Ups and Downs


  • This article was first published in 2005.

Designing a scale model such my Cessna Bobcat is an ongoing process. Generating a set of working plans, enabling the major parts of the airframe (wings, fuselage and tailplane) to be constructed, is just the beginning. As work progresses and the model evolves, the designer’s hat has to be donned as and when – like now, for example, when considering the Bobcat’s engine nacelles.

The first requirement is to make a pair of glass fibre cowls for housing the Laser 80s. The second requirement (and the focus here) is the retracting undercarriage. Most aircraft possess what I describe as ‘90° retraction style’ retracts, where the units fold inwards (as evident on the Mustang, for example).

A number of twin designs employ a different system, where the legs are mounted in the nacelle and retract backwards or forwards. Not the Bobcat, though – its retracts are fairly similar to those of the Avro Anson I made many years ago. Here the main leg had a link portion in the upper section, which hinged on a shaft at its very top. When this shaft was rotated through almost 180° the link would swing forward and then upwards, raising the main leg and wheel. There was also an arm from the wheel axle that went to the rear spar under the wing, which acted as a strut to stop the wheel and leg being pushed backwards. When fully retracted into the nacelle, half the wheel and tyre remained visible to the elements.


Enjoy more RCM&E Magazine reading every month.
Click here to subscribe & save.

Whilst the Bobcat’s wheels also remain half out of the nacelle when retracted, the mechanism for getting them there differs from that used on the Anson. This new model, then, features a rear strut to the main oleo leg, attached to the bottom of the fixed portion. When the leg is retracted it rests at about 45°, the upper section of the leg being housed in the forward section of the engine nacelle.

Unfortunately I haven’t been able to get hold of any useful information on how the retracts of the full-size operate so I’ve had to design a system that produces the right results, given that there’s no commercial unit available. This being the case the basic design requires that a cross block be mounted at the top of the oleo, the former having 6swg piano wire pins protruding from either side. These pins go into square steel blocks, which in turn slide on a pair of 6swg runners. The aforementioned rear strut (running between the bottom of the oleo and the wing underside in the engine nacelle) produces a radius as the top of the leg slides forward and raises the wheel into the nacelle.


So with the design conceived, the time came to turn it into reality! Fortunately, I recently acquired a small metalwork lathe and a milling machine; a Clarke CL300M lathe with a 300mm long bed, and a CMD10 milling machine which, although fairly small, is more than adequate for model work. Both were bought from Machine Mart, costing £422 and £281 respectively (inc. VAT) and both machines have variable speed motors that can be run forward or backwards.

So, where does one start when making a retracting undercarriage? Well, for me the first thing to do was select some lengths of spring, the size of which would determine the sizes of the oleo legs. The outside diameter of the spring I used is about 1/2” whilst the inside diameter is 3/8”; the spring is fairly substantial as the Bobcat is likely to weigh somewhere between 20 – 30 lb and so needs some supporting. When executing what looks an absolute greaser of a landing it’s a dead certainty that you’ll be landing on one leg before the other, which means that the full weight of the model has to be taken on one leg at the initial point of contact with the ground.

Obviously, in a poor landing the model’s going to hit the ground fairly hard and the undercarriage needs to absorb all the forces that are incurred at that moment. What’s less obvious with what looks like a good landing is that there’ll almost certainly be some side effect on the legs, particularly if you’re landing in anything that represents a crosswind. If a model is to be regularly flown from concrete or tarmac then fairly hard springing is used, but if flying off grass soft springing is better as it will better absorb all the lumps and bumps.


Whilst I’d sourced the spring from stock, I didn’t have any steel tube of a size that would provide a decent fit over the spring. My only option therefore was to ‘turn up’ the oleo leg. Whilst such a process involves extra effort it does mean that the tolerances can be made fairly tight, so there’ll be very little slop in the assembly.  Any that is evident would cause the leg to twist and turn, the wheels pointing in various directions, with obvious consequences on take-off and landing! At the bottom of the fixed oleo section is a bracket plate, which has an anchorage point for the drag link and the rear radius strut. This bracket plate (partly turned on the lathe) fits onto the outside of the main tube, allowing the undercarriage leg to go up and down within it. A pair of additional side braces also go into this plate.

With the main parts turned on the lathe the leg was then put into the milling machine, attached via a mandrel and held in a vice so that the portions of metal between the various brackets could be milled away. A time-consuming process as only small light cuts could be taken at a time.


The entire oleo unit and the curved portion of leg that goes around the wheel down to the axle were all made from mild steel; it would have been lighter (and easier to make) had it been machined from aluminium, but this wasn’t an option as aluminium wouldn’t provide the necessary strength, given the size and weight of the model. My hope is that the increase in strength (and reliability, hopefully) should outweigh the extra ounces on the model.

The curved portion of leg that goes around the wheel to the axle was made from 3/8” diameter mild steel rod, bent around a large diameter piece of steel to generate the right radius.  It would be fair to say that this required quite a bit of hammering with a heavy hammer! A pair were produced, then a slot was cut into the bottom of the oleo and the whole lot silver soldered in place. I used a Calor gas blowtorch of quite reasonable size to do this, as the heat dissipation on this amount of metal is quite considerable. Finally, the axle from the wheel was made from 6swg piano wire, once again silver soldered into the lower portion of the leg.

So, how to raise and lower the undercarriage? There are two ways that this can be done, either by high power servos or pneumatics. One of the advantages of servos that use a large gearbox is that they’ll be fairly slow in operation, producing a realistic speed. Some computerised transmitters have the facility to slow a function down which, combined with a powerful servo, could be an option. One of the downsides of using servos on retracts is that if the leg or linkage gets bent it can stall the servo and flatten a battery pack very quickly. For this reason it’s a good idea to use a separate battery to operate retract servos so as to avoid inadvertently draining your receiver pack!

This is all well and good, but the Bobcat u/c demands rather a lot of travel, something not possible with a servo. Enter pneumatics, which have an immediate advantage in that, should the legs get bent and stall the mechanism, the battery won’t get drained because the controller operating the air supply will have reached its full travel and be perfectly happy. The system is fuelled by a large reservoir of high-pressure air that will provide all the retractions you’re likely to require during a flight. There are commercial air rams on the market that provide about 2” of travel, but as I required something like 4” I had to make my own.

First job was to buy some suitable brass tube to make the cylinder. Browsing through the available stock I came across some tube (made by K&S) which was ideal, at just over 5/8” diameter. Partnered by a selection of ‘O’ rings that would be used as seals I then turned up a piston, which would sit on a length of 8swg piano wire. The piston had a very carefully machined groove turned in it at just the right depth to accept an ‘O’ ring, the trick being to get a nice air seal fit between the ring and the brass tube. Making the groove slightly wider than the ‘O’ ring allows the latter to roll slightly when the piston is first moved, which is sufficient to ‘break’ the tightness and free the piston within the bore. (One of the downsides with pneumatic cylinders is that any lubrication will be squeezed out of the way by the ‘O’ ring if they’re left to stand for a long time).

Next, an aluminium block was turned to fit inside the brass tube cylinder at one end, with a hole drilled through the protruding piece of the block to act as a support. A 6BA hole was also drilled for an air nipple, screwed and epoxy glued in place (the air nipple is a 6BA brass bolt with a 1mm diameter hole drilled through its length). This block was then glued into the end of the brass tube to make it airtight. The opposite end of the cylinder (where the actuating rod would slide in and out) needed a second aluminium block to be turned. A hole was drilled through this to accept the 8swg piano wire rod, and a second hole counterbored on the inner face to accept an ‘O’ ring and a small bush to retain it. An air nipple was also added as the piston in this design is ‘push-push’; this is different to the system used by Spring Air, which only requires air to be applied to one end of the piston. The whole unit was then assembled and lubricated with Vaseline prior to the second end block being epoxy glued in place.

As the air supply can be pumped up to 80 or 100 psi there’ll be plenty of power from this piston, for sure! I’ll be using a lemonade bottle to act as the air reservoir, which, I’m reliably informed, can be pressurised to around 250 psi before the top unscrews! Air line restrictors will be fitted to slow the speed of operation, making the gear up / gear down process look far more realistic. Another advantage of pneumatics is that you only require a small, cheap servo to operate the air valve that will control the retracts. Robart make a selection of quick-release air line connectors for connecting the air pipes from the fuselage to those at the wing root. These will need connecting / disconnecting along with the aileron servo leads etc. when the model is assembled or dismantled.

Whilst the system described above is still evolving somewhat, I’m now at the stage of building it into the engine nacelle so it may yet require a little bit of modifying or changing to get it to work reliably.


Article Tags:

About the Author