What is possible with Payne's patent US3987987 figure 5?

I compared the NTS linear track with the rope-drive you sketched (the carousel I sketched being another thing) for two reasons: kite force transfer on a single mobile point, and forth and back motion.

1 Like


I made a rough miniature of this as advised by @Windy_Skies. Here is a picture below:

Then I experimented it (video below) by pulling on the two ends of the lever (as for the initial figure 1b) then on one end (as for OrthoKiteBunch).

The figure 5 is significantly different because there is a linear move of the (power) segment between the two pulleys where the power is taken.

Below is also a miniature of the figure 5 device alone then with steelyards.

Distance between the two pulleys: 0.45 m.
Distance between the vertex (where the kite is, or here the orange steelyard) of the triangle and the point 0 in the middle of the two pulleys on the power segment: 0.45 cm when said vertex faces said point 0.
When the vertex of the triangle covers 0.45 m, the power segment moves linearly by about 0.2 m.

Concerning the measured traction, the orange steelyard located on the vertex indicated a value roughly 2 times the value indicated by the black steelyard located on the power segment. The photo shows where both steelyards are located.

These rough measures could be calculated and refined by using a simulation integrating kinematics then the resulting forces.

1 Like

That did help Pierre. I sometimes just use a loop of string and some extra fingers and tubes. Or maybe a board with some nails sticking out of them.

I misinterpreted this image before. I have now quickly read the patent. Payne knew his stuff to be able to write this in 1975 I think.

Here’s what the patent says about figure 1b:

FIG. 1b shows a first alternative preferred embodiment of the tethering structure shown in FIG. 1a. In this embodiment, the kite 10 is again tethered on cable 20 such that it tacks back and forth across the wind shown by arrow 18. However, rather than employing a succession of pulleys, the single tethering line is coupled to ends of a link member 40 at points 42 and 44. The link member 40 is pivoted about pin 46 such that it tends to oscillate back and forth in a motion directly corresponding to that of the kite. By the use of well-known transfer mechanisms the oscillatory motion of link 40 about pin 46 can be converted into electrical energy or dedicated to other purposes.

I don’t know what I’m talking about, but anyway…

I don’t see the point of this idea, other than for pumping, if I understand it. The kite can move hundreds of meters and here it is just used to wiggle a see-saw a bit? But let’s use it as a starting point. I don’t understand the reason for the see-saw, so let’s delete point 42 and the upper part of link member 40 so that we are left with a single tether that is connected to point 44. Now the tether can rotate link member 40 around 46. Let the boom rotate horizontally instead of vertically, maybe make the boom a bit longer now you can support the end connected with the tether with wheels, add more spokes maybe, maybe make a giant wheel out of it, and you end up with something like that Italian team was doing years ago? That’s no good, too complex and too much material. But it is better than the original idea in that your boom is longer and you can rotate the boom further so your working cycle duration increases and you can do more work. Let’s instead stand the wheel upright again, make it smaller, wider, and put grooves into the rim if tether wear is an issue, so it turns into an oversized drum for your tether to pull on. Now if you can wrap your tether around your 2 meter diameter drum 50 times, you have over 300 meters to work with, instead of maybe 4 meters with the original see-saw idea.

Figure 4 looks like it kind of describes a yoyo system.

Another preferred embodiment for utilizing the motion of a kite that more than supplies its own losses is shown in FIG. 4. The operation of the kite 90 is generally described in American Aircraft Modeller, September, 1972, pages 20 and 21, whereby a motorless control line model is flown in the wind by making the device 90 sweep out a tipped orbit with the high side of the kite upward as for an autogyro rotor. In this mode of flight, power could be collected from a propeller and dynamo in a manner shown in FIG. 1 or, by attaching the tether cable 20 to a large crank 92 pivoted about fixture 94 on the ground.

Superfluous link:The Drag and Shape of Control Lines in Flight

Senior Boeing Engineer, Chris Carlin, was a key mentor to many of us in AWE a decade ago. His long and heroic career was full of major aerospace research, like developing large scale wind turbine reliability, and flying full size airliners as drones, for fiery crash tests.

Perhaps Carlin’s most valuable lesson to AWE R&D was to seek Engineering Similarity Cases to heuristically assess feasibility, and to reuse engineering methods. In AWE, this started with study of giant industrial trawling nets, as mega-kite analogues.

Finding Similarity Cases continues to help us find ways forward in AWE. The latest is a USP3987987fig5 Similarity Case, Reaction Ferries, rigged similarly as a cross-current paravane at MW scale. Reaction Ferries have been around for centuries, and prove the fig5 cable-rig works quite well, at a substantial scale-

Its a close match, but most fig5 rigs are flatter than the original patent artist depicted-


1 Like

Reaction ferry similarity case is not US3987987figure 5 because pulling is on a single moving anchor, a little like the rope-drive of your colleague @Windy_Skies, or rather NTS linear track, or rather the sketch below:
Reaction ferry similarity case

Of course Similarity Cases are always partial matches to a novel concept. Having reviewed many photos of Reaction Ferries, rarer cases use a single tether, but most are bridled directly to a pulley traveling on a cross-line. These are the closest similarity cases to fig5.

All Reaction Ferries have the essential property of tacking paravanes in cross-current motion, just as fig5 is a tacking kite in crosswind motion. Rigging details vary, but the essence is quite similar, and two anchors set crosswise is the common form.

Any cableway gondola moving between two pylons is also a fig5 similarity case. There is no mysterious parasitic limitation on power by static tension. In fact, a higher static cableway tension, up to a point, is the most efficient basis for the dynamic tension of tensile work.

Consider a bicycle chain drive as a fig5 rigging similarity case. Wikipedia-

" Efficiency: A bicycle chain can be very energy efficient: one study reported efficiencies as high as 98.6%.(https://en.wikipedia.org/wiki/Bicycle_chain#cite_note-Spicer-4) The study, performed in a clean laboratory environment, found that efficiency was not greatly affected by the state of lubrication…Higher chain tension was found to be more efficient: “This is actually not in the direction you’d expect, based simply on friction”."

1 Like

Some cases of reaction ferry use several tethers instead of a single tether, both moving on the fixed transverse cable, unlike the device on the figure 5, unlike a bicycle chain, both being not similarity cases in regard to reaction ferry.

Pierre, Sorry if you are unable to see any similarity in these cases. The heuristic reasoning must therefore seem very mysterious. Never give up trying.

kPower can rig fig5 variants by single or multiple lines, and still distinguish key topological constants as a similarity basis for cases like reaction ferries.

I indicated a reaction ferry similarity case. The figure5 is not this as it describes a static force equilibrium. However some flexibility allows a little force transfer from the kite as it moves the vertex horizontally.

Agreed, you see a similarity case with NTS, but not with fig5.

I see both similarity cases, and fig5 is the topic here.

The problem with kite tacking on single line to reverse track direction is the phase delay to cross the static zone. Fig5 topology does not have this problem; it tacks and develops traction immediately.

Here is a video of a system based on Fig. 1a or 5.

As the kite on the video is 2 m² and the wind speed is 5 m/s, assuming a L/D ratio of 4, a rough calculation in pumping (yo-yo) mode would give a mechanical power of about 350 W during reel-out phase, so about 175 W in average, and less in electrical power, but likely more than the given value of 100 W. However said value of 100 W is higher than I expected. And as already mentioned this cannot face all wind directions.

Baptiste also shared info on LinkedIn https://www.linkedin.com/posts/baptiste-labat-01751138_kiteborne-nano-system-test-1711-2018-activity-6663749398582951937-vw2Y

Glad to see, after all these years, someone bothered to try a “crosswind kite-power” system using “a power-kite”.

1 Like

Below is a sketch for a figure 5 system whose the second pulley can move on holes or a rail on only a quarter circle to perfectly adapt to almost all wind directions, and well enough for all.

This is similar in a way to the 3 point anchor, isometric triangle pattern proposed by @kitefreak Dave Santos. Whereby the 2 best suited anchors necessary for Paynes crosswind application would be chosen from the 3 to match to any wind direction within a 60 deg arc

e.g. if you store the operating rig and pulley sets In the space between any 3 of these anchors…
Whichever way the wind is blowing dictates which 2 anchors is the best match for you to deploy and perform side sweep work from

Thank you @Rodread for this drawing that I did not know.

I think my system is far more easy to manage because one (pulley-generator) of the two anchors is always fixed. Only the second anchor (movable pulley) is moving. In the end an automated control could be implemented.

And also a rail on a quarter of circle can save material compared to a complete circle.

1 Like

TBH can’t remember if Dave ever did mention arraying the system… But he was definitely into the 3 point anchor setups. Arraying it may be something I inferred.

In fact I correct my two previous comments as the advantage of the used space for a quarter of circle is not obvious.
Superimposed areas

I think you need 180 degrees rather than 90 degrees, eg to operate in winds from SE on the figure

The drawing shows that almost all wind directions are compatible. There is only two gaps that are indicated by the arrows: SE and NW. If we want assume quite all wind directions, we need 180 degrees instead of 90 degrees, but I think it is not essential.

That said the used area (with 90 degrees) is the same as for a complete circle where the two anchors are equally spaced. So the is no area gain. Perhaps the system is easier to manage as the pulley-generator anchor is always fixed.

So as a circle is required for assuming all wind directions for the figure 5 system, using the same circle for a carousel could be more suitable I think.