AMA Expo West 2018 - NASA's Al Bowers: Prandtl Wing Update

This to me looks like a good design for a crosswind kite. It also looks very similar to what @luke shared here: Ultrakite course project so I think he was on to something.

Short introduction:

Longer talk, 2018:

Same talk for different audience, bit more technical info:

Bit from the above video where he talks a bit about propellers, fans, and wind turbines:

Index of videos from same event.


My design was heavily influenced by Prandtl.
However I designed winglets with rudders because I couldn’t quite figure it out.
Note: My awes should also have flown independently, so stabillity was important.

While this is should definitely looked on closer and its effects on awes examined, I don’t think it is quite neccessary for awes at this point in time. Let’s just build awes, get them to market, scale them up and worry about efficiency gains later.

Why I think flying wings are particularly suited to awes:

  • Stability is much less on an issue on kites than on planes. Any change in orientation can be done by winching bridling instead of using control surfaces.
  • Kites don’t need to be aerodynamicallly stable as soon as the line is tensioned. The line is hardly ever untensioned. Never in existing soft kite designs (Kitepower BV). For this short time one could use thrust vectoring, inducing drag or other less efficient means.
  • When controlling pitch and roll with bridles lift vector change of ailerons as well as induced drag is not present. (The effect is minor compared to the rolling of the wings according to wiki)
    The induced drag from lift difference when changing yaw of any backwards tapered, swept or delta wing still needs to be overcome.
  • If nothing is transported, a fusealge is just dead drag.
  • Without any pilots, passengers or cargo and a limited danger zone it’s easier to experiment in awes than in planes.

I don’t know why ampyx power’s plite has such a large fuselage. My guess is, that it’s full of instruments. I also think they tend to just go with a classic design to get something to work earlier. One could argue that their variable multiwing is evidence against that.
The above considerations are alyo the reason why I was quite astonished to see @Rodread incorporate full aeroplanes with fuselage and empennage in his design. On a daisy turbine one can use bridles in more directions to transfer force than on a yoyo, which makes it even better suited to flying wings / blades.

I’m of course no expert. I state things as facts because showing epistemic humility with every statement makes it hard to read.^^

1 Like

Minimum induced drag (2016):

1 Like

Kite/rc model airplane hybrids should be an excellent research platform to try this out.

How about biomimicry? Articulated wing sections, servos, and string? It occurs to me that bird anatomy seems to almost necessitate the design of their wing. No strength to resist bending at wingtip, so no downwash there. Muscles and tendons close to body, and the need to have a large enough moment force to be able to move the wing, so a thick body under the wing, necessitating a large AoA close to the body. Almost no weight at the tip of the wing, so low inertia. Maybe my logic is off.

I’ve only seen a few attempts to mimic how birds fly. Maybe someone has seen something interesting? This topic is a coming together I think of the study of how birds fly and how airplanes fly, so it would be interesting to see something from the birds’ side.

I dont think the lowest drag is so important for AWE, because you will never get rid of tether drag, so you can only do so much with the wing design.

Note that when you are using bridles rather than a tail, control is relative to line angle. This is often what you want, but not always.

The most interesting thing about this video/concept is the ability to turn by rolling, and therefore get rid of the fuselage and tail. This could have benefits in ground handling and weight. Another opportunity is to design a snag-free kite, which could be useful for scenarios where the wing catches a tether.

Though quite interesting, I doubt this will have much impact on AWE for a while. There is a lot of technical risk related to making brand new aircraft/kite designs. Making something fly is not an easy task. I think those who trailblaze AWE will be those who use something that looks a lot like something that is already flying, with as few modifications as possible. 20 years down the line we may see this concept used in AWE. But is’s far from the most pressing issue for AWE designers

1 Like

Please explain that. I don’t understand it yet.

If you make a multi-wing AWE rig, there might be a time when some wings are not flying on taught tethers and perhaps in strong winds. In this scenario, a wing could happen to crash into a tether. If the kite is concave in all directions, there is no way the tether could snag the kite.

This is more probably with a wing plus fuselage design

Another scenario is during a controlled launch or landing. Snagging will make life more difficult


I searched for this and found:

For example:

Wings of gliding birds increase aerodynamic performance by continuously changing shape and size.

Here I’m interested in first principles, not so much what current players are doing. If they are smart they will have different parallel research programs exploring different ideas. This one can be two guys building and flying kites every other Saturday.

The claim is a possible 12.5 percent increase in wing efficiency, 20-30 percent increase in efficiency by eliminating the tail, and 15.4 increase in propulsive efficiency.

How much of the total is the tether drag?

A crosswind kite is always turning. Some Youtube comments on the first video:

Here are some more counterarguments to this wing:

Although looking at nature is in general a food thing thanks to evolution, some structures are just not a good fit. There is no flying creature that relies on a tether. This makes natures potential for good influence a lot less compare to just flying for AWE.

For tight turning radius mesh AWE, yawing left when rolling right is probably a good idea, as you need extra «centrifugal» force for keep your mesh at the correct size. For such designs this doesnt make sense.

I still thing the design idea has some
potential though. Its not good or bad, somewhere in between, depending on lots of other factors

1 Like

Are you referring to a blade in a meshed rotor here? I think by Rolling right, you meant what I’d previously called banking the outer tip closer to the ground root. (that’s quite bio sounding) The banking is set by tethering so it’s not an actuated wing rolling force needed there. The wing can be flexed along the leech if needed. I reckon Twist as in the likes of a windsurfer sail with loose leech is going to be over the top for optimal rotor blades in a mesh. I do like the look of those seal whisker bump wings, those sections could easy give against each other up and down with a wee glide surface for local AoA tuning. (please never talk whale bumps again anyone) Windsurfer sail type twist efficiency could be usefully incorporated into paired wings for a mesh rotor blade. (They are also an under recognised development in line with Prandtl’s works /// again/// I reckon.)
Two masts tied foot to boom clamp make a nice sweep reconfigurable glider.
Getting rotary flight down a dune face from these is quite easy. Throw forward and leap out catching the booms (The landings not so. he he) It’s kinda silly and impossible to ride.

Whether or not I had enough sweep in my recent sketch… Not a clue. The wings were asymmetrically swept. The fuselage was based on a cord of the flightpath.
What Prandtl found was that the L/D will have gone up, and he went on to find the distribution that gives optimum L/D if the span is not constrained but the maximum bending moment is.

We have limits to our bending moments constrained by the amount of weight we are willing to carry to stiffen and control our wing.
Our inner and outer tip are constrained downward in a kite rotor so we can afford a fair bit of stiffened mech.
A fuselage ring through the mid line of our yaw sweeping wing pair can hold a lot of clew tension stiffening and tuning the whole rig.

So there’s a trade-off here: you do get to remove the vertical stabilizer, with its associated drag, but you narrow the efficient speed range of the aircraft. Sailplanes actually need wide speed ranges, so despite eliminating the drag of the vertical stabilizer, I doubt that this is going to be a winner in a sailplane application, even in the Open class (unlimited span).

Best hod on to the tails while we’re learing

“Build NASA’s PRANDTL Flying Wing”:


I knew about the details being in the papers. Haven’t looked at the Igs files yet but they should be more convenient.
It’s for a 3.75m plane. While that isn’t large for an inflatable kite, building something rigid of that size is a lot more challenging.
No idea, how one would go about scaling it.

Dave Santos posted:

On again commenting on New Forum discussion from Old Forum knowledge:

Let flying wings be defined as pure wing with no tail or fuselage; in powered flight originating with Dunne 1910. Such wings are lowest theoretic drag but poorly maneuverable. To execute a turn requires two steps- Roll (slowly), then Pitch. When the wing is tilted sideways, no fuselage is present to act as default horizontal wing surface for Sideslip capability.

Kites at the edge of the kite window depend on vertical surface to tilt into horizontal surface. Without such lift, they fall down. Such vertical lift surface is called “Keel”, and adds some drag, but is as necessary as the dagger-board on a sailing dingy. High turning rate is also essential for competitive kite sports and vertical area is designed-in (esp. C-kites).

Many AWES designers fall into the flying-wing trap, like Makani Wing3 era, Enerkite, etc. They end up adding back vertical surfaces, like Makani’s flygen pylons. Prandtl’s 1933 work did not earn him naming rights to flying wings. He was a Nazi scientist who worked alongside Betz, who also somehow usurped proper naming priority (from Lanchester, for the 16/27 ratio).


Adding that a kite wing’s “keel” area needs to be balanced just forward of CG. Just adding winglets to a flying wing makes crashing happen even easier, by added aerodynamic imbalance. If the wing is high enough or suspended, or the turns very big, less vertical surface will suffice. The ultimate goal is a balanced turn behavior, just like most aircraft.

Forward vertical surface also can create a tricycle landing gear structure with a high flare/stall angle, and also act as a “chicken-stick” to promote STOL/VTOL.

To the extent these standing claims on the AWES Forum are original and valid, they are part of the Open-AWE_IP-CLoud

I still think flying wings have their place. Especially in constantly tensioned systems where they are employed e.g. by @kitepower , kps, Enerkite, skysails and airseas as well as common kitesurfing.

There are more ways to turn a kite. To generate a moment, you need a force and an arm. A tail gives you the arm, and the wing on the tail gives you a large force with little drag.

Another way to turn a kite would be to add propulsion or drag somewhere on each wing. This may be easier than you’d initially think, as using propellers to harvest energy is a feature many rigid wing awe will have anyways.

Also, you are perhaps missing the boat a bit here. The Prandtl (?) wind was designes to have «proverse» yaw meaning that it would yaw by itself when it was rolled. This happens due to change in L/D somehow, I will not go too deep into the details where I’m no expert. Anyways, it seems a tail is not necessary.

The other reason to have the tail is of course pitch stability. This is a more challenging problem imho. The tether cannot always be used to find the optimum angle-of-attack I believe

1 Like

It doesn’t have to find, but only control the aoa, which can be done actively like on the kitepower system.

Thats true. But then you also need to meaure AoA and sideslip. I am out of my comfort zone here, but it seems this problem is difficult for normal airplanes, in particular for a real-life situation with snow and ice

1 Like

This guy Bowers is worrying me just a few minutes into the first video of him, above.
He’s still using the discredited Bernoulli theory of lift, and complaining about his paper not being published…

Wow it ends at El Mirage dry lake - heck we were there Saturday taking a ride in a friend’s motorized hang-glider.