How a sky serpent shaft can scale up?

I would think since this is slightly borderline infeasible to begin with it doesnt make sense to test anything other than carbon fiber in well engineered dimensions and weave layup. Better to start with some calculations (like I did), then prototype something that might work.

I also did some tests with spinning tethers and swivels, seeing like you did that the amount of torsion propagated in a string is quite small

I suggested using this for @Rodread ‘s stuff but he advised against it, so probably a bad idea, but still here it is, again:

This stuff could act as a torque transferring tether if drag is of lesser concern. You could add fringes or a tail for lesser drag, and get the most advanced tether design ever (in a bad way) :slight_smile:

1 Like

Yes, and above all when with the tether is tensioned.

This sounds really dangerous. Why do you need to twist the shaft?

I just experimented torque transmission with a rubber rope. As a result it is the same as previously with a non elastic rope: when it is slack, the rotation is easy; and when it is tight, the rotation is difficult.

Tallak
I would appreciate if you could do a ballpark scale-up analysis of the multi-turbine system I propose. Unlike the Sky Serpent, this system has no cosine losses or shadowing effects. The cable drive enables operation at higher altitudes with stronger and more consistent winds which increases power and increases lifter kite pulling force. How do belt drives and cable drives scale?
DOUBLE TRAIN BELT DRIVE.pdf (72.2 KB)
DOUBLE TRAIN.pdf (84.2 KB)

Hi… Looking at the “double train” figures.

My initial thought is that there is no lifting force shown. Depending on the altitude you want to attain here, a lifting force is required more or less of equal force as the downwind pull of the blades/turbines if you want the general elevation angle to be 45 degrees. As all the blades are positioned directly downwind, they will not provide lift. I guess you could possibly tilt the props upwards slightly for a gyrocopter effect. Nevertheless, the lifting force required is pretty massive. But of course, thats just a matter of adding a large enough lifting surface above these.

Otherwise, the design seems sound to me, limited by practical issues concerning weight mostly. I think one would need to calculate the exact weight expected to compare the cost of lifting that relative to the power output. I am afraid that a regular HAWT could compare positively in such a “LCOE” kind of analysis.

I think there are some mechanical issues regarding the cable drives. One issue is how to get the amount of tension necessary to make the cable drive not slip. This means having spacers to keep either end of a cable drive apart. This essentially adds weight but only to counteract a force that you created yourself.

I think also of great importance is the material and design of the cable of the cable drive itself. It must be lightweight and durable, and not easily slip. So I expect regular UHWMPE (dyneema) may not be most suitable. Not my area of expertice this, still I am slightly concerned.

The cable drives incur each a loss of power. Look eg at this page:

It’s difficult to say how much the loss is, but I see that the upper turbines are connected to the lower level, which again is connected to a lower layer. For each pulley, you will certainly see some losses. If the loss is 2%, if you have 4 layers of pulleys, almost 8% is lost in the rope drives. I think more than 2% loss is pretty realistic in each cable drive to be honest.

To conclude, I don’t right now believe that you could build this in an economical manner, considering losses and lifting force required. You could perhaps convince me by supplying numbers that look more positive than what I am guessing at this point.

If I was to suggest a possible improvement, perhaps only have one rope drive (or none) and then make all the turbines move synchronized connected through cables connecting the generators electrically rather than mechanically. Finish the design by having a motor run the rope drive towards ground, or just transfer the power by cable to the ground. I’m not sure if this approach would be better though, as all this equipment is also rather heavy.

windmill.pdf (6.6 KB)

Showing my concept sketch ans a possible evolution of yours. A large frame with many smaller blades/generators, all connected together through wires, which are further just fed to the ground for extraction.

Blockquote
If you consider the multi-turbine system I propose is a modular system, then each additional module added (turbine, 2 pulleys and a belt) will have an efficiency loss of say 2% so the overall efficiency loss is 2%, not multiples of 2%. Would there be a greater efficiency loss if you scale up the turbine and belt drive size?

I originally designed the system with universal joints or bevel gears and found the efficiency to be much worse than belt drives. Also gears and U-joints are much heavier than belt drives. Will belt drives scale better than gears or U-joints?

Blockquote

Blockquote
The Kitewinder system works just fine without slipping. Why would there be a greater tendency to slip if you scale up? All you need is a larger lifter kite. Bear in mind that a large component of the turbine thrust is transmitted to the tether (cable drive).

Kitewinder does not have cable drives in series, just a single cable drive with a 90 degree bend I believe.

Anyways, I think maybe mass scaling of a cable drive may not be positive… that is, the weight of the cable drive to support a scaled up blade.

Your design doesnt need to scale op the blades, rather they may be multiplied. But still, more rotating elements would lead to more layers of cable drives…

Do not confuse cable drive with belt drive. There is only one cable drive connecting the output from the system to the ground gen. The belt drives connect the power from all the turbines together. Each belt drive can have its own belt tensioner so slippage is no problem. The cable drive must have external forces (lifter kite/ turbine thrust) to prevent slippage.

Certainly, but on the other hand the tension of each of the belts will be transferred to the structure which will be weighed down accordingly.

As shown on the video from my old previous comment, the torque transfer was not good with a single rope, because the rotation was difficult and slowed down when the rope was stretched. But perhaps a combination of several ropes (the two messages I quote) could work better, like a solid shaft. Who wants and can test this?

I just saw the video of the medieval fair.
I really like that kids’ ride at the beginning.
The lady is turning the whole thing using just her hands on the central pole.
No motor required.
Looks like great fun for little kids.

Essentially ancient technology but for modern applications.