Quetzalcoatl high-performance all-paper glider

**Leif Ohlsson** · #1 09-29-2009, 01:56 AM

In the downloads section for aircraft you will find a kit for the "Quetzalcoatl" glider. It is a trial version developed to get a grip on the technique of designing and building all-paper, high-performance model aircraft, which I did not believe was possible until now.

You will find that the model is very colourful, with colours sampled from the Quetzalcoatl ("Winged Serpent") mythological figure of Mayan origin reproduced in the wing roundels. Reproduced on the tail fins is the Resplendent Quetzal, Guatemala's national bird.

The design features a Jedelsky profile, double-surface wing with an internal spar, and nine-layered fuselage hollowed out as much as possible. To get the C.G. in the correct position with as little weight as possible added, the nose carries a compact ballast of melted solder inside. The kit provides instruction for how to achieve this without leaving any visible marks on the finished model.

The kit comes as a six-page pdf, containing a full-colour version requiring reverse printing (on the backside) of one page. The kit contains basic instructions for this, and further tips will be given below. In addition, there is a black & white version for recolouring or test build.

The basic data of the model are:

Span: 375mm - 15"
Length: 300mm - 12"
Weight: 17.5g (including ballast)
Aspect ratio: 7.8 (span / mean chord; 85% of root chord)

**Leif Ohlsson** · #2 09-29-2009, 02:03 AM

It all started with the "Neato gliders" thread started by Outersketcher. In that thread there were tips by others leading to the simple G1 & G4 paper gliders by Mark Ivey.

I built those, and was stunned by the performance of the simplest G1 glider - it was fast, but seemed to have a L/D performance which astonished me, from what could be determined by living-room glides (a MUCH too confined space for this kind of gliders!).

I then studied the models from the Kobe Wings site. Letting the computer translate the pages from Japanese, I realized this was quite a movement in Japan, elderly men out with their highly-advanced, all-paper gliders, launching them with rubber catapults and achieving flights of quite some duration. The designs seemed very aerodynamically advanced and challenging.

I concentrated on studying several of the gliders found here,and finally chose the "Quetzal" for a number of reasons:

• It was the largest of the models there, and I wished to enlarge it further
• It had a Jedelsky profile wing, which would be most interesting to try out
• There were double fins, which would lend increased rigidity to the otherwise flimsy stab
• The fuselage was simple and constructed in a way that lent itself to easy redesign
• All in all, it could be made into a paper model resembling a classic model airplane glider.

Here you can see the original Kobe Wings "Quetzal" model in front of Mark Ivey's G1 & G4. They all come as white models, but what attracted me was the possibility of making a very colourful scheme. A quick look for "Quetzal" in Wikipedia soon confirmed this, and from there on it was all a lot of fun.

First, the original "Quetzal" download was enlarged to 110 percent, making the fuselage fill up an A4 sheet length. I then "stretched" the wings & stab parts a further 112 percent span-wise. This is how much they could be stretched and still be kept on the same page.

The Japanese instructions didn't come out very intelligible through the computer translation, but I gleaned as much as that tip dihedral was supposed to be 20°, and centre dihedral 2 x 7°. The C.G. marks in the original are centered on 67 percent of the root chord, with a ± margin of 4 percent each way. The original relation between wing & stab angles of attack was carefully kept. This is it as far as design parameters go.

I then redrew the whole enlarged and stretched model, and redesigned the fuselage completely with a more "bulby" front part and pleasing lines (to me). Full use was made of the remaining space on the A4 sheet to lengthen the nose a further 10mm, in order to reduce the amount of ballast as much as possible. Cut-outs in the interior layers were made as large as I dared to save weight.

A white test-build served to check the viability of the redesign, and to develop the method of using melted solder for ballast (more about this in the build report below). It was also very interesting to make a complete evaluation of weights of the different parts and sections of the model:

Weight distribution (g):

Cut-outs in fuse 1,5
Fuse parts 4,1
Stab 1,1
Fins (2) 0,2

Total fuselage 5,5

Wing supp. (3) 0,4
Spars (2) 0,7
Wing parts (2) 7,4

Wings, no supports 8,0

Wing parts total 8,4

All parts 14,0

Glued sections:

Fuselage with stab & fins 6,2 = 0,7 glue
Wings, no supports 8,5 = 0,5 glue
Complete model, no ballast 15,2 = 1,2 glue (total amount)
Ballast (melted solder) 2,2
Coat of spray fixative 0,1

Complete model 17,5

Cut-outs in fuselage would seem to save ca 10 percent of total part weights, and more than a third of the fuselage weight. This is not entirely true, since cut-outs in the front will have to be replaced with ballast, although less than what the corresponding parts weigh, since the ballast is at the very front. So making the cut-outs is worth the trouble, particularly in the main section under the wings (and in the tail, were it possible).

Main weight (more than half of finished model without ballast) is in the wings, so anything that could be saved there would be beneficial. Anything saved at the tail & empennage would save double or more in required ballast. This is why the tail of the original design is so thin. I particularly appreciate the extra thin reinforcement strips of the outer layers - the wider parts at the front cover the cut-outs in the main section of the fuse, and the aft thin stripes add to strength of the tail, with a minimum of extra weight.

A light coat of spray fixative added only 0,1g. Trials showed that a coat of matt solvent-based varnish, brushed on, would add 0.7-1.0g, depending on how thick you lay it on. A sprayed coat of similar varnish would be lighter. It is probably a wise move for weather protection.

**Leif Ohlsson** · #3 09-29-2009, 02:04 AM

The kit is designed for utilizing A4 size to its maximum. If your paper size is US letter, print at 94 percent and you should be alright. In the Adobe Reader advanced print dialogue, there is an option for automatic centering. Make sure this is turned on.

Making the colour scheme was very satisfying. The result of sampling the four main colours from the Mayan "Quetzalcoatl" figure reproduced in the wing roundels, and letting them flow out over the entire model, really stunned me. If you want to enjoy the same result on your model, you'll need to try out reverse (backside) printing. So here's the deal:

• If you want to study the design further for a bit, and read the instructions, print pages 5-6 on ordinary paper. This is the black-and-white version, and every piece of instruction is included in that, too.

• If you want to make a test-build without wasting expensive ink, do the same, but this time use the kind of paper you've chosen for building the model. For the test-builds, I used heavy paper, intended for making business cards (220 g).

• For the full-colour version, first print page 2-3 on the heavy paper. Then turn page 2 over, and feed it same short end first into the printer, and print page 4 on the back of page 2.

If you experience problems with a mismatch between the front and back side, check your actual paper size. The heavy paper sheets I used had shrunk by 1 mm, from 210 to 209 mm width as compared to its stated A4 size. It took me three misprints and numerous attempts to adjust the kit to finally find this obvious source of mismatch.

There was nothing wrong with the kit after all. 1 mm less width of the paper than what the printer believes is the case results in 2 mm dislocation of the backside printing, which is more than the margin I allowed for when making the coloured reversed printing backside.

The solution to this particular problem was to feed the paper in 2 mm away from the main, fixed, paper guide at the second printing.

There should be no major mismatch lengthwise, even if your paper has shrunk in that direction, since you feed the paper in same short end first at both printings.

If you want to recolour the model, the original was made in Illustrator (vector parts), and the pdf file should open in any other vector (recommended) or bitmap program.

**Leif Ohlsson** · #4 09-29-2009, 02:08 AM

The idea behind a Jedelsky profile wing is that the underside is flat and angled like an inverted V, covered by a curved upper surface, glued on flat to the rear flat section, and forming a bulb at the forward half.

To achieve this, score the underside fold line from the printed (black & white pattern) side (but do not score the corresponding line on the top surface - this line is just for alignment and actually superfluous). Make the cuts at the wing tip front halves, glue in the four-ply spar, and fold the leading edge to a nice rounded fold. Dryfit the two sides so that the trailing edges meet.

To help achieving the correct bottom angle, a simple jig might be helpful. Make two 1 mm sheets 210 x 120 mm. Score and fold each in the middle (60 mm). Tape a 1-1,5 mm wire in the fold to get correct angle.

Start with the main (centre) section of the wing. Apply glue to the areas indicated, fold and match the trailing edges. Lay the wing down on the jig, and "massage" the leading edge top part so that a nice curved top profile is achieved, while glueing it down to the spar.

An alternative to a jig is to press down the rear, flat sections of the wing to the building mat, while the front halves are bent down toward the table underneath.

While the wings dry out completely under pressure like this, you could glue up the fins to the stab. Note that both fins and stab have ever so slightly curved profiles. This is to provide increased rigidity, and if you fold the tabs on the stab and follow the glue lines on the fins, they will help each other to maintain their respective curves.

In the photos, I am building two models in parallell, just to be sure at least one would come out right (they were both fine, actually).

When dry, bend the wing-tip parts 20° (use the template provided in the kit), and glue up the tips much the same way. Unavoidably with the present kind of wing design, there will be a small gap at the front tip joint cuts. This will be filled out by small extra parts.

When the two wing parts are thoroughly dry, use the centerpiece dihedral template to rig a provisional jig (of sundry wood blocks or whatever) for the 2 x 7° center dihedral. Trim the center joint for a straight fit, and glue them together, taking care that the main part of the leading edge forms a straight line. Add the top center joining piece, to hold it together. When dry, add the bottom center joining strip.

To complete the wing, trim the tip joint covering pieces to a good fit, and glue them in very carefully. If you wish to make a quick job of it, omit them and fill the gap with some extra glue.

This part of the design comes from the original and it is still slightly unsatisfactory. I ought to try to come up with something a bit better eventually.

**Leif Ohlsson** · #5 09-29-2009, 02:16 AM

There are nine layers in the fuselage. A suggested glueing order is to first make three heaps of three layers each: The centre section with the optional launching hook for a rubber band, and two outer sections, one on each side. Be sure not to glue the forward section of one outer layer (4L or 4R), leaving it as flap to allow for the melted solder ballast later.

Let the three heaps dry out completely under pressure before glueing them together.

Both wing and stab sections are supported by the folded tabs included in the fuselage parts. Score these already before cutting out the fuse parts.

Building two models at one go has some advantages - alignment is easier for one thing. It is a good idea to mount the fuselage in a vice between some pieces of wood or thick card when glueing the stab and wings to the fuse. It will make handing easier, and alignment more exact.

When the glued-up model is all dried out, the fun part starts.

Weighing is fun, if you have one of these small scales. All my three test builds came out at 15.2g, possibly 0.1 g heavier in one case.

Now, on to the all-important balancing.

Mount a couple of suitable wood or thick card slivers in a vice, and place the model on them so that they coincide exactly with the CG center marks. Then cut a piece of solder and bend over the section of the nose where it is destined. Cut, or add to, the solder piece by piece, until the model balances well at the correct CG position.

Now for some smoke in your eyes.

Mount the model against a flat piece of metal, and fold up the unglued fuselage side part flap to access the front cut-out inside. Melt the piece of solder you so carefully measured out into that compartment. It looks really horrible, but is virtually harmless. The iron below will suck up most of the heat.

When you've melted all the solder, it will likely form a blob much to high to fit into the thin compartment. Further careful touching up with the soldering-iron and a suitable piece of wood or card will enable you to spread the half-melted solder evenly within the compartment, so that the flap can be glued up all tight and smooth.

You're done - if you deem it necessary, spray the model with whatever you prefer, and go flying!

If you've achieved the correct CG some further trimming is possible by carefully bending the trailing edge of the stab up or down just a fraction at a time. My test models flew well pretty straight off the board.

Due to weather being what it is here in the autumn, I have not been able to try them out on more advanced, high-performance flights, or rubber band launches. Any reports of such would be most welcome!

Leif

**airdave** · #6 09-29-2009, 06:17 AM

Not personally interested in building this plane, but still an interesting and engaging project...and, once again, you have presented a detailed tutorial in an intelligent and thorough manner. Thank you very very much!

I will be waiting to see the future test flight results!