircraft with roll away from the wind and get blown sideways from a strong gust. A "weathervane" motion of the tail speeds up the leeward wing, helping mitigate the roll. Birds mitigate cross wind roll by anhedralling their wing tips (again the wrist).

This relationship of aircraft dihedral and vertical stabilizers is expressed in discussions of "Dutch roll" (vertical stabilizer too small) and spiral instability (too large). This may be why the B52 reduced, but did not eliminate theirs.enter image description here


Maybe they have tried winglets as part of their design iteration and then decided to leave them off, or rather bend them downward thus obtaining a bigger wingspan.

"Horizontal winglet" seems to be a term used in windtunnel studies to indicate "the winglets are bent down" as opposed to 60 degree, etc. winglets.

The difficulty with these "horizontal winglets" is that essentially you have a different wing from the one you started out with, one with a longer span. This is sort of like cheating on yourself.

So, there are no winglets in the given design. They just increased the wingspan and call this "horizontal winglets", which is a misnomer.

And I also think their explanation with regards to thrust being created by the winglets is wrong.

  • $\begingroup$ I think you appreciate my confusion better than most, thank you! If you get a chance and can find a link or an image that helps me to visualize what you are explaining, that will be greatly appreciated. Thanks! $\endgroup$ – uhoh Jan 1 at 0:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.