Dabbler

Here's my progress on designing and building the Dabbler. A lot of the information on this page is technical kite stuff, often my own thoughts on what I should do next, so I've tried to highlight things to make this page easier to read. For a quick summary, just read the first paragraph of each section, in the blue boxes. Often, these have nothing to do with the design of the kite, but are just a record of where I went and what I did, and how much fun I had. OK, so pale blue is not the prettiest colour in the world, but it does the trick.

Thursday 12th July, 2001 - Finished design of the kite using my own modelling and calculation program. For those who are interested, here's the details of the final kite design and how the program works. For those who aren't interested, you can skip to the next blue box...

First, an aerofoil profile was defined. I based mine on the profile for the top skin of the Dominator II ram-air foil, but I smoothed it by hand to get rid of what my uneducated eye saw as bumps.

dabprofile.gif (2322 bytes)

The next stage was to define the skin shape for the kite. I wanted it tapered toward the edges, so I used parabolic curves for the leading and trailing edges. Guessing that most of the lift occurs at 1/3 back from the leading edge of the kite, I tapered the back more than the front. (I have since questioned whether this was the right thing to do...) I didn't taper it to a point, because I wasn't sure where the balance point would be. Also, I wanted four line control.

dabblerskin.gif (3210 bytes)

The next thing was to guess what shape the canopy was going to form. I figured it was going to be close to a a semi-circle, but I wasn't sure, so I wrote a finite element analysis function to model the vector forces for lift and tension on each panel, and the resultant canopy shape is...

dabcanopy.gif (3105 bytes)

It's very close to a half ellipse (90% wide 105% high), with about 20cm of tangential extension at the wingtips.

Each point on the skin and canopy is the point for a seam in the skin. To fully model the kite, the profile is scaled to the correct length for each rib, and rotated/translated onto the correct position for the canopy shape. The result is a 3d model of the kite.

dabblersm.jpg (3117 bytes)

Final kite specs...

Wingspan - 3.37m

Max chord - 1.2m

Flat sail area - 3.02 m²

Foil equivalent sail area (area/1.36, as used by Cabrinha and Wipika) = 2.22 m². Hence the name Dabbler 2.2.

Aspect ratio (winspan squared/sail area) - 3.76. I designed to an aspect ratio of 2.8, based on (wingspan/max chord), but technically, this formula should not be used for a tapered kite, because the max chord is not applicable for the whole kite.

To project the panels back to a 2d pattern, each panel is logically broken into triangles and then folded out flat to give the 2d pattern for sewing. Because the panels are actually a string of non-flat 'rectangles' (bow-tie polygons), the rectangles are divided into triangles by cutting diagonally (triangles must be flat, because 3 points are always coplanar). The diagonals are alternated to minimise the cumulative error by cancellation. Once the triangles are defined, it's a matter of simple geometry to programmatically unfold it flat.

Once the 2d projections were completed, the output points were imported into Excel to be printed as an Excel chart. That was not much fun, because Excel can't even force a 1-to-1 aspect ratio, let alone print out to an exact scale, so there was lots of trial and error in getting the scaling right. Excel also seems to have a size limit on charts at around 15 pages (above which it prints blank pages), so I charted it in sections of four pages. Seam allowances were added manually at the stage of cutting out.

Sun 22nd July, 2001 - I sewed the panels together today. Made from Ripstop nylon, four purple panels in the centre, two yellow panels on either side, and two red panels on each wingtip. Took the dog for a walk, and ran through the park wearing the half finished kite as a batman cape.

Monday 23rd July, 2001 - Finished hems and played with swinging the kite over my head. The nose was very apparently folding under the rest of the kite, so I added ribs on every second seam with dowels along the bottom edge of the ribs. I ran up the street at midnight, holding the edges of the kite, and the kite seemed to hold it's shape ok.

Wednesday 25th July, 2001 - Spent an hour down in the park last night. It was very eerie with the mist lying 3 metres deep over the whole park. Both myself and the kite got rather wet from the mist. It must have looked very surreal to the guys sitting on the balcony of the soccer clubhouse, watching a billowed kite drifting through the mist. Anyway, back to the kite...

The kite billows over my head well when running, but does require the rear of the wingtips angled in to keep the sail filled. With the wingtips held at the correct angle, I could run the length of the field, turn around and run back with the sail filled above my head. This required applying plenty of brake, although to get the kite in the air first required a brief pull on the front edge. I tried flying it on four lines, but could not get the sail to fill while running backwards - it would just fold in half on itself and drag along the ground. I'll probably have to wait for some wind, and possibly a friend to help me launch it.

My current ideas to improve the stability of the kite are...

Test out the kite with a leading edge spar, and see if this helps anything - particularly launchability. This will make the kite harder to pack up and transport, but I'll do it if I have to. Eventually, I'll work on an inflatable spar for the leading edge so it's easy to pack up. I tested out a 6mm dowel in the leading edge, and it flew beautifully for about 1 minute on 1m flying lines, before the dowel snapped. At least I know the concept is sound.
Resew long darts into the trailing edge to increase the angle of attack of the profile. As far as I can think it through, this should also be the same result as designing the kite with the front edge tapering more than the back edge - moving the centre of lift forward w.r.t the tips, and putting more pull on the front lines.
Try wire supports along the top curved edge of the profile. This may support the aerofoil profile better than the straight spars along the bottom edge of the ribs.

Saturday 28th July, 2001 - I resewed the dowel pockets so the dowels won't fall out, and tried plastic tubing in the leading edge to see if this gives enough support. The answer is no, it doesn't, so I pulled out the tubing again. The 20+ knot westerly winds today managed to snap the rib dowels when the kite inverted, so fibreglass rods are definitely the go all round. Also managed to snap a flying line and the bridle on my speedwing, so some serious string is on the shopping list too.

Wednesday 15th August, 2001 - I've finally got my fibreglass rods from Kite Magic in Sydney. The short version of today's update is that the kite now flies! Yaaay. I flew it on 20m lines. The wind was a bit gusty, so it never stayed up for long, but it did go up. Flew it in the soccer fields near work. Unfortunately, nobody was around to take a photo of the kite for me.

I've replaced the wooden dowel rib supports with 3mm fibreglass, and the leading edge has a 5mm rod.

Anyway, I now have a new list of what needs doing...

Sew rib pockets to the leading edge properly so they aren't so loose.
Sew a support line into the leading edge to stop it stretching. I haven't seen this happening, but it is needed for step 3...
Sew pockets into each end of leading edge to tuck in the leading edge spar to stop it falling out, and to stop kite bunching up when launching.
There's currently an obvious pulling out of shape at the rear line attachment points, so I need to resew these along the trailing edge.
Sew darts into the sail to bring trailing edge lower and slightly curved down to catch the wind.

Some flight characteristics were interesting...

The normal take-off procedure was lie the kite top down with one end of the kite pointing directly toward me, and the flying lines slightly off the direction of the wind. I would then pick up the upwind handle first, and then walk back toward the kite and get the downwind handle. Walking away from the kite would peel the far tip of the kite up off the ground toward me and once the wind caught it, it was a fairly easy matter to pull it up to a wingstand and take off. I think this is the recommended launch method for a C-Quad. In low wind, it seemed to work better if done quickly, kind of like peeling off a band-aid. The only problem was that the far edge would sometimes bunch up along the leading edge spar instead of nicely peeling the kite off the ground, so that's why I need to secure the leading edge to the spar a bit better.

I also noticed that even from a standing start (i.e. the kite is leading edge up, and ready to head for the sky) the kite wouldn't launch directly vertically. I had to roll it over on it's side and get it to launch sideways into the wind. This is quite probably normal for traction kites, but I just thought it was interesting. Fortunately, when ground level wind was available, the kite was fairly easy to roll onto it's side by juggling the lines. Perhaps in stronger ground level wind, it is possible to launch directly vertical and directly downwind, but it's not a good idea with a traction kite anyway.

Predicatably, when the wind dropped to nothing, the kite flattened out from lack of tension on the leading edge spar. To combat this, I may tie the wingtips together to prevent the kite from totally flattening, although I will have to change my launching technique. I'll wait and see if this behaviour changes when I sew the darts.

Occasionally, when flying to one side, the kite collapses on the top edge, and with the top edge tucked under, the whole kite collapses. I'm hoping the darts will stop this from happening. If not, then possibly a stronger leading edge spar is required to hold the kite in shape more firmly. Again, I'll wait and see.

Oh, one last thing - when I got my spars delivered, they were wrapped in some sort of thin plastic stuff, like a really strong plastic bag open at both ends, 5cm wide and 5m long (presumably it came off a roll). I'm thinking that this stuff would be ideal to make inflatable spars. I'll keep this in mind for the Dabbler stage 2.

Sunday 19th August, 2001 - I was playing in an Eisteddfod at the Gold Coast, so I went a few hours early and spent time on the beach with the kite. I'm getting the hang of flying it now, so I don't crash as often. There were some guys there flying NGens, and they let me have a fly of a 3.5m NGen - I was very impressed. The power was very smooth and the kite was very controllable. I was also surprised to find that they could be controlled like a 2-line kite - whereas the Dabbler does not respond at all to pulling one hand more than the other, but instead only responds to twisting the handles as a four line kite.

The Dabbler's wind window is about 20 degrees off - i.e. it flies at about 70 degress off the ground, and same to each side. With the good steady wind at the beach, it was flyable, but needed careful attention. It still has a tendency for the wingtips to flap inwards which needs to be combatted by pulling quickly on the back lines. I've got some ideas about moving the line attachment points, or possibly putting a simple bridle on each side to help this.

Fri 31 August, 2001 - I haven't done any more flying, but I've had some more thoughts.

The current problem with the Dabbler is that the wingtips flap around. When there is not enough tension on the rear lines, the wingtips are only held by the front, so the lift in the middle of the wingtip causes the aoa (angle of attack) to go to zero, or less. This has the double effect of losing all lift at the wingtips, and also collapsing the foil profile. The centre of the kite still has lift, so it pull the kite into a really tall ellipse, and the wingtips just flap around. There's no steering control without the wingtips, and the kite drops. For the wingtips to not collapse, they must be balanced by the tension in the lines which requires the brake lines to always be under tension. The easiest way to enforce this is to add a bridle on the front lines, so that the main flying lines support both the front and back of the kite wingtips. The line tension balance point must be behind the centre of lift for the wingtips, so that the wingtips tend toward an increased aoa, rather than a decreased aoa, but only slightly. Brake lines can still be seperately attached to the rear of the wingtips for four line control. Once I have found a good balance point, Dabbler phase II will probably replace the wingtip bridle with another panel. A design to combat this could use more of a Flexifoil profile on the wingtips which gives more lift at the front edge, and less at the trailing edge, which would then require moving the wingtips further rearward to keep the centre of lift in a line across the kite.

Currently the Dabbler only just takes off with full front lines, so with the wingtip bridle in place, I may need to take out the darts to move the wingtips forward and give the kite a better angle of attack to take off. We'll see on this one.

To allow two line flying (or four line flying with two-line turning characteristics), another small bridle is required, where the brake of one side is attached to the upper front edge of the opposite side. This line should just have no tension in normal flight, but as soon as one side of the kite is pulled more than the other side, it tightens pulling the front of one side, and applying brake on the other side.

Sunday 2nd September - Off to the beach again for another test run. I flew the kite for several hours on a magnificent beach at Marcoola, Sunshine Coast. Flying the kite on an almost deserted beach with a fantastic sunset was an amazing experience, and reminded me of how beautiful God's world is. A few families with young children wandered past and stopped to watch the kite. Some fishermen came over and had a chat. Others just quietly fished while the kite sailed over their heads. A few planes took off overhead, because I wasn't far from Coolum Airport. I was mildly concerned at how low they were, but my lines were only 20m long, so I'm fairly sure it was legal. At one stage I accidentally dunked the kite in the shallows, and in 2 seconds it gained about ten kilograms of water and sand. I had to wade out into the waves to wash off the kite, and then hold it in the wind for 10 minutes to dry before it would take off again. Eventually it got too dark to see, so I packed up the kite and just sat on the beach for a while and watched the full moon rise over the water.

Anyway, about the kite...It is coming along very well. I tried quite a few simple bridling systems, with the result is that the kite is now very stable. It was actually stable enough to fly with both handles in one hand while I picked up my shoes (although I had some delicate finger movement to keep it centred). I've actually tried to upset it by pulling and shaking the handles in wierd ways, and it keeps it's shape very well. The wingtips don't flap anymore, and the kite responds well to 2-line or 4-line steering styles. The kite is quite fast, probably due to the 12.8% profile thickness. (The profile graph at the top of this page shows the max height as 15cm, but don't forget that this is on a chord length of 1.2m). I'll increase the thickness to around 15% or 16% on the next version. There were a few times the kite got itself completely twisted up and fell to the ground in a knot, but I think this was while I was still experimenting with weird bridles.

The black bit is the kite (very detailed picture, I know). The thick black line is the leading edge toward the viewer, and the thinner line is the trailing edge. The flying lines are not shown.

The bridle comes in 2 parts with 2 very different functions. The first part of the bridle (red) basically keeps the wingtips curved. This makes a very big improvement in stability, and also improves takeoff, although one person landings are a little more difficult, because the curved kite no longer lands easily flat on the ground, and tends to blow away. :)

The second part of the bridle (blue) has 2 functions. Firstly to keep the trailing edges of the wingtips in (i.e. keeps positive aoa on the wingtips), and also has the double function of giving 2-line steering. I still have to play with the exact measurements, because it is not totally symmetrical at the moment, and still has the occasional unstable turn, but it is far and away better than it was.

Most of the time, the kite can be flown as a 2-line kite, with the special note that the tension is fairly evenly divided between front and back lines, unlike most 4-line foils where the tension is fully on the front lines. Even taking off does not require any 4-line steering. At the top of the window, the kite can be depowered to an extent, although it doesn't like it much, and tends to lose stability and turn to one side or the other. This is easily counteracted by steering back to the middle, but it is a bit 'loose'.

Upwind performance has not really changed that much from where it was. It's not great, but I think chikara skin and spectra bridles will help reduce the weight and the drag. I have proven by demonstration that it's not a water relaunchable kite at the moment, but this might be made possible by using chikara skin and airbags.

I also need to fix up the wingtips - the wingtips ribs don't have the reinforcing that the other ribs do, and time has taken it's toll. I've discovered a rule of kite making - if anything can break, it will. If you think some part of a traction kite is not under stress, you will be proven wrong when you fly it.

Saturday 22nd September - Well, I've been flying the Dabbler a few times, but not actually making changes to it. Today, I was forced to...I left my quad-line handles at home. I flew the speedwing for a bit, but the wooden dowels didn't last long, and the speedwing hugged the ground after less than five minutes in the air. I really have to get some proper spars. Anyway, I had the Dabbler all setup, and only dual-line handles. I've always said it was possible to fly the Dabbler on two lines, so I hooked up the wingtips to fly on two lines, and I was very pleased when it went up first go. We got out the camera and took some nice shots. I've put some of them up on the kites gallery page. The wind was a bit gusty and kept changing direction, but there was a nice sunset, and it was fun anyway.

Attaching the dual lines was a simple matter of tying a 1m loop from the front to the back of each wingtip, and hooking the flying lines to the centre of the loops. This gave equal balance to the front and back. I was kind of surprised when it went up first time. It flew really well, and I didn't bother adjusting it at all. The only real disadvantage I noticed of flying on two lines was that it made it harder to take off. Vertical takeoffs were impossible - I had to roll the kite to one side and take off along the ground, then turn into the air. Also, it was harder to roll the kite over into a good position for takeoff. I think with good ground level wind, this wouldn't be too hard, but quad-lines definitely give more control and make takeoff far easier.

Once it was in the air, it flew really well, and dual-line control is certainly a lot simpler than quad-line.

The photos allowed me to see the Dabbler from angles I haven't seen before, which was very educational. The first thing I noticed was that the foil profile is distorted, as evident in this photo, particularly in the yellow section on the far side of the kite. The front half of the profile is a nice curved shape, but the back half dips in instead of being straight. I think this may be caused by the fact that I haven't used a zero-height profile on the wingtips. This allows the skin tension to pull the wingtip profiles around. I think this will be easy to fix to see if it helps.

Another factor in the profile distortion is probably caused by the material pulling out of shape diagonally to the material grain. This kite is built from garment-grade ripstop, which is not ideal for kite-making. I may try sewing diagonal support lines into the kite surface. Proper kite making ripstop does not stretch much diagonally, and therefore I won't need this in the future, but I may as well fix this kite.

The other thing I have noticed is that the bridle does actually pull the canopy out of shape very slightly. I don't think this is a problem, but it is something for me to keep in mind when I write the instructions on how to build the Dabbler.

Sunday 21st October - Not many updates recently, with my son distracting my attention. It's amazing how much more distracting children become when they are outside the womb. I've been out flying a few times since the last post, and here's what I've found.

I've added reinforcing strings to the wingtips. These are in an X shape on each wingtip. This seems to stop the profiles from deforming. I haven't had a real close look - I've only seen it from 30m away :) - but it seems to work.

The other major issue is that when flying on two lines, there is a lot more stress on the wingtip supports. This is a problem particularly in stronger winds. I now recommend fully reinforcing the wingtips front and back with nylon webbing or dacron, and using 5mm fibreglass supports. 3mm supports are just not strong enough for the wingtips.

Wednesday 7th November - Joshua is 1 month old today!

I have resewed the wingtips of the Dabbler 2.2 with no profile, and some nylon strap re-inforcement to try to make them stronger and easier to sew. I have tested this out, and it has made a marked improvement in the kite stability. The extra inherent stability has allowed me to lengthen the cross bridle slightly, giving less drag and a wider wind window.

I'm also working on importing the plans into a CAD program, so the plans will be much easier to print. Fortunately, the DXF format is well documented on the internet, so implementing this into my design program was not too difficult. DXF format plans are now on my web page.