My paraplane design.
The Black Rainbow.

Note that there is video of the successor to BR on Nerdipedia.


I want to design and build my own model paraplane. There are already designs out there but I want to do something a little different. Chutes look cool and are fun to fly but my model will be made as payload lifter. My primary payload will be a 120 gram 5 mega pixel digital camera.

I'm planning on having a payload pod which - like the one on thunderbird-two – can be changed for different needs. Wheels will be optional.
A small chute should be quite adequate for lifting a 150 gram payload. Even the cyber-chute could lift that and I'm sure I can do better than that by using carbon fibre and a brushless motor.

Some experiments still need to be to be done to finalize the design but I've made a start – I made the propeller shroud.

This project started with a trip to “Crazy Clark's” to buy a kilo or two of cheap candles. I also had a measuring stick in my pocket – a 230 mm length of bamboo skewer – just in case I found a convenient round object I could use instead of wax. The cyber-chute propeller is 230 mm and I'd like to used the same size for the new beast.

I came home and melted some of the candles in a rice cooker – the small one I use for melting wax – not the one I cook in.
The molten wax was poured into a round plastic container about 250 mm across. The depth was perhaps 60 mm.

More wax was melted and poured into a round Chinese food container.

I could have found a ready made former to make my shroud around but carving wax is more fun.
I would have liked to shape it in my lathe but my lathe has a “swing over bed” of 230 mm and I needed 250 or so.
The shaft which turns the chuck comes through to the left hand end of the lathe – you can feed 20mm material into the chuck through the hollow shaft.
That is where the second lot of wax comes in. I turned it on the lathe to make a mould for a carbon-fibre (CF) composite flange that I could use to attach my lump of wax on the outside of the lathe where I could spin a larger piece.

I G-clamped some steel to the bench frame to steady my hand and carved the spinning block free hand.
I was aiming to make a shroud around 240 mm ID, about 2mm thick and 20mm wide. I guessed it would be 50 grams or so.
I was just sort of practising – I trued up the block by eye and carved a grove into it that looked about right.
After a cup of tea I measured the block and found it was already too small for a second try. My “practice” grove would have to be the real thing.

I laid the carbon without removing the block from the lathe, I was hoping it would remain in position so I could grind it round after it had cured.
Wrapping some CF should be simple but it wasn't.
I wanted to use CF tissue for the first layer or two and had some long strips cut to wrap around. The problem was the binder in the tissue didn't dissolve in the resin.

For those who don't know – what should happen is the binder dissolves so the fibres are free to move and conform to the mould. A good example of this is glass fibre chopped strain mat (CSM) has a binder which dissolves in polyester resin – this works well. However if you use CSM with my epoxy the binder doesn't dissolve and the matt won't conform like it should.

Later tests found that acetone will dissolve the binder so it is possible a small amount added to the epoxy might have worked – I'll need to test this some time.

CF is not noted for it stretchiness so it was impossible to make the tissue conform to a compound curve and I gave up trying.
I had some uni-directional material prepared as well but it had a similar problem. I pulled this material apart and wound the fibres around as best I could. I used short pieces of CF cloth and then managed to use the tissue once the surface was a simple curve. Once I build up what seemed like a good thickness of material I wrapped several layers of peel-ply around it. Normally I'd vacuum bag it to remove excess resin but with simple convex shapes like this stretching ply over it works pretty well.

This is how it looked next day when it had cured. The brown cloth is polypropylene peel ply. Click the image to enlarge.

This is a photo after the ply was peeled off.

This is how it looked with most of the excess sanded off. If you click the image you can see the peel ply pattern in the bits which haven't been sanded off yet.

Then after the resin had cured I spun the block and and trimmed the excess CF from the edge with a hand held lathe tool. I tried to grind the CF using a small low-power tool. It hardly left a mark so I grabbed my belt sander and went for it.
I should have removed more wax before I started because the wax coated the sanding belt and slowed down the work. It did the job though and I'm happy enough with the result.

Most times when I use a wax mould like this I melt the wax to free the piece. This time however I just spun it and parted the shroud and 10mm of wax from the rest of the block by using a 2 mm CF rod as a parting tool. Once separated the wax was easily broken and removed from the CF shroud.

The shroud ended up 240 mm ID, about 2mm thick and 21mm wide – it weighed in at 38 grams and is just what I had in mind.

I also noticed my big rice cooker- the one I cook in and don't use for wax – is almost the perfect size to use as a former to make a shroud like this. I'd have a wrap a few layers of tape around to build it up a little and the result wouldn't have the same profile but it would work.
I'm sure a trip to the shops with a measuring stick would also find something suitable.

I ordered a motor on the weekend from
I ordered their
Single Stator - Special Edition” which should have plenty of thrust and only weighs 27 grams. It is a direct drive motor which can spin a 230 mm prop without needing a gearbox. The whole power system should weigh in at well under 150 grams. That means prop, shroud, ESC, motor, motor mount and supporting structure.

My kite arrived.

The “rainbow stardust” kite arrived so there was a detour to see if it works.

This is how it came. It is “cross bridled” and the lines are pretty long. Unlike the cyber-chute the lines have a two tiered arrangement.

I removed the cross bridle lines and shorted the main lines to match the length of the lines on the cyber-chute canopy.

An end view showed the three sets of lines. Note that they don't line up as they do on a square shaped chute.

This is more or less the rigging I had the best results with. It seems like no one has made a cloth canopy fly as well as the plastic one. Why the shock chute works so well is a mystery that would be well worth solving.

After I shortened the lines I moved the main arm of the cyber-chute gondola back a little, attached the kite and tried to fly it.
The thing turned right on a throw launch (as usual) and crashed. The best I could do was to get it to fly just above the ground for a short distance. Since I was flying in a small space I had to cut power after twenty meters or so.

I was pretty sure I need a higher AOA (angle of attack) on the kite but also suspected the two tier thing wasn't a good idea. I later found para planes which do use a tiered bridle but I restrung mine to be more “normal”.

I didn't know what line lengths to use and too a guess. I had too much AOA first up but it was well behaved, it refused to climb and just floated down and landed. The server pull to the right was also gone.

The best results I had were with the front lines 78cm second 80 and rear 84. With these lengths it would start to climb then give up and come down again.

This behaviour was a mystery until I measure the static thrust. It was also 300 grams but decreased rapidly by almost a third. Apparently I'm overloading the battery.

I'm sure this will fly but it obviously needs more power than the plastic cyber-chute canopy. I'm not sure exactly why. It is a little heavier but the overall weight is pretty much what it should be. More power may just be the price to pay for using a cloth chute.

Motor arrived.

Back to building.
The motor looked good but came fitted for a straight bearing tube. I had to make some jigs to push out the tube and replace it with the firewall style mount.

It took three attempts to make the motor mounting tube. First time it got too hot and the wax plug melted. Second time the CF looked OK but I accidentally cut through the side when I was trying to tidy it up on the lathe.

On the third try I did a vacuum infusion of the resin instead of a wet layup and this was much easier and less messy.

There was a problem machining it because I had a nice inner surface but nothing much to hold it straight in the chuck.

So I made a matching plug out of cured epoxy which I could hold in the chuck and press the tube onto it and turn it with some accuracy.

The end product weight in at 11 grams but a little more was removed later and there will be ventilation holes drill. This should shave another gram or two off the weight.

The wax plug for making the CF motor tube.

This is all the wind making equipment minus the ESC and wiring. It is tacked together with CA. This is the back of the pusher and the foam core is visible inside the main fin.

The main fin is made from 10mm PVC foam covered with one layer of 200 GSM carbon – a single piece. This was also vacuum infused. Embedded in the leading edge is a 10mm CF tube for running wires up to the ESC. The resin gelled before the infusion was complete but I managed to add more resin and finish the job without any visible defect in the finished piece. After trimming to size, some foam was removed from the inside of the top of the fin to allows the motor wires to be brought back to the ESC on the outside of the tube.

Since the motor is an “out runner” I need to keep the wires away from the spinning can, so it is easier to take them past on the outside of the tube and inside the fin.

The two upper fins are sheets made from 4 or 5 layers of the same carbon cloth.
Its come out a little heavier than I thought but should be pretty close to the 150 gram mark, I'll never know exactly because this section won't be completed until it is fused to the other parts.



The top common steering mechanisms that I know of are a “tilt bar” and the system used by the sky surfer and it's clones. I don't like tilt bars because they will get damaged when a crash occurs. I think the surfer systems is better in this regard but I though I could do even better by putting the main mechanism below deck where is is well protected and bring lines topside inside strong CF tubes.

I thought about different ways to do this and decided to use a servo controlled pulley with captive control lines.
There are at least three advantages of a pulley over a control horn.

The first is it that it has a linear relationship when converting rotary to linear motion. By this I mean one degree rotation will produce the same amount of travel in the line regardless of the amount of deflection. On the other hand the servo arms on the sky surfer will have the greatest travel per degree when the controls are level – this is when you want the finest control not the coarsest, this could be compensated for the electronics. The sky surfer's arms move in and out as well as up and down so it isn't that simple.

The second advantage is longer travel for a given width. By making the cable travel around the perimeter of a circle you increase the travel by over 50 percent (pi/2) for a 180 deg servo deflection.

The third advantage is the line stays in the same path so it can (easily) be run through a tube.

My pulley was another case of a practice piece turning out good enough to use. I didn't have any suitable stock material to use so I poured some epoxy resin into various containers to make rods and disks of various sizes. I found my wax bath is excellent for post curing. It takes many days for resin to cure. It will turn solid fairly rapidly but doesn't reach maximum strength for quite a while. This resin splinters if machined with a day or two of setting but a few hours sitting in 120 deg C wax makes it really hard.

The pulley is 50mm in diameter with two groves which are just deep enough to hold the cable. The cable is gold-n-cable #507 - this has a brass plated stainless steel cable approx 0.8mm in diameter plus a low friction nylon sleeve.

When assembled into my system the cable is captive and shouldn't pop out when the line goes slack or into mild compression. The puzzle is how do you assemble it.

My pulley set. The servo disk is press fitted into the pulley and will be glued later. The two groves a clearly visible. Two pieces of cable are laying on each side. One cable cut in half gives just about the perfect lengths I need.

This is a trial fit the see if the holes in the housing matched up with the grooves (they did). This is before I drilled, slotted, glued and all that. On the left is the yellow (golden?) sleeve. This side mounts against a CF bracket so the pulley is totally captive. The servo mounts onto the other side.

I drilled a hole where I wanted to anchor the internal end. It was from inside the groove and at an angle. I tried to shape it so the pulley could be assembled inside the housing and the wire fed in last and glued in place. I drilled another hole in the housing I could poke some piano wire though the encourage the line to go where I wanted. This was difficult but seemed possible. However the pulley would sometimes pop out.

So I figured that if it can pop out I can make it pop in too. It was very easy to put one wire in this way. I bent the wire ends 90 degrees and glued them in place. Putting one wire is easy but two wires is not possible. So the trick is to make the wire on the right in the photos pass through a slot instead of a hole.

The slot was made the correct size to take the sleeve and this side will mount up against servo bracket making it a snug fit.

The bottom skin.

The bottom part of the paraplane is an interchangeable payload module but for testing all that is required is a protective shell. The skins of the future module will be vacuum infused carbon or kevlar.
The prototype I made is just a wet layup so it is heavier than it needs to be. The world wide shortage of carbon fibre has prevented me buying any more so I used glass kelvar and carbon tissue. It weighed in at 84 grams which is a little on the heavy side.

I have already cast a block of wax to carve the plug for the camera module but wanted something quick and dirty for the test flights. I looked around for something I could use to make a pattern and found some plastic containers at the supermarket. I bought everything that looked about the right size and choose the best fit at home. One of the containers was close to being the correct size but need to be spread a little. Some pieces cut from a fruit bottle gave it the extra size and shape I needed. It ended up looking a bit like something from a montie-python movie but it will do.

Willow fresh choice 700ml is the right size for the existing steering gear but not big enough for the radio gear or wiring.

It was spread 10mm and super glued to a plastic cutting board. Pieces of bottle were glued to the ends then the gaps sealed with hot-melt glue.

This is the view from inside after the composite was cured with the pattern still in place. The yellow kevlar is visible through the glass fibre.

The deck was crudely shaped and drilled. Matching 3mm CF post were glued onto the inner side of the skin and two rubber bands hold the shell in place. There are lots of ways to do this.

July 19'th
Almost ready to fly.

Next the CF tubes for the gold-n-cable where trimmed to length. Flexible tubing and heat shrink were attached to help protect the wire against kinking in the case of a crash. The ends of the wire were bent back, bound with fine copper wire and soldered. This provides a good attachment point for a brass swivel.

A pheonix-10 brushless ESC was recycled from the carbon-sixty project and initially connected to the motor with clips so I could make sure the motor direction was correct before soldering. I hung the model from two strings hooked to the ceiling. The thrust when powered by a two cells LiPo was disappointing. The power to weight seems worse than the stock cyber-chute. Using three cells make a big difference but it still isn't spectacular.

There was a gondola stability problem. This didn't surprise me too much. The gondola would oscillate from side to side, sometimes quite violently. Adding weight or reducing power definitely helped. For comparison I suspended the cyber chute gondola the same way and it was much worse so I figure it won't be a problem when attached to a real chute.

I soldered the motor directly to the ESC and ran three wires down the inside of the fin. Normally you'd expect five wires. Because I'm not using the BEC on the ESC (in other words I'm not using the ESC to provide 5 volt power) I don't need the two 5V power wires. I only need the two battery wires and one signal wire for the throttle. The ESC was hot glued in place.

To supply 5V for the radio and servo(s) I'm using a switched mode park-BEC. This plugs into the radio's ESC channel.
Three wires also run through the base of the fin. These are two battery wires to power the BEC and the ESC control signal which comes from the BEC and goes all the way up to the ESC.
The radio is a 6 channel JR which was also recycled from the C60 project.

The chute attachment point. The sheath of the gold-n-cable can be seen through the clear tube.

ESC glued in place. Yellow wire is battery positive, black is battery ground, white is the ESC control line.

View from the rear underside- steering servo, radio and BEC. The BEC is the plastic wrapped thing on the right. Radio top left. Servo middle with the end of the gold-n-cables and the CF tubes visible at the back.

July 20'th - Its too heavy.

The weight was pretty much what I expected but the empty cyber-chute gondola is lighter than I realized. My lift module alone (no shell) is 48 grams heavier than the cyber-chute gondola.
I didn't think this was good enough so I started making holes to let the weight out. I took 16 grams off the servo mount, it was originally intended to be strong enough to land on but not any more.

The hole in the main fin only saved five grams but does created a nice carry handle. The holes in the upper fins only saved two grams. I took 3 mm off the front of the shroud, it was a bit sharp before and it still pretty strong now (see photo below).

I got the weight down to 302 grams and stopped.
So apart from having a tougher more rigid frame and experience I haven't gained anything over hacking a sky surfer.

The bottom shell still heavy but this doesn't matter because it is temporary.

With the lighter weight and my new three cell 1040 mAh battery I think I have over a 1:1 power to weight ratio on the gondola with it's shell.

The biggest potential for weight saving is in replacing the standard servo for a mini.

I bought a “blue-bird” brand servo that looks like it will be up to the job and is 30 grams lighter! Just to make life difficult the spline on this servo is a little bigger than the hi-tech one. If I can't make it fit it will take a complete disassembly of the steering gear to replace the servo wheel.

I choose not to replace the servo for the test flight and laced the servo mount in place instead of gluing and left the servo just tacked in place because it is all coming out ASAP.

The new holey looks.

It is hard to convey how stiff the shroud is. This photo of a 3.5 kilo house brick balanced on top might help. The prop still has clearance.

July 22'th – D-Day.
My friend Barry came over to video the event. The weather wasn't great, a little gusty with a shower or two. I wasn't sure which canopy to try. The plastic chute had never flown in a two string mode so I tested it on the proper gondola first.
To convert the four line chute to two lines – I joined the front and back lines together with a 15cm extension on the front. I tried to fly it in a small partly tree covered area beside my house. I had a lot of trouble controlling the chute enough to even launch it. When I did, control was poor and it didn't want to climb. I thought the AOA was too high and shortened the extensions by 3 cm by clove-hitching some twigs into the lines. It then flew better.

Barry's time was limited and I decided to try the kite on the carbon creation. The kite is much more stable. I launched it while standing beside a tree. My videographer and I were both caught by surprise because the model climbed almost vertically, turned and got catch in the tree. My attempt to dislodge it by applying power wrapped the lines around the prop. After much throwing of sticks to try to dislodge it a handy six meter length of steel tube I had nearby broke to pesky branch and I caught the gondola as it fell.

I had no doubt now that this thing could fly :-)
We untangled the lines and there appeared to be no damage.

We went to my mother's place which has a big yard.
This time I launched with less power but it wasn't enough and it almost kissed the ground. More power and it lifted and swung left (as you'd expect do the prop torque). I let it come round then straightened up, cut power and made a perfect catch (which Barry missed).

Next launch was full power again and it climbed ok. It had better control than the cyber chute but I need a lot more practice. It was only a short flight with a few turns when I steered it towards me and cut power. It had too much height for another catch.

Then I lost it, it turned down wind. I may have reserve controls and steered to there, I'm not sure. It was on a nice stable glide into the neighbours yard and I didn't try to stop it because I would have ended up on the fence.
I'm not sure exactly what it ended up hitting, video evidence may help figure it out later.

One rubber band broke but the gondola shell was still in place and undamaged but the that poorly secured steering assembly broke loose and that was the end of the test flying – but it flew and that's a good thing. I'd even say it flew well.

We didn't get any stills of the flying – only video. These two photos are posed throws.

As I said earlier, if you just want a working parafoil there are easier ways to do it. I did this because I had some ideas I wanted to try out. If I built another one I'd possibly do it differently but until you try building one, it is hard to imagining exactly how things will work out. I had to take a lot of guesses and some were a little off.
If model paraplanes were as popular as some other types of aero-models we'd probably see some new products for sale. In particular I'd like to see a robust compact steering assembly. The one I made works but it is large.

I've put a ridiculous amount of time into this and will be reducing this to work on other things.

Before building the camera payload module I'll be improving my vacuum infusion techniques.
For more see page two.

Cheers Eddie.M.

My models page is here and my home page is here.