Hi DaveS,

“The Sharp-Rotor Arch Kite unit-WECS is closest in previous design to Magenn's AWES. The differences are in the number of "bucket" vanes and the Sharp arrangement as an arch array.”

Not true. The Magenn is a buoyant Savonius rotor. The Sharp Rotor is a variation of the two-sided Donaldson rotor. The difference in the number of “buckets” is critical. A four sided Sharp Rotor does not auto-rotate, so it does not create lift by rotating itself. We don’t know if the Magenn produces lift due to the Kramer effect. A few years ago, I told Magenn about the Sharp Rotor and why they should use 3 surfaces, but they did not reply. However, they seem to have moved toward 3 surfaces. I would like to see their test results, if any, on this design.

http://comcollab.com/featured-magenn.html 

“As usual, the open challenges are in lifting gas economics and operations, and the high parasitic cycle-drag of a cross-axis rotor's upwind-moving side in headwind.”

The Sharp Rotor can be made to be self-launching from the surface of the water by using small pontoons on legs that do not rotate with the rotor. Once one or two rotors are creating lift, they can begin to lift the next rotor, which creates additional lift, and so on, until the entire arch is in the air. So a buoyant gas is not a necessity for a Sharp Rotor Arch Kite to self-launch. But if hydrogen containment can be perfected, then it would make sense to use it. Turning a Sharp Rotor sideways to the wind eliminates lift, and that is why I use that technique for overspeed control for the arch kite. The arch would become a almost a straight line, extending mostly downwind, flying a little above the sea surface. But testing and refinements would be necessary.

I have a variant of the Sharp Cycloturbine that can operate on its side, so it could be used between Sharp Rotors to greatly increase the torque of an Arch Kite. Part of the power from the Sharp Cycloturbines could be used to spin the Sharp Rotors at a spin ratio greater than 1 so as to greatly increase their lift. That is because a Sharp Cycloturbine normally spins at a spin-ratio (tip speed ratio) of 3 to 3.5.

The amount of lift created by the Sharp Rotors of an Arch Kite can be increased by using a wider arch, proportionally, so that the rotors are more horizontal. But if they are being used to create torque, then they need only create enough lift to support themselves, or reasonably close to it. If they are buoyant, then the angle is irrelevant. If the rotors are spun at a spin-ratio greater than one by using Sharp Cycloturbines, then their lift will be much greater, and their angle can be more vertical while still supporting themselves.

“Note the geometric and dynamic similarities with Aerology Lab's SkyBow, which could be held up in lulls by a single aerostat at the center. The SkyBow ends operate as Archimedes' screws, while the Sharp rotor loses rotation off-axis at the ends, unless it adopts handed helical vanes. The Skybow's center tumbles in tight orbits as a DS wing, while the Sharp rotor is a Savonius refinement.”

Please note that the Sharp Rotor Arch Kite orients to the wind automatically, so there is no need for it to operate “off-axis at the ends”. The Sharp Rotor is not a Savonius refinement. It is a Donaldson rotor variation. Turning the Sharp Rotor sideways to the wind can be used for overspeed control.

“Have been unable to find documentation of the "Kramer Effect" yet. Maybe PeterS can help with a definitive link. We have tended to see a Magnus Effect along a continuum from smooth Flettner rotors to textured rotor variants (like the stitches on a curving baseball).”

I have explained the Kramer effect previously. It’s a simple concept, but the aerodynamics can become very complex. It refers to when a wing is rapidly increasing its angle of attack. Stall will be delayed to a much higher angle of attack. In the case of the Sharp Rotor, stall seems to be delayed so much that there is no stall. The evidence for that is that there is no vibration. In contrast, a two-sided Donaldson rotor does stall and vibrates strongly as a result. In fact, as JoeF has shown, the vibration is so strong that it can be used for short-pull pumping.

The Kramer effect is well known to aerodynamicists and it occurs in many different situations, even insect wings. There are thousands of references to it. I did a Google search and found a great many. I searched for “Kramer effect delayed stall”. If you find one that you believe is definitive, please let us know.

As I previously explained, in the case of the Sharp Rotor, the Kramer effect and the Magnus effect smoothly overlap. If the Sharp Rotor is spun using external power so that the spin ratio exceeds 1, then the Magnus effect will begin to increase. But at a spin ratio of 1 -- which is the self-spinning spin ratio of the Sharp Rotor -- the Magnus effect contributes almost no lift. The Sharp Rotor can create extreme lift by using the Magnus effect, but a little more power is required to spin it at a high spin ratio, as compared to a cylinder (Flettner rotor). In contrast, a Donaldson rotor requires a huge increase in power to produce Magnus effect lift because the rotor acts like a centrifugal air pump, so it is not practical to create Magnus effect lift with a Donaldson rotor.

The Sharp Rotor has a unique combination of favorable characteristics, with the result that it can be used to create new kinds of kites. When focusing on similarities, it is important to also focus on the differences, lest they be overlooked. In my experience, neglecting differences leads to accepting false similarities as true similarities.

Hi DaveS,

“The Stretch Kite Flying Wing concept by PeterS is apparently workable, but with known implications. The dependence of the tacking "flying wing" (kiteplane) on rigid spars sets a far lower scaling limit than a hypothetical soft-kite version.”

I’ve shown how soft kites could be used for Stretch Kites.

A flying wing moving directly across the wind (which is not “tacking”; it is “reaching back and forth”), and without dragging a tether across through the wind, can produce maximum power which is far greater than a soft kite moving in an arc across the wind while pulling a tether through the wind. So scaling is just not an important issue.

The tip speed ratio (TSR) of the flying wing could exceed 10 or even 15, and its lift and thrust will be proportional to the square of its TSR. So it could produce the same power as a soft kite a great many times its size. That implies that you may have the scaling problem backwards. It would be quite difficult to construct a soft kite that could be large enough to produce as much power as even a moderately sized flying wing used in this way. The flying wing could be quite light.

“The single anchor-point "turntable" makes a wide separation of the two pilot-lifter kites unstable, compared to spread-anchors for stable ultra-wide separation, unless the turntable is massively huge.”

What you are saying is not true. If the support kites fly higher, the angle between them becomes smaller, while maintaining the same distance between them. So it is not necessary to use an extremely wide separation angle. Consequently, there is no need to fly the kites at an unstable angle. Also, you seem to be assuming that the support kites must be soft kites, but they could be Sharp Rotor kites, and there is, as yet, no evidence that Sharp Rotor kites would be unstable when flying at a large separation angle. The support kites could be flying wings, and I am not aware of any evidence that flying wings would be unstable at a wide separation angle. If you are, then please show me.

“Based on industrial pick-and-place operations with moving "anchors", like crane work. kPower proposes it will usually be more effective, esp for large-scale unit "stretch kite" designs (incl kPower and AWElab's) to belay between spread anchor-points around an anchor field, rather than onlydesigning single-point anchoring for passive self-rotation. Let operational research settle the question of which AWES anchoring approach is more powerful.”

Please don’t create unnecessary confusion by improperly grouping Stretch Kites with other designs based on overly generalized similarities. Please pay attention to the important differences lest you fail to see the advantages of the new concept.

If you have a paper that analyzes in detail why spread anchor points would be superior to one central anchor point, please let me read it. Increased costs, complexity, and space requirements seem to be obvious disadvantages of using spread anchor points, so please address those problems. Thanks.

Hi DaveS,

“If it is not true that Magenn's unit is "closest" to a Sharp Rotor unit, then, of known AWES concepts, what is closer?” 

As I already said, the Magenn is a modified Savonius rotor in its initial conception using many buckets mounted on a cylinder.

http://peswiki.com/directory:magenn 

Savonius rotors can have many blades. (When they have two blades, they create some lift and  have an L/D of 1. A Sharp rotor has an L/D of 2.) When more than 3 buckets are used, as Magenn used initially, the rotor will create almost no aerodynamic lift. So there is no similarity with a Sharp Rotor, other than it being a cylinder-like object rotating around its long axis. Their later design using 3 buckets begins to approximate a Sharp Rotor, which is what I recommended to them, but they may have discovered that on their own. How well their 3-bucket design approximation creates lift is not known. Similarly, I don’t yet know how much torque I can get from a Sharp Rotor by using more concave trailing edges to each of the 3 surfaces of a Sharp Rotor without decreasing the lift.

“Also, if  Sharp Rotors are strung along an arch, are not the end-unit axes angled more upwind, at the catenary tangent angle?”

I think maybe I can now translate what you meant originally. I think what you meant to say above is that the “end-unit” (bottom rotor) “axes” are angled more downwind, not upwind, since the bottom of a vertical object is assumed to be stationary, and the lean angle occurs relative to that bottom point. If that is what you meant to say, then yes, that’s a good point. The lower rotors will all lean away from the wind, and so there will be some reduction in lift in addition to their being mostly vertical. There will also be some down force, which might be a serious problem. But their function is to spread the bottom of the arch.

A potential way to help deal with that reduced lift might be to minimize the lower end disc of each rotor because the wind will, to some extent, act like that end disc by pressing against air trying to flow off the lower end of the rotor. Another approach, to eliminate the down force, might be to add a light tether between the upwind anchor buoy and the top rotors so that the top of the arch tipped toward the wind. And making the arch proportionally wider will increase the lift of each rotor by facing the rotors more into the wind.

“Note that for many a problematic AWES design, its easier to write extensive supportive claims than to throw together a compelling prototype. The best AWES prototype basis might be so simple, easy, and amazing, that writing about it is the bigger job. Your own best ideas all fall along a spectrum of what is easier to prototype than prove by argument alone,”

What? That’s a false choice. The first step is a proposal, which includes a preliminary analysis of a design, such as how a device is intended to work. Then comes experimentation/testing of a first approximation prototype. Then comes an improved analysis based on that new information. Then the process repeats as a design improves. A prototype is no substitute for an analysis. Both are necessary. A prototype without an analysis based on experimentation/testing is not even a completed prototype. That is because a device that “works” may not be working nearly as well as it could, or may not work under other conditions. If you perhaps assume that a working prototype makes everything self-evident, it doesn’t. Without the analysis, it is still just a proof-of-concept model. The analysis is needed to explain “what, when, where, and why, including the limitations on each of those”.

Hi DaveS, 

“Ok, "max Cl varies in proportion to the rate of increase in AoA", which merely means there is a transient lift-force spike, but of course, no net increase in total energy (under conservation-of-energy law).

Therefore, even a Flettner Rotor exhibits the Kramer effect by any sudden acceleration in rotation, just as Kramer himself experimentally imposed a sudden rotation on a flat wing to produce the effect.”

You statement is not a known fact and so should be presented as a hypothesis, not a conclusion. Your hypothesis seems wrong because a Flettner rotor simply does not seem to benefit from the Kramer effect.

You could test your hypothesis easily by rapidly increasing the rpm of a cylinder. If your hypothesis is correct, then you would see a spike in lift which exceeded the lift of the Flettner rotor at its higher, end-point, rpm. But, to my knowledge, no researcher has ever noted such a spike. Maybe they missed it. But I doubt that because there is no mechanism by which a Flettner rotor can generate the Kramer effect. So you could start by showing precisely how the spike could occur in order to clarify your hypothesis. What is the mechanism? I say there is none.

It might be argued that a rotating cylinder can produce the equivalent of the Kramer effect by causing the air to stay attached to the cylinder by rotating the cylinder, which could be seen as the equivalent of a wing rotating (increasing its angle of attack). But if that is true, then there would be no additional mechanism by which a cylinder could once again increase lift by causing the air to stay attached -- because the air is already attached.

It could also be argued that a steadily spinning cylinder does produce the Kramer effect, but it does not do it as efficiently as a Sharp Rotor because the methods of inducing attachment are different.

But how could the Kramer effect and the Magnus effect be differentiated in that case of the steadily spinning cylinder? If there is no way to differentiate, then it becomes an explanation involving the Kramer effect is a complication with nothing gained in understanding. In that case,  the principle of Occam’s Razor applies: the simpler explanation is to be preferred; entities must not be multiplied without necessity.

I have launched many free-flight cylinders (Flettner rotors) by spinning them with rubber bands. If there is a spike in lift due to angular acceleration, I have not observed it. So even if it exists, it must be relatively small.