Why? A weak motor?
Without a control pod you’d have to overcome the delay from the speed of sound in the tethers though, and the mushy control from controlling the kite through somewhat slack tethers?
You could probably test this out now if you have a power kite, and try to fly it with longer and longer tethers and then seeing how much of a delay there is from your input to when the kite responds.
According to Multidisciplinary Optimization of
Drag Power Kites
by Florian Bauer
It seems the altitude isn’t a limiting factor at only a few hundred meters, I think the Control Pod scales well as long as kites continue to get bigger, but if we want to use existing smaller kites, then apparent wind and kite speed are paramount.
1. Control Pod: Limited to Steering
Taking that into consideration the control pod, while innovative, has a significant limitation: it’s primarily designed for steering. Its mechanism, often a belt or similar device, adjusts to change the kite’s direction. However, its Achilles’ heel is glaringly evident: it doesn’t actively manage the tension in the kite line. This means that even if the steering mechanism is adjusted swiftly, the kite’s response might lag. The kite’s reaction is intrinsically tied to the line’s tension. Without a system to actively manage this tension, there’s an inherent delay between the control input and the kite’s movement.
2. Grounded Tether Systems: Mastery over Tension Systems like the OKE take a different approach. By grounding the kite tethers, they harness direct tension to influence the kite’s direction. This method offers several distinct advantages:
I would assume therefore that such a system to manage the tension usually is implemented? You’d need that anyway to optimize the power harvesting and the tension with which the tether is wrapped around the drum.
If the system has that, your Instantaneous Feedback and Navigating Challenges points become moot I think. And then my comment about the speed of sound in the tether and, let’s say something like the control input from the ground station first needing to result in a minimum tether tension, or a minimum tether sag, before it translates into effective control inputs, contradicts your instantaneous feedback point. Your precision in flight paths I also don’t see why that couldn’t be achieved to a similar or better degree with control pods and tether tension management.
This is not to say I don’t like ground control for shorter tethers, before my points and the tether drag become too much of an issue.
I wonder if it would be a good idea to divide the point load that the control pods now are into multiple load points, or a longer load point, resulting in a less of an abrupt bend in the tether.
The necessity for a tension management system is directly linked to the design of the control pod. Even If the pod can adjust the lengths of the lines individually, the tension control is primarily dictated by the link between the pod and the ground tether anchor point (behind the pod).
Skysails uses a large reel capable of drawing in the line to adjust tension, although this requires energy. This approach has several challenges:
The control pod acts as a connection point for both lines. Thus, any reel tension adjustment impacts both lines, leading to decreased efficiency. A ground-based Control Pod could provide a more direct connection, which may reduce the intermediary steps and enhance tension management efficiency.
On top of that, the significant size and weight of the reel and tethers introduce barriers to rapid tension adjustments. The scale of Skysails’ drum is a notable factor when dealing with fast and agile kites. Such kites, with their swift energy cycles, necessitate that the transmission system rapidly alternate between powering the reel and engaging the generator mode. Moreover, the gearbox must handle large power fluctuations, with peaks potentially reaching thirty times the base power. This on top of the first issue, about how the tension compensation affects both of the control pod’s tethers.
In kite systems, the control pod’s primary function is to adjust the direction of the kite. It achieves this by varying the length of the control lines attached to different points on the kite, much like how a paraglider pilot manipulates the brake lines to turn or change the wing’s shape.
However, the control pod does not have a direct mechanism to alter the tension in the main tether that connects the kite to the ground. This tension is critical for several reasons:
In so saying, “programing” kite paths using pulses of tension might not be as possible with a Control Pod system because the lines to the kite are so short… The kite would just wiggle
As far as “the speed of sound” goes, even with a direct RF connection to the Control Pod, tension in the line still dictates steering, one would still have to wait for tension to mount in order to see a corresponding change in course, so this actually adds a step.
Well the devil is in the details. The tension at the kite and the ground is not the same. This is due to tether mass and bending and aero drag on the tether.
Another complicating factor is that tension can’t easily be measured for a reeling tether. You need to introduce a bend assisted by a pulley, force measurements on the whole reel or some fancy acoustic or video device to measure it. Either way it will not be super accurate.
The other option is to measure at the kite but this requires telemetry (radio probably) and electricity to work. So you may as well have the pod.
Some sluggishness anyhow is to be expected. I think the only way to get rid of sluggishness is a ground video camera to film the orientation of the kite. Anyhow the actuator (tension control of the tether) must have seriously high bandwidth to get good stability margin and agile steering.
The main problem with a pod I think is to manage it swinging about. Though I have little practical experience to show for.
You state that a pod has no tension control. This could be added. If its not there, its probably just not useful enough. It could be as simple as monitoring the current in the steering servo motor. Almost zero cost that is…
Tension Control in Kite Systems and Its Impacts:
What the Control Pod Can and Can’t Do: Right now, the Control Pod doesn’t change the tension in the main tether. It’s designed to turn the steering lines via belt, not to handle tension or reel line in any capacity. If we wanted it to control the tension, we’d need a complete overhaul in design. But that would make it heavier (as your reel would be aloft), and the kite would need to carry that extra weight.
How Changes in the Control Pod Affect the Kite: Since the Control Pod doesn’t handle tension directly, if you make changes to it, those changes have to move through the lines before you see a difference in how the kite flies. Ground based systems do not have the intermediary steps and manage tension on multiple lines much better than a control pod (that must remain lightweight).
Measuring Tension Isn’t That Hard: We can measure the tension in the lines without too much trouble. We place sensors on the reels or control tethers, and they’ll tell you how much tension is in each line. The actual units of tension isn’t a concern. The relative difference between the control line tensions is what we are concerned about. (You can fly a kite blind as long as you can feel the difference in tension in the control lines. Kite surfers do not need to physically watch the kite; they monitor the difference between the line tensions) What’s important is whether one line has more tension than the other. If they have the same tension, the kite flies straight. If one line has more tension, the kite turns. Adding a camera would help in controlling the kite too. (I prefer to keep telemetry and associated equipment ground based to reduce total inertia aloft) A machine learning algorithm can help optimize and calibrate with enough data.
The Problem with the Control Pod Moving Around: The Control Pod swings introduce parasitic inertia to the overall system. It also means the kite cant fly as high because of the associated weight. This subsequently curtails size of the available wind window.
How the Control Pod Controls the Kite: The Control Pod is what makes the kite turn. But because it’s up in the air, any changes it makes need to move down the line to the anchor point (a gradual build up in tension) This is different from systems on the ground, where the kite controls are directly connected to the ground.
Issues with the Control Pod Design: If Skysails wanted the Control Pod to reel control lines, they would need at least two independent reels incorporated into the design. This means clutches and transmissions, x2 the battery capacity, subsequent code, etc. That’s a fundamental change to the design. The control pod is designed to roll a pulley. This is the metaphorical difference between rolling a car and lifting a car. The pod is not designed to reel multiple lines. To do so would imply significant extra weight. Suffice it to say, this would require a complete redesign. And even if that was done, the Control Pod is still wont manage ground tether tension, resulting in similar issues as previously stated.
The Control Pod system is a good idea, and it can scale with increasingly bigger kites; but, if we want to use smaller kites, increase performance, maximize apparent wind and power output, we might need to think about controlling the tension in each line separately for a more agile, performance based kite
I think we are not entirely on the same page here. I did not study the SkySails pod design. But other pods could easily be imagined. Though the Skysails pod is likely an example of what is possible, and increasing functionality a lot may increase the size and weight above that…
Anyhow I think arguing for a new design based on a single existing design is a bit off. Because maybe you should compare to what could be possible, if you had your way in terms of designing a pod.
Research on the Control Pod can be found here:
Im not convinced that practical improvements can be “added” without a new design/overhaul.
All pods would suffer the same ground tether tension issue. That is the nature of a control pod. We’ve outlined several flaws and solutions with the Control Pod, there is no problem with taking another look at how it’s designed. It might be that other AWE systems suffer similar issues.
In addition to that:
I would opt not to use a control pod. Especially a heavier one that compensated for a portion (and not all) of these design issues. If I had to design a pod it would likely reflect the OKE Precision Winch.
A pod doesnt make much sense to me as well because its hard to manage with less tether tension. Also it needs to be strong and lightweight, and that is at odds with design requirements of it being agile and powerful.
But still, ground based steering has its own issues. I know some
AWE companies went there earlier and then moved on. I believe E-Kite (now merged with Kitemill) did this. But of course they did not have OKE tech (no sarcasm intended).
Exactly.
The excerpts below are comments about the behavior of the Control Pod, Robust Automatic Pumping Cycle Operation of Airborne Wind Energy Systems These are hallmark issues of tension control problems we were outlining above. I don’t know if the fundamental issues can be addressed with a Control Pod.
Dylan Eichelhoff (Control engineer) emphasized:
The challenges of managing tether dynamics during specific flight transitions, particularly when shifting between traction and retraction phases. He pointed out that, due to the scale he’s working with, the tether forces are extremely high. These forces spike dramatically when transitioning flight paths. A noticeable issue arises when the tether develops a sag, leading to the dynamics becoming asynchronous. This misalignment, coupled with the tension from the sag straightening, results in particularly pronounced force peaks, raising concerns about the system’s overall robustness and safety.
Dylan Eichelhoff, TU Delft, Skysails
Sebastian Rapp goes on to say,
"Sometimes you have fundamental limitations what a controller can achieve right fundamentally like physical axioms you cannot do better than because your system has certain characteristics so if your system is designed in a way that it’s super hard and super difficult to control then the control engineer will have a very hard time actually finding something that will work."
“Figuring this out is very difficult because you’re doing very nonlinear flights you have very nonlinear dynamics you could for a linear system there are tools to test this but for nonlinear system it’s difficult so the only let’s say approach that systematic approach would be to put this into an optimizer and there are a lot of optimal control toolboxes also for airborne wind energy use your aerodynamic model use your geometry characteristics and then use these optimal control toolboxes to get kind of an idea if this is even possible to fly your system.”
Sebastian Rapp
TU Delft
AWES PhD Defense
This reflects exactly the issue brought up earlier.
Referring to the following message with the video:
@ChristianH , Is there a simple way of ensuring a constant height of the kite loop, for example by cyclically lengthening and shortening one of the two lines, the orbital line or the central line, using the winch, or by balancing the baler which would be off-center?
Any combination of the above can be used to modulate the foil.
The system is designed for redundancy, and would be a good research platform. All of the schemes you mentioned could work to modulate the orbit of the kite.
All below modes are for conducting clockwise orbits where the left line (orbital line) rotates around the tether line
Orbital Line:
Tether Line:
Baler:
Depower Line:
An instructive video about OKE:
I would like find a mean to ensure reel-out and rotate in the same time, for pumping (reeling, yo-yo) mode. Maybe you already had an idea about this @ChristianH .
Hi Doug: the interesting part of this video is from 0:24 to 0:33.
Oh, sorry, I must have fallen asleep for that part! 
I wonder if with the Manual Reel-in Retrieval from Pacific Sky Power (@dantmaui), we could ensure both reel-out power phase in yo-yo mode and rotation. If yes it would be a step for both OKE Precision Winch “Reel and Rotate” Technology and Low radius loop.
For this it would be necessary on the one hand to mix the two systems, on the other hand to find (if possible) the device allowing to maintain an inequality of length between the 2 lines to allow rotation during reel-out power phase.
It looks a little like FlygenKite.
We can see in this video the double groove winch allowing the 2 lines to be wound or unwound at the same time. Besides that the video shows some nice experiments.
Perhaps if one of the two grooves of the winch is of a smaller diameter, the rotation would be carried out while the reel-out power phase is occurring.
The OKE system allowing to prevent tangling would then be adapted. Depending on a favorable eventuality, it would be enough to turn the whole winch including its periphery (and the lines)?
I’m sorry it’s been some time, I’ve been getting some serious coding and hardware done, This is the latest iteration of the project
I love it! impressive work
