| Return to Builders corner home | Return to Story page | Tämä sivu suomeksi |
With that fuel tank redesign, it was noted that aileron control, initially done with cables, was flawed. Parts that would produce the desired aileron differential movement would not fit inside the reserved space. So, after several iterations, a move to a pushrod system was made. We are working on it. Not straightforward. The system should be such that the wing can be removed or assembled from/to the fuselage so that the pushrod system inside the wing does not need to be removed for the operation.
The fuel tank was originally thought to be a rubber bag tank on the wing's leading edge. It would use the entire space from the spar box forward. In the meantime, we ordered four tanks for use in the existing aircraft. A rubber bag tank has a specific useful life, after which it hardens and gradually leaks. It turned out that, contrary to the information received at the fair, there are problems with their availability. Even if the price and factors are OK, 1.5 years is too long for a builder. It would help if you continued to herd to get them even in that schedule.
So, we started looking for another solution that would be equally resistant to different fuels. It can be manufactured and usable in terms of volume. And, of course, one that doesn't leak in use.
So, we started looking for another solution that would be equally resistant to different fuels. It can be manufactured and usable in terms of volume. And, of course, one that doesn't leak in use.
There are sometimes setbacks in planning.
If you look for a 3D CAD model of the Pik28, it is available (at current status) in GrabCAD pages. Link to pages. But only in Rhinoceros v7 format.
Time has gone to everything else. But the plane needs a group of small form parts, wingtips, root fairings, and tips of the tail units. Some aircraft that are offered as drawings contain channels in which these can be bought. But we liked to create a possibility of making them by the builder. But the making of the mold is a monumental amount of work. And as the need is only for one piece, it did not quite inspire. So the idea of printing 3D emerged.
So time to experiment. Firstly traditional model + mold route was tried. But the work needed to smoothen the 3D printed model to the standard that is needed for mold was long. And results were good only to trashbin. That was not the right road. A wild idea was directly print the mold without an intermediate stage was tried. As we have a background of composite production, we tried to smoothen out the printed surface to an expected standard. Did not work out.
But then skipping all finishing stages was tried. And it worked! The work instructions on how it is carried out will be part of the drawings as there are some steps that must be followed. Below are pictures of steps needed to produce wingtip fairing. The printer worked three days, but manual work needed was less than hours.
"Errare humanum est". But pretending that you can not make mistakes is stupid. So, along with the normal design work, an error capturing procedure is started. This is done by building a scale model using the actual 3D model of the aircraft. And as a side product, a printable model is produced, and eventually, it will be published. It is on a scale of 1:8.
Not a measure of merit, but the number of individual documents just passed 132. And the number of sheet 162. You can check the actual number; first, navigate from the main page to "Builders corner (drawings)". Then the bottom link to "Numbering of Drawings". Below GENERAL there is a link to "List of Drawings". It is a libre office spreadsheet (ODS). In the top, cell O1 is the number of documents, and in cell W7 is the number of sheets.
Wing root shape refined. That picture shows magnitude of turbulence at different cross sections. At cruise speed of 220 km/h. As you can see srom section just behind wing, there is no wake. Which is the preferred situation -> less drag.
It has been publicly quiet.
As the philosophy is also to tell why and how the design evolved, a question how to show wing structural design process has caused work.
Eventually a large spreadsheet was constructed, which contains answer to this question.
And will be part of published design data (ATA 5).
As you can imagine, this spreadsheet is not really a simple and using it requires at least keen interest on aeroplane design process.
Otherwise results are not safe.
So if you use it, remember the golden engineering GIGO principle. "gargabe in, garbage out". See also wiki definition.
General meeting of Polyteknikkojen Ilmailukerho ry granted type designator Pik-28 for the aircraft.
Fuselage desing iterated. idea, model, analysis with FEM, conclusions, change layot, return to start and start again. Finalized autumn.
During experimenta winter meeting, aircraft was introduced.
Time has been used to define aircraft basic forces. And these forces are used for strength analysis (design) of the fuselage.
During our woodworking course, aircraft was introduced in public for the first time.
Where to attach safety belts? It must be a strong point, it must tolerate a force on each is at least 1135/230 kp (lap belt / shoulder belt), to be useful during crach landing. When the attachment point for safety belt is strong enough, it would help if the attachment point for safety belt would stay in same package with seat during crash landing. If belts do not follow seat, belts are not really useful. An obvious solution to this is to arrange a sturdy cage, inside which pilot's primary safety equipment are, Meaning that cage has attachment for seat and belts. So the idea of cage is to sustain crash without breaking apart.
For some reason, traditional single curvature surface is arranged so that the straight line is longitudinal. Visible from plywood designed gliders and also in sheet metal designs. Like Cessna 150. This produces easily round fuselage shapes. But air is not flowing up/down direction, but from nose to tail and the traditional single curvature way of doing it makes the streamline not flowing on a smooth line, but a polyline with abrupt change of direction. Other influence is that plywood must be bended to two directions.
Another structural matter is those longitudinal longerons. In traditional way, those longerons must be bent abruptly when they flow from one plywood area to next. The only practical way to make it happen is to chop off longerons and make a split joint to them at corners. Not a very sound way to make structure!
Thinking further about streamlines. All streamlines follow longitudinal direction around fuselage. So surface which has sharp corners along these streamlines is not a sensible solution if thinking of drag. Boats are also designed for low drag, and the basic need for useful space is about the same, a long object, with space for people in the middle. See pictures on this page .
Not many boats are constructed so that plywood seams are vertical. They are arranged to be longitudinal (like the kayak in link above). This makes streamlines to be smooth lines. So why could we also use this arrangement. And it is not strange in aircrafts. See the layout of this PIK-5 glider. And the final shape is not at all funny looking. Actually after first shock, it looks pretty nice. About this time it was obvious that the aircraft was not a small refining of Pik-11. The sensible decision was to change from modification mode to designing a new aircraft. Which takes its inspiration from Pik-11. Board of Polyteknikkojen Ilmailukerho ry was contacted to start procedures for the application of new PIK series aircraft number. The rules for new serial number are known and application is filed.
So, to make room for landing gear attachment, wing will be double tapered. Also to make more room for landing gear, wing profile is for rethinking along with wing redesign. Wing profile is selected to be NLF(1)-0115, which has lower drag and relatively large leading edge radius (for easier manufacturing). With this profile wing tank volumes would be 117 litres.
In Pik-11 landing gear attachment was such that the brackets were the first parts of the fuselage! All structure must be built around these brackets.
It was learned that landing gear of Pik-15 was evolution of Pik-11 solution. But even that was a forged steel sping leg. Needing very specialised subcontractor to make it.
During building of PIK-27 a ready custom built landing gear supplier (Groove) was found as easy to integrate solution. So this much easier solution for builder was selected. This one piece solution requires total redesign of fuselage, but is worth the small weight penalty. It looked that original wing shape takes room just where landing gear would be. Rethinking wing plan form (from single tapered to double tapered) makes room just here. leading edge of root chord moves aft, making integration of this single piece landing gear possible. Dismantling of wings without removing landing gear is high in the wish list. It will be tight fit, but this will be addressed by fine tuning wing root. Also by installing gear legs so that the tapering edge is forward moves gear attachment a bit forward. For gear attachment hardware a rectangular steel beam is positioned on outer edges, below cockpit floor. The front end of this beam will be the lower attachment points for engine mount.
During visit to Aero fairs spring 2017 a company making custom bladder fuel tanks was found. Price for a pair of tanks sounded reasonable. But benefits would be very appealing. In safety, tanks are foam filled, and crashworthiness of rubber tank is better than hard surface tanks. Operational matters (tanks can fill all the space) and it is good for any mixture of gasoline and alcohol, even ethanol. And as tanks are made by professionals, quality of seams is better than any builder can hope for own manufactured tanks. And all fittings for the tanks are made by them.
Drawback is that normal wing structure, frames and skin, can not be used. Inner surface of the space where the tank is must be relatively smooth. After several iterations, a cold laminated plywood structure was selected for this area. Leading edge for the outer part of wing will be normal.
Originally in Pik-11 fuel tank was a cylinder welded from sheet metal, situated in front of instrument panel. Not the safest solution during any crash landing. And the long term leak-proof ness of welded tank is not the best. Depends so much of the quality of welds. So better solution was clearly needed. Looking for ready made solutions resulted only solutions that had to be installed inside fuselage. Or Lockheed Jetstar type blisters on wings. Neither was appealing for team.
Initially a welded tank inside wing leading edge was composed. Cylindrical shape that would fit inside wing leading edge would be quite tiny in volume (total about 40 litre for two wings) and shape would be quite difficult to make. A composite tank felt better solution, even though it need sealing compound to be usefully. Sealing fluids are quite fuel specific and own experience on composite tanks are not the best. Any leak would mean destruction of wing structure. meaning rebuilding wing structure, not something you would like to do.
Fuselage drawings consists of line drawings of the shape and separate drawings of each fuselage frame. By combining these it was possible to guess what was the original intention of shape. Principle of the fuselage was that field borders were vertical. The same principle used in many plywood fuselages used in gliders. Using this prociple it is easy to produce round rear fuselage. See photos of Ka-6 gliders. But to create smooth transition between plywood areas, you must bend plywood to sperical forms. Which plywood does not naturally do.
It turned out that every piece of the surface area was drawn to be spherical shapes! Meaning that following drawings was impossible. And explains those buckling waves visible in every Pik-11 fuselages. Rumours that Pik-11 was difficult to built have solid grounds!
Shape has to be changed to single curvature surfaces, to make it feasible to be made from plywood.
When 3D model of the Pik-11, using drwaings and oral information was startedm the magnitude of the task was rapidly evident. Wing drawing was drawn with diherdal of 4 degrees. But there was a hand correction that it actually must be 5 degrees. Mid section of spars were drawn with less curvature that actually needed. Wing's position in the fuselage was moved and the correct position was documented only in one drawing. Meaning that parts from different drawings were not compatible to each other.
First decision was to correct drawings to such state, that you could actually built Pik-11 using corrected set of drawings.
At this stage team in the project included several persons, so decisions could be made using opinions of the team. As corrections included design work, not just correcting obvious flaws. As the set of drawings was not entirely complete. Some details were not included (were never documented). Luckily on complete Pik-11 aircraft was near, so it was possible to look and measure missing details.
Drawings were available, so copies of Antti and Jouni (Laukkanen) we loaned for scanning. They told that these were probably copied from originals by Pentti (Saaristo). These two sets were not equal, so a wider understanding was received. With these copies in hand and also article in book by Jukka Raunio told same story. There were many modifications in the aircrafts, which are poorly documented or not documented at all, with only oral information available.
So first step was to scan all drawings. No previous digitizing of Pik-11 drawings was known to exist. So if needed they are now available, about 17 GB of data, so they are not online.
Aki Suokas, who would be the chief designer of this aircraft, had built simulator models of ATOL 650LSA and Flynano for X-plane 10 and was thinking what next. Mainly to learn more of this art. He had just flown PIK-11 made by Antti Kääriäinen and found that this is really a fun aircraft to fly. So why not make simulator model of PIK-11.
So first the flight dynamics model. That is what describes how the aircraft feels and performs in simulator. After it was ready it was tested and found by couple of persons that the flight characteristics are close to PIK-11. But the external shape was not presentable.
