3D Printed Undercarriage

Introduction

My mates and I fly up in the hills east of Perth, Western Australia. Its quiet, wide open and mostly void of people. When you do encounter anyone they are invariably friendly and obliging. While our flying site has these advantages it also has a downside. The open area is mostly dirt and gravel with lots of small weeds. The weeds can hide the occasional rock. Our runway is a rough gravel road. Landing and taking off from the gravel road might seem a negative but my ability to line up and land on a narrow runway has improved drastically since I started flying there so I think its a positive. However I don't always keep it on the road and sometimes I miss the road entirely. The road itself is not exactly smooth either. Until recently the solution has always been to pick the type of plane to fly there. I started with belly landers because they don't need a smooth surface. The next step was fixed gear and big wheels with really well enforced piano wire undercarriages.

My other flying buddies have flown models with retracts there but there is no fudge factor. If you miss the road then you will have a model to repair. My landing skills are not up to par with my buddies. I can land a plane on the road but not every time. Its one thing to land into the wind with only a rough idea of touchdown point. Its much harder to land on a 8 foot wide strip, not into the wind and having to touchdown at a particular point. Don't get me wrong. I love flying there and landing on the road has improved my flying skills enormously. However there is a limitation in terms of aircraft types I can fly there and I wanted to do something about that.

About two years ago I purchased a 3D printer. It was the first and last thing I will ever buy on a pre-order as it arrived months late and with lots of problems and issues. It was called the Soliddoodle Press and it is one of the worst printers out there. Prior to the press, Soliddoodle had a good reputation but design issues and production quality issues led to a very poor product which ulitmately sent the company broke. It took me two weeks to figure out how to make it work at all and months to figure out what its limitations were and what I could or could not do to get around them.  However, despite all the problems I found being able to design and print the model plane parts I needed an absolute wonder. Okay my prints were a bit rough and I would spend more time modifying the design for the part so that it would actually print on my printer than I did designing the part in the first place. Never the less I could still produce the bits and pieces I needed for my designs. I have another printer on the way (Flashforge Creator Pro) and I have high hopes I will be able to extend what I can print. Mind you it cost just over twice the price of my original printer.

So back to undercarriages. The one thing I didn't like about the Flitetest Spitfire and my own design Hurricane was the fact that you could not put the trademark radiators on the bottom of the wings because they were both belly landers.

That is when I started to think about designing an undercarriage with suspension built in. If I could design and build an underriage leg with some kind of suspension built in then landing on the road (or missing it entirely) would not result in damage to the model or undercarriage. Given the size of my models (around 1.2m wingspan) there are constraints in terms of size and weight which eliminates the possibility of using ready made undercarriage legs with oleos. I sat down in front of the computer and started playing until I came up with a preliminary design..

As can be seen from the above diagram, each leg consists of an upper and lower part. The upper part is anchored to the wing (or retract system) and the lower part pivots to absorb shock loads. Provided I clould find a balance between how much the lower leg pivots I figured it should work fine. I envisioned using rubber bands wrapped around the leg to oppose the forces trying to pivot the leg away from the vertical. If you could dissipate the landing forces via a suspension system then the undercarriage could take a lot more knocks without breaking.

Thinking about what material to use led me to the idea of 3D printing. A 3D designed and printed part would be light enough and could be easily modified and reprinted based upon how things went. Strength was an issue though.  In terms of the loads on a model plane airframe landing in the rough, plastic just isn't strong enough. Added to that is the possibility of the part splintering between the different layers of the print. For thse not familar with 3D printing, the object is built by layering thin horizontal slices of melted plastic one on top of the other. However, under stress, there is always a tendancy for the object to split along the boundary between two layers. This meant I had to keep that in mind. In a sense it is a bit like allowing for the wood grain direction when working with balsa. The other variable was how solid should the inside of each part be? It could be completely solid and heavier or almost hollow and lighter. How solid an object is printed at is expressed as percentage infill. A 100% infill is a completely solid object. A 50% infill means half the internal volume will be solid and the other be air. Some form of 3D pattern is used to fill the space (I use a honeycomb pattern). The design has to be spot on with regard to distributing the force or it will break instantly. The game being to balance strength against weight. Isn't that always the way with model plane design?

To Retract or not to Retract?

I decided it would make more sense to start with a fixed undercarriage, sort the problems out and then proceed onto a rectractable one. I used my PT-19 as a test bed for the fixed version and my Hurricane as a testbed for the retract version.

In the first attempt the rubber band absorbed the energy but the upper part proved too weak. I was printing with 40% infil so I increased that to 100% (i.e. completely solid) for the upper part.  That solved the strength problem and the suspension worked really well. In fact a bit too well. When the wheels hit something and moved back they moved behind the centre of gravity which caused the plane to instantly nose over.

The next thing I tried was to replace the rubber bands with elastic.  Elastic stretches so far and then won't stretch further which restricts how far back the wheels can move. Problem being the second the elastic stopped stretching the upper strut snapped. I next tried adjusting the angle of the lower undercarriage leg to reduce the distance that the wheels would move back. That also failed and resulted in snapped undercarriage legs. I pondered a while and realised that both modifications failed because, while they prevented the wheels moving back too far, they did so by substantially reducing the amount of landing force that could be absorbed. Finally I hit upon the idea of angling the upper leg forward. That had the effect of reducing the how far back the wheels end up without reducing the ability of the undercarriage to absorb the landing impact forces.

All three legs in the above image are positioned as they would sit against the upper leg. The first two (left and middle) sit against a vertical upper leg and the third (right) sits against an upper leg which has been angled forward. The left hand leg resulted in the wheels being moved too far back. The middle leg is almost verticsl and so its ability to pivot is reduced which meant the amount of impact force which could be abosrbed was reduced by too much. The final version on the right is the one which worked perfectly. One thing to point out is if I was designing a new model I could position the undercarriage as far forward as possible to help things out.

The solution was to angle the upper strut forward and it all worked beautifully.

The above image shows the final undercarriage mounted on the PT-19.  Note that I also had to design and print a mounting point for the upper strut which sat inside the wing. These were a modification of the original mounting points that I designed for the pianol wire undercarriage. Okay I'll confess! I didn't design anything. I simply removed the piano wire and drilled a hole in each mounting point. The hole was the same diameter as the upper leg.  Hot glue isn't very good for gluing ABS plastic but in this case it is perfect. If you land stupidly hard then the glue snaps before the undercarriage leg. Since mounting this undercarriage on the PT-19 I have missed the road completely, ran off the road and stalled in from about 3 feet without damaging the undercarriage. It looks very agricultural and isn't very scale but It was an experiment and it works fine so I am happy to leave it on the plane.

The Retract Version

The next step was to modify the design so that it could be fitted to a standard worm drive retract. Also I had to modify my Hurricane so that it could take retracts.

The above image shows the result. I had to design a plastic insert for the wing so that I could screw the retract unit to it. Modifying the undercarriage legs from the PT-19 was the easiest part. All I had to do was taper the upper leg so that it would fit against the retract unit when the leg is retracted. I used a cheap set of retracts from hobbyking and the 3D printed part is simply glued onto the piano wire leg in the retract unit. I used epoxy and roughened up the piano wire to get the strongest bond possible.

The above image shows the leg in the retracted position. Note I had to bend the piano wire leg forward to maintain the angled forward orientation of the upper leg. Also, because I thinned the upper leg to fit onto the piano wire strut in the retract unit, I wrapped some fine wire around the leg and covered it with epoxy to increase the strength and combat the possibility of the upper leg splintering away from the piano wire in a bad landing. Also note the wheels are not contained withing the wing due to the wimg thickness being too small. I didn't care because it still meant I could mount my radiator and up in the air it wouldn't show that much. Again, if I was designing a model from scratch I could make allowances for this.

The first flight was brilliant and everything worked perfectly. The hurricane took off fine. I got a buzz watching the wheels retract it flew past and the landing was perfect. Straight down the gravel road, flared just right and nice touchdown. The second flight was a destruction test for the undercarriage. I have no idea how or why I let it happen but on approach, at a height of about 10 feet I managed to let the Hurricane get too slow and it stalled and went in. As it dropped I applied elevator and it started to pitch up. Just enough for the undercarriage to get the full impact force of the plane hitting the ground. Cursing myself I went to inspect the carnage. I was very surprised to see the damage was minimal. The piano wire had been bent back quite a way and the legs had come loose from the piano wire. However the 3D printed parts were completely undamaged!

Back at home i straightened the piano wire. This involved dis-assembling each retract and putting the struts in a vice to bend them back and then re-assembling. With it all back together I did exactly the same thing on the next flight! This time there was less bend because this time I stalled in from about 3 feet. In fact only one leg was bent the other was completely fine.

Conclusion

I feel the experiment was a success. If the idiot pilot could manage to not stall his plane in on final then all would be well even on our rough gravel road/runway. The fixed undercarriage on the PT-19 has taken months of abuse and is still perectly fine. I haven't got around to bending the strut back on the Hurricane undercarriage because the thought of taking apart the retract unit again sends my motivation for the task into a death dive. I have enjoyed the problem solving and the experimentation, and of course the buzz of success. I think I will work on my landings a bit more before I put it back in the air. A major point of this is that the journey was as much fun as the final result.  3D printers are going to be a massive boon to everyone in this hobby. At this point they are still a tad rough around the edges and can be an absolute trauma to get to work. However I am old enough to remember the excitement of actually getting a properly printed colour image from an ink jet printer after hours and hours of unclogging jets and mucking around. Since I purchased my 3D printer, the standard and quality has improved at an astronomical rate. I beleive this will continue to the point where 3D printers become cheap and reliable. Even now I would encourage anyone to get one. The main point at the moment is to understand there will be a learning curve and you won't be printing model plane parts on the day your printer arrives. However you will, after some trial and error, get to the point where, designing and printing anything from control horns to cowlings, even machine guns is an every day possibility for you. Who knows maybe even an entire model plane!

Data Files

The PT-19 sketchup and stil files can be found here. I included some wing mounting supports modifed from the original piano wire undercarriage to suit the 3D printed undercarriage. I never used them because I took a drill to the piano wire ones that were already in my PT19 wing.

The Hurricane sketchup and stl files can be found here.

I believe that anyone who knows how to play with sketchup could easily modify either undercarriage to suit different models. Anyway I hope you enjoyed the article and I hope some of you at least try out my undercarriage designs on your models.