Peter Lynn ARCHIVE
Ram air (soft) kites are those built by the Jalbert principle of using pressure captured from flow stagnation points to inflate the kite's internal spaces and using these as the kite's structural elements.
In their pure form they use no other framing or battens and have the great virtues of easy packing and no rigid parts that can cause injury to users or innocent bystanders.
Some single line soft kites are built purely for efficiency and stability with no concessions to other aesthetic considerations except maybe for surface decoration. Examples are the Jalbert Parafoil and the Sutton Flowform. Increasingly though, soft kites are built as theme kites with some specific representational shape. For these, flying efficiency is not the prime requirement. For theme kites, that they look like what they are supposed to represent and that they fly reliably are the prime requirements rather than efficiency as measured by angle of flight or pull-for size. In fact, especially for larger theme type kites, NOT developing much pull is usually desirable.
Soft kites are all relatively new, the first flew in about 1953 and theme style soft kites are even more recent than this, not becoming common until the middle 1980's and after.
This has not been nearly long enough for there to be yet much understanding of how to make such kites fly satisfactorily. As at 2001, designing such theme kites is still very much a matter of trial and error,- and more often error unfortunately.
The following are some suggestions that can help, many derived from the theories of single line kite stability expressed in the first section under the heading "Single Line Kite Stability"( referred to from now as SLKS).
Some examples will be explained by reference to a simple parafoil style kite, called the "Pilot"- because it is much more basic and allows various influences to be isolated- but the principles derived can be applied more generally. "Pilot" plans are generally available.
The starting point for designing new kites is shape selection.
The inviolable principle here, (and the only principle of kite designing that I'm absolutely certain of) is that:
"Not All Kites are Created Equal".
Shapes that are difficult to contrive using only ram air inflation.
*It is difficult to make any shape with long thin appendages using ram air inflation as the sole structural element unless they can trail to rearward. Even when such appendages can be caused to inflate and hold their required position such kites will still be less reliable than they would otherwise be during launching and in unstable winds.
This makes most insect shapes difficult to do because they usually have feelers etc projecting forwards. An exception is some beetles that have rearward sloping feelers.
Another exception, or rather, a way around this problem is to design such kites to fly tail first so that the long thin things can trail to rear rather than project at the front. Jurgen Ebinghaus has a beetle that flies this way.
*Flat surfaces are also difficult to contrive on ram air inflated kites because any pressure difference across a flexible membrane will cause it to become convex, even if it has been cut flat. For example most of the flattish panels on car shaped soft kites (the sides, roof, and bonnet) tend to bulge out grotesquely.
There are some ways to control this bulging.
Firstly, any surface that lies in or near a plane at right angles to the wind direction will be a stagnation area for airflow and hence have near enough no pressure difference across it from outside to inside. This will therefore cause it to conform approximately to whatever shape it is cut to rather than bulging out- but such surfaces generally flap a lot so the effect won't be smooth. Unfortunately also, surfaces that are required to be flat can rarely be placed so as to meet this criteria.
Secondly, it is possible to rig multiple internal cords or very close spaced ribs inside the kite to hold a surface flat or even concave. The problem with this is that unless there are a huge number of such cords or ribs the surface will still either have an obvious quilted or ribbed appearance.
The third is to construct the areas that are required to appear flat so that they are recessed- either by tying them back with cords or ribs or by contriving some variation of a ram air inflated tubular perimeter and then to stretch a second fabric skin across this recessed area and vent it to atmosphere rather than to the ram air pressure so that the outer skin retains it's flat appearance. Because of the tiny margin over atmospheric that ram air inflation allows (except in ferocious winds), there is not enough compressive strength in any ram air structure to stretch such secondary skins really tight- but it is still a useful, though complex, solution.
Ram air inflated shapes that fail the "aesthetic" test.
*Precision shapes can be difficult. The human eye immediately notes even minor asymmetries in common geometric forms. Circles, spheres, squares, straight lines etc have to be perfect or they leave most viewers with a sense of dissatisfaction, of "something not quite right".
An example of this response is the wheels on ram air car shaped kites- because there is a stagnation area along each "tire's" leading edge they always look distorted and wrong.
*Representational forms that we are accustomed to viewing from the top often look unsatisfactory when viewed from underneath. The much belaboured ram-air inflated car is a typical example of this. Car underbodies are the most visible part of car kites but are unfamiliar except to mechanics and roadkill.
Shapes that have stability problems
There are shapes that are more likely or less likely to be stable. However, because kite stability is so complex and poorly understood there aren't many rules about shape/stability that can be stated with confidence.
*From experience to date, long thin ram air inflated forms (when length is, say, more than 5 times diameter) tend to have serious, even terminal, problems with volatile instability because there is usually not enough rigidity in the body to prevent the head developing independent and destructive lateral oscillations. However, I strongly suspect that some new insight, or a few thousand more hours of stuffing around, or getting lucky just once, might make a nonsense of this rule.
*Kites with wide blunt noses tend to have superstability problems (of the fourth type when referring to the companion paper on kite stability). This happens when such a kite gets a bit misaligned with the wind. If the leverage exerted by the drag on what has become the more downwind side of the nose sufficiently exceeds that from the side that is momentarily more upwind, the misalignment will increase rather than correct. Without changing the overall shape there are two things that can be done to minimise this tendency. The first is, for obvious reasons, to make the "wing" tips as thin and drag free as possible. The second is to contrive some sweepback in the tips or "wedging", so that when viewed from above, the kite's widest point is somewhat back from it's nose. This increases tip drag on the side that finds itself more upwind when any disturbance occurs, automatically opposing and correcting the displacement. When parafoils or flowforms exhibit this form of superstability a sure cure is to taper each cell so that they are wider at their trailing edge than across their leading edge. The amount of "wedging required to effect a cure can be very small. For our "pilot" kites (a parafoil) reducing the leading-edge width by just 5% while not changing the trailing edge dimensions was enough to completely eliminate this type of superstability.
*Appendages projecting out laterally anywhere in a kite's "head" or "shoulder" area appear to benefit stability. Analogously to the use of a balancing pole by tight rope artists, there are good theoretical reasons why this should be so. Human form kites (including anthropomorphic examples such as teddy bears) can therefore be very good fliers.
*Kites that are tubular in their main form and bridled to mid-range angles of attack tend to suffer terminally from a form of instability called Rolling instability, as explained in "Single Line Kite Stability).
*Aerodynamically efficient forms tend to be less stable. This might seem paradoxical but is true because lift forces are the driving engine of instability, while drag forces generally assist stability. The higher the ratio of lift to drag (usually written as L/D, the main measure of aerodynamic efficiency for everything from aeroplanes to kites), the less inherently stable a kite will be. This doesn't mean that high performance kites cannot fly reliably. They can, but for such kites, the demands that stability makes on bridling and detail shape are such that there can be little freedom left to allow for aesthetic considerations. Our Ray kites are a good example. These are aerodynamically efficient but have been the very devil to develop. Starting in 1988, I have spent more time trying to understand the characteristics of Rays than the total for all the other soft kites I have designed and developed. Notwithstanding the occasional random success (balanced by some miserable failures), it took me 13 years to get a reasonable understanding of what makes them work and not work.
Yes But:
While it is true that some shapes will inherently look better and fly more reliably than others, skill, luck, endless iterations and persistence beyond reason makes a difference. Don't expect anything to be easy and keep the overweening belief in your own kite designing ability that will surface unbidden if ever you chance on success well-hidden until it is confirmed by repetition. There are many shapes that would seem to be ideal as kites but which nobody has yet succeeded with and others that appear unsuitable which fly well first time. I've always found single line kite designing to be hard going, having rarely had new designs that just "fall up into the sky". Mark Abernathy, the principal designer of our Teddy Bears (which flew perfectly, immediately) was wondering why I think it is so difficult- until his next design that was!
Luck aside, Rolf Zimmerman's Lobster is an example of how careful design and clever thinking can make possible soft kites in shapes that do not at first thought appear to be suitable. This is a very fine example of soft theme kite design.
.
Changing and Tuning for Stability
Because most soft theme kites are, almost by definition, not ideal shapes from a purely performance perspective they tend to exhibit volatile instability- often violently so. Further, they are also usually unable, for visual reasons, to have flares fins or keels to assist stability
As explained in "Single Line Kite Stability" (SLKS) there are three basic cures for volatile instability that are available for kites that are unable to utilise extra lateral area devices for this purpose.
The most obvious is to attach drag devices. These can often be incorporated into the shape without causing visual offence or are already part of the shape that has been chosen anyway.
An example is octopus tentacles. The Octopus was an obvious choice for a theme kite. However, I found that the drag from tentacles alone was not quite sufficient to eliminate volatile instability for the ram air octopus kites we make so at first we added an extra drogue to the central tentacles. Although this was made in the form of a fish so as not to be visually offensive, it was still less than 100% satisfactory because it often snagged on things (like innocent bystanders) when launching and tended to cause the second type of superstability problem as explained in the companion paper. The eventual solution was to construct the suckers on the tentacles so as to form drag buckets- which solved the problems completely without offending visual requirements.
Another answer is to increase the kite's angle of attack by changing the bridles. The reason why this can reduce volatile instability is explained in the companion paper on kite instability. It is usually done by lengthening the top bridles rather than shortening the rear bridles because this also introduces some "banana" or rocker into the shape which can be desirable for reasons also explained in the companion paper. Unfortunately, theme type soft kites commonly start life as not at all naturally stable shapes, so bridling changes by themselves rarely if ever reduces volatile instability sufficiently without also unacceptably increasing the minimum wind speed in which the kite will fly. For larger kites, the increase in line pull that accompanies increasing the kite's angle of attack is also undesirable. Our Gecko kites are a case in point. For small versions, increasing the angle of attack until remaining volatile instability is just controllable by adding a moderate sized drogue works quite well- albeit that these kites when tuned this way do not fly great in light winds and have a lot of pull for their size. For larger versions this approach is not very satisfactory for average kitefliers as the pull can easily exceed 300kgms, hence current attempts to utilise the third approach as below.
The third answer is to remove all rear bridles (or lengthen them until they have only an occasional restraining effect) and change the front bridles so that the kite has a strongly "nose up" aspect curving away to minimal angle of attack from say 1/3rd of the body's length. This effectively makes the front of the body fly as a kite, stabilised by the remainder of the body functioning as a tail.
For this system to work there should be at least some minimum flexibility between the "head" and the "tail" but soft kites are generally naturally flexible enough to exceed this minimum.
One type of soft kite that doesn't have sufficient flexibility to allow the above bridling/stabilising technique to function is our Turtle kites from 10 years or so ago. The widest part of their body is exactly where lateral flexibility would be required - and the front body shape looks most un-turtle like if bent upwards sufficiently to make some useful difference to stability.
The two remaining choices for stabilising these Turtles are: To use the entire body as a lifting surface and increase the angle of attack until remaining volatile instability can be (just) controlled with a large drogue, accepting the ferocious increase in line pull that will result. Or, to reduce the angle of attack to almost zero and use a large drogue on a "Y" attachment to control volatile instability, accepting the very low angle of flight that will then pertain.
I have never succeeded in getting this type of turtle kite to fly satisfactorily at an angle of attack within the desirable 5 degrees to 15 degrees range by any combination of bridling and drogue(s). It becomes possible if weight is added to the drogue as well, but this is generally unsatisfactory because of the narrow wind range that results, and danger to bystanders. It should also be possible if the drogue is replaced by a long tail because tails with their combination of weight, drag and lateral area are more efficient stabilisers than pure drag devices. Unfortunately, a tail sufficiently large and long to stabilise these turtles would be visually intrusive.
Considering carefully in advance what will likely be required to stabilise a new design of kite can avoid a lot of what are basically "selection" problems - that is caused by inherent characteristics of the shape initially chosen or by specifics of the details of that shape as it is built.
By far the most reliable predictive indicator of how suitable a particular shape will be is to judge from similar shape kites. The best example of this is how human form shaped kites of many different specific forms are proliferating. Although not automatically perfect fliers, once a few examples such as Pierre Fabre's "Spaceman", No Limit's "Super Grover" and Jos Valcke's "Clown" showed what is possible, many variants such as our Teddy bear and Cat became obvious. Human form shapes are suitable candidates for soft kites because their head and shoulders can be set up as the lifting surfaces while the body and especially the shoulders, arms and legs function as the "tail".
Inflation Points for Ram Air Kites.
The basic principle for inflation is to create an entry point to internal spaces from, and only from some point where the airflow over the kite is stalled. Anywhere that airflow is stopped, by Bernoulli's theorem, the kinetic energy of the wind will be converted to static pressure, which by Jalbert's inspired discovery is sufficiently greater than the ambient to be useable as the structural element in inflatable kites. This is called ram air inflation to distinguish it from pump inflation which is required for kites that use pressurised bladders structurally.
The points at which airflow is stalled are called stagnation points. A classic example is the centre line of the leading edge of a cylinder placed at right angles to the wind direction. For most kite shapes, there will be many such stagnation points, at all of which exactly the same maximum pressure is available.
It makes no difference whether a stagnation point at the rear of the kite or at the front is chosen for a given wind speed, they all offer the same boost to the ambient atmospheric pressure.
However, like kite shapes, from a practical point of view, not all stagnation points are created equal. For example, the leading edge of a fin or flare is a stagnation point and such have been used as inflation points for soft kites by using two layers of fabric for the fin and venting the gap between them to the kite's interior. The problem with this system is that the volume of pressurised air from such a slit is not great. The actual internal pressure developed in a kite is generally markedly less than stagnation pressure because kites leak (seams especially unless taped) and more so as kites get older. All leakage reduces the internal pressure but if there is a higher volume of replacement pressurised air bleeding in, the loss of internal pressure will be less. The stagnation points from which the highest volume of adequately pressurised air is available will be from locations where there is the greatest area of stalled airflow. Typically, this will be in the middle of blunt noses or of relatively large blunt sections.
This is assuming that it is desirable to have internal spaces at maximum available inflation and this is generally only true when the kite has forward or laterally projecting appendages which will otherwise buckle or appear malformed.
If there are no such appendages, lowering inflation pressure slightly can be of great assistance to stability by creating more drag. An example of this is how soft kites such as our octopus become more stable as they get older. As fabric becomes porous and stitched seams stretch kites lose some inflation and their leading edge indents a bit, creating more drag. In the case of the octopus a small reduction in inflation is beneficial in that stability matters more than efficiency for these kites. Too much reduction and the tentacles tangle however.
It is possible to cause some deliberate reduction in internal pressure by providing some adjustable vent from the kite that can then be used to let some air to continually escape.
For kites that require maximum inflation in some parts but benefit from reduced inflation in other parts to assist stability, it is possible to make separate internal spaces and have them independently inflated and controlled.
Because volatile instability tends to worsen with increasing wind speed, sometimes inflation points are deliberately placed where they will be correct when the kite is flying at a higher angle of attack, as is usual in lighter winds, but incorrect as the kites angle increases. This is so that some proportional inflation is automatically lost to assist stability as the wind gets stronger.
The Pilot, A Simple Parafoil
Using the "Pilot" as an example illustrates many cause and effect relationships for soft kites in general.
It is quite possible to make a variant of a Pilot without any flares that will fly satisfactorily albeit a bit nervously until is settles into steady flight at a reasonable altitude but its stability is then very sensitive to bridling angle.
The stability of the standard (with flares) Pilot is almost completely independent of bridling angle over a wide range.
About the only effect that changing the standard Pilot's angle of attack has is that increases the pull and the minimum wind speed in which flying is possible.
A major determinant of stability for the Pilot is weight. As is predicted in SLKS above, increasing weight without changing the C of G position- by for example using heavier fabric- will cause any tendency to volatile instability to increase. This is because although the C of G position relative to it's C of L doesn't change in this case, the moment effect of the kite's weight acting as a pendulum does increase (by proportion to the weight increase). Hence, the kite's response to disturbances will be faster, generating higher aerodynamic forces that can cause over-correction.
The Pilot's main resistance to volatile instability comes from frontal area drag. Decreasing the depth of the kite's open front by even 20mm on a 4sq.m. Pilot made in 35gm/sq.m fabric will cause the onset of volatile instability to occur at around 20km/hr instead of 40km/hr.
Of course increasing the open front depth will proportionally increase resistance to volatile instability but will also cause some loss of flying efficiency as measured by angle.
A way to win-win in this would be to increase the frontal depth only at the outside cells while decreasing the depth of the opening to the central two cells. This is because the kite's L/D is determined by total drag, which wouldn't change, but volatile instability is controlled best by drag that has as much moment around the axis of rotation as possible. Extra drag at the outside edges has more beneficial effect in damping out volatile instability.
The Pilot also has a minor amount of taper towards the front of each cell. This is to prevent frontal area superstability- which is often a problem with this type of kite (see the accompanying paper on kite stability). Actually, the amount of longitudinal wedge built into the Pilot is only about enough to offset pull-in of the trailing edge caused by "coal sacking"- which would otherwise cause the shape to be wider at the leading edge than at the rear (a common but often unrecognised cause of superstable behaviour in parafoils).
One other design feature of the pilot is worth discussing. This is the profile shape which has it's maximum depth quite close to the kite's leading edge and is then exactly straight on it's top edge back to the trailing edge.
This is to resist a form of instability that can be called "rolling" instability which parafoils and other kites with cambered top surfaces are prone to. Rolling over happens when, as the kite climbs to a higher angle of flight the airflow over the top surface attaches progressively further and further towards the trailing edge before separating. The effect of this is to move the centre of lift progressively towards the trailing edge also, which, can eventually lift the trailing edge over the leading edge in a luff or more commonly cause the kite to arc over to one side or the other until it crashes. Although sometimes difficult to distinguish from some types of superstability and volatile instability it is quite distinct, characterised by increasing speed (superstable sideways movements are at fairly constant speed) and absolutely no tendency to oscillation (volatile instability). As a blunt rule, open front parafoils with profiles for which the maximum camber point is forward of 15% chord do not usually suffer from roll over instability.
Using the Pilot it is also easy to experiment with the difference between open and closed leading edges. When Pilots are built with closing flaps (one-way valves at the open end of each cell) so as to make them easier to launch (especially larger sizes) and to prevent loss of inflation collapses when letting out line fast, they become more volatile unstable. There are two effects happening. Firstly, the extra fabric at the leading edge shift the kite's C of G forward which will improve volatile stability. Secondly, the smoother leading-edge shape, especially around the top edge, which was previously sharp, decreases frontal area drag and reduces volatile stability. The second effect is by far the greatest in this case and quite a large increase is required in the depth of the nose to add enough extra drag to compensate for the drag reduction that derives from the smoother shape.
There is currently a strong trend towards using closed leading edges for soft theme kites, either by way of gauze or now more often flap valves. Advantages of flap valves are greater resistance to loss of inflation during turbulence and much improved visual effect, but extra anti-volatile instability resistance has to come from somewhere else to compensate for this change.
Building Techniques.
I have written various papers from time to time about soft kite constructional techniques. These are available from the Peter Lynn website.
There are a few things that are new or worth repeating.
The "Super Ripstop" constructional system, beginning just a few years ago, has become the standard system for most builders of soft theme style kites.
It's advantages are;
*Requires less fabric, often 30% less which saves weight and cost.
*Allows considerable adjustment of shape without re-sewing- great for new shapes.
*Provides natural reinforced places for bridle attachments.
*Is very strong- 1000sq.m kites are easily possible with 50gm/sq.m ripstop.
*Not as aerodynamically efficient as ribs- more drag, which is advantageous for theme kites.
The only serious disadvantage of the super ripstop system as it is currently used is that through cords have no shear strength in the way that ribs have. This sometimes allows distortions of shape by fabric "rolling" around the leading edge and sides. It would of course be possible to rig diagonal cords as well in some approximation of a latticed girder so this is at least to some extent solvable.
For soft kite of complex shape crutch areas are the place where stresses can be extreme and where rips and damage usually starts. Crutches are places like under arms, between legs, at the roots of wings and other appendages where there is significant concave shaping. Stresses in these places are inherently high because many tension forces converge where there are internal corners. They become even higher when an adjacent appendage thrashes around in turbulent conditions. Reinforcement in these places should not be missed. Cording should converge from every direction that tension forces can develop from.
Generally, in terms of constructional approach to complex shapes, the choice is between two systems.
The first is to make separate component units that are than fitted together with vent holes at the connections to allow cross inflation. This has the advantage of easier visualisation of each panel and block by block construction.
The other is to consider the entire shape as a whole and make flowing panels that sew together to create the entire shape in one unit. This is often visually tidier but creates more crutch areas that need reinforcing and is a bit more difficult to create the first time.
Combinations of these two techniques are common. Eyes and other features like these are almost always planted on.
A trick when closing one space from another is to first sew in a cylindrical sleeve by one circumferential edge and close the opposite end with a fabric disc- this eliminates the visually annoying sucked in appearance that comes from just sewing a disc in directly. Inflation points, separate spaces.
And lastly, just a rough rule of thumb about bridles and bridle spacing. Towards a kite's front/leading area where most bridles are required, the bridles should be spaced longitudinally and laterally about the same distance apart as the through cords are long in that area. Towards the rear where less bridles are required the spacings can go out to about twice this.
The shortest bridles should be about twice as long as the spacing between the two most separated bridles- or else components of bridle tension will tend to cause buckling of the kites body.
Definitions:
Ram-air inflation: Inflating internal spaces in a kite by connecting them to a stagnation point(s).
Soft kites: Kites that are ram air inflated and have no rigid members.
Lift to Drag ratio, L/D: The ratio of the lift forces generated by a kite to the total drag forces applying to it. It is exactly the tangent of the angle between the flying line and the apparent wind direction at the kite.
Kite stability: For single line kites, is the ability to consistently maintain a position higher than the tether point and preferably in a plane parallel to the wind direction and through that tether point.
Centre of Gravity, C of G: The notional point at which weight forces act on a kite to pull it downwards.
Centre of Lift, C of L: The notional point at which aerodynamic lift forces act on a kite to drive it upwards.
Axis of rotation, A of R: The axis around which a kite oscillates at the initiation of volatile instability - somewhere between it's C of L (the point where an extension of the flying line would intersect with the kite's body) and it's C of G, that is, on the kite's centre line and somewhere in it's mid area.
True wind speed: The velocity of the wind relative to the ground or tether point.
Apparent wind speed: The velocity of the airflow as experienced by the kite
Peter Lynn, Den Hague June '01
