Winch Systems for Kite Sailing.

 Using mechanical winches to control kites for kite sailing might seem like a simple enough proposition, but it's not.

A fundamental problem is that lines are tension-only elements; you can pull them in, but you can't push them out.  This matters because very often when kite flying, one or all of the kite's lines will sag, go completely slack.  Experienced kite fliers using handles or bars can usually cope with this by anticipating and avoiding its onset from line tension changes and other indicators, they can pull back very quickly, and they instinctively compensate for the control effects of slack lines - for example, by pulling in more than they would on a tight line for the same response. They are assisted in this by being able to feel how much pull each line is subject to from second to second.  Also, for kitesurfers, lines don't hang loose very often because when line tension decreases, they sink into the water, ceasing to move towards the kite, which automatically increases the kite's apparent wind speed and re-tensions the lines. Unfortunately, kite sailing boats don't behave this way, they over-run their kites like crazy given any chance, and if their kite is connected via a winch system, control will be sluggish and insensitive.

And, when lines between a boat and its kite are loose, sooner or later, inevitably, a line gets wrapped around something on the deck (bad) or underneath the boat around the rudder etc (worse) or around something that isn't part of the boat- another boat for example (worst). 

Kitesailing will be regulated out of existence before getting properly started if this isn't solved before those who are not believers notice it.   

However, external line snags are not the only problem that comes from lines occasionally hanging loose. If a line is wound even relatively loosely onto a drum and then the line tension (when pulling in or letting out) increases substantially for any reason, the tighter line will catch in under the looser turns and snag- often terminally, with complete loss of kite control.  And, if the line between the let-out pulleys and the drums is loose when a drum is letting out, that line WILL snarl and jam.

Using large diameter winch drums with fewer turns helps but doesn't solve these problems entirely. Using stiffer lines, placing them in tubes, and eliminating as many snag points as possible also helps, but a more comprehensive approach is to make powered (mechanically or electrically) let-out pulleys that rotate faster than the drum when letting out, slower when pulling in so as to keep the lines tight within the winch system.  Another approach would be to put each line in a tube between the drum and its pulley(s), place a seal around the line at the drum end, and pump water into the tube.  I did a test of this and found that a few litres of water at just 2 bar (from a bilge pump for example) is sufficient to quietly 'extrude' the line from the tube- keeping some minimum tension on the line at all times.
Another serious problem with winch systems is line length registration. Line length varies significantly with line tension- up to 10%- so as lines are wound on and off adjacent drums, they can get out of register by some metres eventually.  This makes control very difficult because drum position, (counting the turns on each drum) is therefore an unreliable and inconsistent indicator of line length. For hand- held traction kites, the angle of the bar or relative position of the handles is the flier's MOST useful indicator.  With simple winches, this indicator is unavailable-and with indirect flying, many of the other feedbacks that competent fliers use are also unavailable.  Of course, many different systems that register the lines are possible.  Rather than using how many turns are on a particular drum, measurement could be made completely independent of tension - readable marks on the lines, with a counter at each pulley has been proposed, and I used markers in sequential colours at 1m intervals in an early attempt. It should also be possible to get enough register accuracy by separating the winding-in part of the winch from the line holding spool. With this, line always goes on and off the spool with the same tension- so its number of turns are a relatively reliable measure of line length.

Line length register is a solvable problem- but it's not simple. 

And it is certainly possible to build winch systems that address all the above problems, but they will be complex, exponentially, so as the number of lines to be controlled increases from 2 to 3, to 4 and often now 5, and the potential for catastrophic snarl ups can probably never be eliminated. 

Some accompanying pictures are of a simple four-line winch system for a 6m purpose-built kite sailing catamaran in 2002.  It has full control while letting out and pulling in, but does not have any system for retaining tension in the lines between the winch drum and the end pulleys, has no direct way for the flier to sense line tensions, nor any registration correction

This boat was potentially lethal- quite a few people have tried it, none were very keen to go again- because of  the risk of drum snarls, looping between the drums and end pulleys, lack of feel, and very nervous-making loss of control as line lengths lost register when they were pulled in and let out under varying tension.

However, for kite sailing with larger boats (when kites cannot be hand flown), some sort of winch system is necessary. And winch systems, for all their challenges, have an advantage over hand flying in allowing line lengths to be varied. Obviously, this facilitates kite launching and retrieval, but it also allows short lines to be used for upwind sailing, when minimising line drag is crucial, and long lines to be used downwind when looping or figure eighting the kite increases it's average pull be a big factor. 

Because of the problems with multiple line systems touched on above, the likely choices for blue water kite sailing are either a single line system with radio-controlled servos on the kite for steering and pull control, or a three line system -the minimum number of lines necessary for direct steering and pull control from the deck.  Obviously, a single line winch is easier to build than a multiple line system, and will be inherently more reliable.  The less lines there are, the less there are to snag and tangle- and a single line system completely eliminates line registration problems.  The disadvantages of single line systems are the requirement for a power source, servos and control systems on the kite. These increase its weight, decrease its performance, are sources of unreliability and will limit flying time if batteries are used and there is no on-board recharging system.

Another factor that favours single line systems over the three-line alternative is line drag.  This causes significant loss of kite L/D (lift to drag ratio, a measure of sailing efficiency), which is most noticeable when sailing upwind- and especially so by comparison to conventional sail boats. If, one day, we can make kites with an L/D of 15 (currently they're generally under 5, against 10 or more for conventional sails), then if this kite flies with just 3 degrees of sag in its line(s), the L/D of the line/kite system will be reduced to 8; misery!

And, for a typical multi-line kite, every line has to be able to take nearly the maximum pull that the kite will provide (because line tensions vary so much).  This redundancy adds even more drag.

A single line kite can fly more than twice as far out for the same total line drag compared to a multi-line kite, and at the same length will have noticeably better performance.
Also, with a single line, multiple looping is possible, causes almost no problems, while with multi-line systems on larger boats, looping is not likely to be practicable because control is progressively lost by line friction as the lines wrap a few times- and there's little enough control sensitivity when using winch systems as it is.

Another factor that affects the single line- three line choice is that multi line systems are not as compatible with autonomous flying (kite auto pilots) which are seen by many in the field as an essential element of blue water kite sailing.  As I see it, kite auto pilots could work through multi-line systems, in theory at least, but would most likely not be satisfactory using only line angle as the input, so would still require sensors on the kite and a radio link- which rather spoils the simplicity of such an approach.  I remain skeptical of the practicality of autonomous flying- perhaps kite auto pilots can work in strong steady winds, but I doubt they can be developed to work reliably in the light shifty and turbulent winds that have been what I've experienced during 90% of the thousand or so hours of kitesailing I've done so far.

Will single line or three line systems prevail for kite sailing then?

The single line system appears to have the advantage- but problems of making reliable on-board steering and de-power systems for kites that don't cause unacceptable loss of performance or require frequent battery changes remain formidable.

So, it's too close to call as yet- or perhaps both- or maybe something completely different, like using a kite power generation system to run an electric motor driving a propeller.  .

 

                               Peter Lynn, Pasir Gudang, Malaysia, Feb '09