A Stuntmodel with Retracts
A 20 Year Dream
by Uwe Kehnen
I have often thought about stunt models with retracts. At the 1982 World Championships in Sweden, I saw for the first time a stuntplane with retracts. It belonged to one of the Chinese pilots. It was difficult at that time to communicate with them, but I think the system was operated by a timer.
My first attempt for a stuntmodel with retracts was in 1984. I started with a self-built pneumatic gear controlled by engine pressure. The gear worked fine, but the controller did not.
Following that first experiment, a good friend developed for me a simple electronic circuit which was able to change polarity at a certain temperature measured by a sensor mounted in the exhaust. That means a connected electric motor would change directions at certain temperatures. In other words: Hot sensor means the engine is running so the gear is to be up; Cold ensor means the engine has stopped running, so the gear is to be down. The circuitry could be adjusted for appropriate motor direction at the desired temperatures. In front of the circuit was a time-delay-module, so that after switching it on, there was an adjustable time delay before the sensor was active and gave current to the circuit to initiate the retraction sequence. The whole thing was connected to a servo without internal electronics (just the motor and the gear) to operate the retract mechanism. The next problem to solve was the interruption of the current, when the gear was in the all up or down position. I solved this with several microswitches. The cable bundle going from the wing to the fuselage was quite large, somewhat like a finger.
The system worked fine, but sometimes, the gear would collapse during a landing. The gear itself, which was the only commercial part in the system, was not strong enough. My development then stopped for several years when I had to begin to earn a living.
The September 1994 issue of Flying Models had articles by Bob Whitely and Joe Utasi (Stunt News May/June 2007) about an electronic system for a retractable gear in a control line model. This concept and design was perfect for me. That system was based on a switch operated at model release to provide a time delay allowing the model to lift off, thereafter the circuit will be hot. A small microphone senses the noise from the engine. If it is loud -- the engine is running and the gear will go up; if it is quiet -- the engine is not running and the gear will come down. A normal servo can be connected and the large cable bundle would not be required.
I ordered the electronics at Jomar Elektronic, built a unit on a test stand and was very happy with the results. Generating noise for my test runs encountered problems. First I used an electric organ until the family suffered from sleep disturbances. I then used a walkman with extremly loud rock music with earphones directly on the microphone. In the later building phase I used a servo tester to make test runs. It must be very noisy so that the microphone will sense the engine is running. Mounted on a test stand, the gear retracted when it was as close as 1.5m to a team race engine running at full power.
At that time, I again stopped working on the project for several years. Competition flying, training, building houses, occupation and family (kids are only young once) simply left no time for uncertain experiments. I was with my son Felix on the national team at the 2004 World Championships, my largest success up to that time. After the those Championships, I chose to take competition flying a little bit less serious and to dare an experiment.
I decided to build a new wing with a retractable landing gear for one of my existing Impact fuselages. I wanted to accommodate the whole technology in the wing and to have no wiring between wing and fuselage assemblies.
It was then time to search for parts. It should be easy to order mass production parts. In addition, the increase in weight should be within tolerable limits. For the single servo necessary, I used a special retract-servo from Hi-Tec # HS-75 BB (which is called a 180° servo). This servo has the advantage that it is very sturdy and only functions by driving from one end position to the other. I first thought to use two small servos to avoid complex mechanical problems, but on my test stand, the two servo approach did not look very satisfactory. There are strong forces on each gear leg during their retract and extension sequence that need to be accommodated by each respective servo. The one-servo solution allows the forces to compensate each other.
For the power supply, I found 330 mAh NimH cells with only 6g weight per cell. Four of these cells were soldered together to an Akkupack of 4.8V. (If you have no experience in soldering cells ask someone to do this work for you as you can destroy these things easily). I soldered connectors to the battery pack (servo connectors are strong enough). I covered them with shrink tubing to increase their strength. I ordered the electronics at Jomar Electronic. They are not very fast, so allow a few weeks for shipping. If you have the FM Sep 94 issue, maybe you can build it by yourself. It might be difficult to get the right microphone. The one listed in the article is out of production.
There are a lot of retracts on the market. The lightest I found was the Robart 600. There are some heavier units available that look stronger, but I do not know if the extra weight and strength is necessary. I used a mini- switch from the electronics shop. Use a water resistant version if you can find one since contests take place in all weather conditions.
The retracts I used are approximately 15-20g heavier than with a normal torsion wireleg. The battery and the switch can fly for no additional weight when used as tip weight. Electronics will weigh approximately 10-15g, depending if you install it with or without a case. For cables, I would estimate another 20 g. The mechanical linkage will have some additional weight. In the ideal case, the increase in weight could be 60-100g (2-3.5 ounces). My first attempt was over 200g (7 ounces). I consider this a rough prototype. On the next wing I will change a few things and will be able to reduce the weight to the level estimated above.
Now to the installation:
I selected the same point for the gear as with the normal version. This was a big error which I noticed much too late. At that location, there was no place between the wheels and the fuselage and I had to build some complicated and heavy linkages and bellcranks. On the next model, I will place the retracts more outboard so that only two small wires will be needed from the servo to operate the system. The servo will be in the outboard wing, between the wheel and the fuselage.
The gear ribs were converted as follows. Make a simple 1:1 drawing to show where the wheels will be located when retracted so you can determine the mounting angles for the retract mechanism. For the gear mount, I used a U-shaped piece of plywood, doubled with thin maple bearers to carry the gear. The ribs holding the gear mounts were reinforced with 1 mm plywood, ending just behind the spars.
The wing is built in the normal way. The gear ribs carry the U-shaped gear bearer. Then, you can test fit the retracts. At the points where the retracts rest in flight, you will have to remove the lower part of the ribs and build in a wheel bay.
Now some thoughts regarding wheel fixing:
Before bending the wire I made a 0.5mm deep and 0.3mm wide notch into the wire, so that a small spring collar can be used to retain the wheel. You have to work very precisely at this point. The advantage of this system is that the wheels do not retract very deep into the wing to allow clearance or the leadout wire. You can leave the bellcrank in the middle of root ribs.
After mounting the wheels, you can build the wheel bay. I use home built carbon parts, but something built up from wood will have the same effect. The parts were built in and framed with balsa. You can then contour the lower rib parts to match the other ribs. Before sheeting, you have to build in the servo.
The electronic parts are in the space behind the Servo. Remember to build in removable covers over the servo, electronics and battery. Then you can build in the battery and the switch in the wingtip. Connect everything and test it. If you are satisfied you can build the rest of the wing.
The strut covers were built as follows:
Cut four pieces of carbon mat that are oversize to fit over the wheel wells. Saturate two of these pieces with epoxy on a non-sticking piece of foil (The foil of the backside of iron-on foil is perfect). On the top of the carbon package, lay another piece of nonsticking foil. Lay this foil-wrapped carbon package on top of the wing over the gear bay area. Do the same thing on the other side. After curing, you can remove the laminated carbon gear doors from the foil packages and trim to final shape to match the gear bays. Then mount the right top to the left gear and vice versa. The mounting is pretty simple. I solder thin metal straps on the gear leg to provide more glueing surface. I use the same metal used to build fuel tanks. Now you have to retract the gear and put some large spots of good silicone glue on the gear leg and the metal sheet. Fill the gaps betwen leg and cover with the silicone glue. To make it bulletproof, I secured the covers with a thin wire to the legs. The result is a nearly perfect covered wheelbay with true wing contours and shapes over the wheel wells.
Before the first flight I recommend extensive testing on the ground:
Switch to off, start the engine, lift the model and then switch to on. Depending in the setting, the gear should retract after 5 to 15 seconds. Now, stop the engine and the gear should extend without any delay. I set the time delay to maximum so I can have more time when I have a helper who is not familiar with the system.
Always remember: Activate the switch at the release moment and not earlier.
You will need to compensate for a noseup moment during the retraction sequence. There will be a more equal feeling with the inside and outside loops. I have the feeling that the model is less sensitive in turbulent air. At the present time, there is not much more to say because my current model is too heavy.
I think the biggest effect is the show and imnpression. The only problem I have encountered is that the gear has occasionally extended twice two times during flight. This was most probably caused by a bad soldering point. In the first year of operating this syste, I never experienced a gear-up landing. Unfortunately, at the beginning of last year, I did experience a ger-up landing. Some judges are now waiting for it. I hope I can make them wait for a long time before any gear up contest landing.
Improvements on the next model (maybe by the end of 2007 or in 2008):
- The weight has to be more equally distributed on the model. I put so much heavy stuff on the outboard wing that I had to compensate with added weight on the inbord wing (about 2oz).
- The retract gear assemblies have to mount more outside to save the complicated mechanics.
- I will build lighter wheelbays from thin composite.
- The most important thing (joke) is to find a way to retract the tailwheel. First it looks ugly and second is it boring to hear many complaints at a contest about the tailwheel. (That is from people I call friends.)
Prices if you like to try it:
Gear 30 -70$
Battery and switch 30$
If you have any questions send me an e-mail or make a thread in the Stuka Stunt Forum and let me know about it.
Thanks a lot to Keith Trostle for the hard work to translate my English.
This and the articles from Bob Whitely and Joe Utasi are also published in Stunt News May/June 2007