Flying a Plane Powered by Thin Air

When I was a kid I had one of these air hogs planes that was powered by a small plastic engine. This engine used compressed air as its fuel and I remember pumping it up to high pressure with the small included pump before throwing it into the sky and never seeing it again. And ever since then I've wanted to build my own version but a little larger. This video is sponsored by Opera. So I started 3D printing some lightweight aerofoil ribs in ABS plastic that will make up the internal structure of a wing. I could have used other materials like balsa wood or carbon fibre but those are more difficult to manufacture and wouldn't save a significant amount of weight as these ribs are very thin. Once glued to the carbon fibre spar I glued a carbon rod to the leading edge and a carbon strip to the trailing edge as the wing will be covered in a lightweight film so these will help hold the aerofoil shape. Now I've never covered a wing like this before so I had a quick read of the instructions. This stuff is very thin and delicate so it creases easily but on the underside it has a heat set adhesive which can be ironed onto the wing structure which is why they're printed from ABS due to its high melting point. Once ironed onto the ribs a quick blast with a heat gun will shrink the film down to create a smooth wing surface. But there are some issues. The first being the ribs at the wing tip bend inwards due to the tension in the covering film and this tension also causes the aerofoil to shrink between the wing ribs making an inconsistent profile along the wing span. Plus I may have held the heat gun in one position for a little too long causing the covering film to melt. I just want to talk about wing design quickly as this is the fourth air-powered plane I've attempted to build and fly. The first plane I built was the heaviest but used a very similar wing to what I'm building now though it was constructed from foam board and despite the plane using a very inefficient engine it actually holds my personal best flight time record of 16 and a half seconds. Whereas my last air-powered plane which had a far more efficient engine and was much lighter only flew for eight seconds which is partly due to it hitting a fence but if the wing had a high lift to drag ratio it may have cleared the fence. I've been trying to make these planes lighter and lighter whilst trying to make the engines more and more powerful but what I haven't been looking at is the drag caused by the aircraft. The last wing I built was very light but also had a lot of drag and because the thrust has to oppose the drag the very little thrust produced by the engines struggled to get the airplane up to speed. So this new wing design uses a known aerofoil profile which with the right airspeed and angle of attack can achieve a lift to drag ratio of 60 to 1 so a little extra weight won't be much of an issue and if you want to design a custom wing like this with your own aerofoils I highly recommend checking out Onshape. I simply import the aerofoil profile from aerofoiltools.com then with a few added design features it's basically ready to print. These ribs are the same colour as my t-shirt. To fix the issues caused by the covering film I decided to use a thin sheet of balsam wood to reinforce the leading edge. This stuff bends really easily and if you take it outside in the British weather it'll bend even easier. So much so that I could bend it around a 3D printed mould of the wing and pin it in place until it dries. Once dry it holds the curved shape of the leading edge and I can print some new ribs with a small step where the balsa will mount to. I then gave the wood a few coats of bright orange paint to make it look nice which gave me some extra time to build a new wing structure. This new wing is similar to the previous one but has much thinner ribs as I'm hoping the balsa wood will add some strength as well as some thin carbon rods to prevent the wing tips bending inwards. Once the paint was dry I glued the balsa wood onto the wing structure with some thin CA glue and the wing was looking really nice with the orange leading edge and grey ribs. Then the covering film can be applied to check how the aerofoil profile turns out which looks much better. Also with the thinner 3D printed ribs this wing is actually lighter than the first wing and with two of these wings making up a two meter wingspan it's over three and a half times larger than the air hogs plane so we need a powerful engine to fly it. Fortunately I've been developing 3D printed air engines for a while now and have tried many different designs such as spring operated inlet valves, cam and pushrod valves, diaphragm actuated pistons and the latest design that uses a custom moulded piston seal that expands under pressure to make the piston airtight which is very similar to the air hogs engine and this engine has been a huge step up in performance compared to my older designs producing 350 grams of thrust from an engine that only weighs 40 grams. The engine is very simple with only a few moving parts. High pressure air flows in through the top of the engine and is sealed by this ball bearing then as the engine turns a small pin on top of the piston moves the ball upwards opening the valve and letting high pressure air fill the cylinder but this seal on top of the piston is a vital component to the engine's efficiency as it expands due to the pressure and seals against the cylinder walls. This causes the pressure to push the piston down closing the ball valve and turning the crankshaft then as the piston reaches the bottom the air flows out the exhaust ports and the seal contracts removing any compression as the piston travels back up. Now the airworthy version of this engine is slightly different but works exactly the same it just has this 3D printed thread to attach to a plastic drinks bottle and the air flows up this thin tube into the cylinder head. I also decided to buy a high quality propeller to hopefully maximise the thrust output and surely the fact that this propeller comes with its own socks must mean it's good. I know many of you are thinking why don't I use CO2 cartridges instead of plastic bottles as they can store far more pressure. Well this CO2 cartridge is pressurised to about 800 psi or about 54 times atmospheric pressure which is a lot but its volume is very small at just 21 millilitres so if we let the CO2 gas flow out of the cartridge to equalise at 1 atmosphere of pressure we'd have just over a litre of gas but if we pressurise a 2 litre drinks bottle to 120 psi or about 8 times atmospheric pressure we can store 16 litres of air. This means we'd need 14 CO2 cartridges to equalise the storage capacity of the drinks bottle and considering the single cartridge weighs more than the bottle the air storage to weight ratio is over 14 times greater for the drinks bottle. So moving to CO2 cartridges would be a huge downgrade but you know what isn't a downgrade? Trying out the Opera browser. Opera 1 is the latest version of Opera desktop browser and it really helps me with my workflow when creating these projects for YouTube. Like whilst researching and designing a project my browser tabs can really stack up but Opera has these tab islands so I can arrange the tabs into groups like here's the island full of tabs for researching wing aerofoils and here's the island full of product tabs for purchasing the parts required to build it. Also with the built-in ARIA AI finding an answer to anything has been made far easier. For example if you read something that you need more information about simply highlight it to bring up the ARIA AI prompt then select between explain briefly, explore or translate depending on what you need to know. Personally I love using the explain briefly option as I come across a lot of technical words whilst researching my projects and this feature saves me a lot of time and it also has a command line where you can ask it various questions which I find really useful if I need a calculation done fast. With many other features such as the free built-in VPN, ad blocker and messenger apps I highly recommend you try it by using my link below to download Opera today and thanks to Opera for sponsoring this video. So to increase the runtime I'm going to mount two 2 litre bottles back to back using a nylon bolt. This requires drilling out the underside of the bottles which is the thickest part and shouldn't reduce its strength by much as well as hollowing out the nylon bolt but in order to attach the bottles together I need to get the nylon bolt inside of the bottle which requires cutting up an old arrow shaft and attaching a 3D printed socket which perfectly fits the bolt and just squeezes through the bottleneck. This allows me to sandwich a large rubber washer between each bottle before tightening it down with a nylon nut which should hopefully create an airtight seal but before we test it for leaks we need a way to pressurize it. The design I've come up with is a 3D printed bottle cap that has a small ball valve inside and to fill the bottle with air I can use this hose with a 3D printed latch attached to it. This hose fits perfectly through the o-ring in the ball valve and the latch locks into the bolts with a small turn. Once the plane is up to pressure I simply twist the latch, pull the hose out and the ball should seal the air inside. So I filled the bottles with water to reduce the potential energy if something were to go wrong and pumped them up to 120 psi and both the nylon bolt and the 3D printed filler cap held up without a leak. So it's time to build the tail of the aircraft. For this I chose to use balsa wood sheets but this 4mm stuff is a little thick so I ordered some 3mm sheets that were, believe it or not, shipped in a paper bag. So let's reduce the weight of the thicker sheets by drilling some holes. First with a small drill bit and then with the large countersink bit to reduce splintering and then finally with a 3D printed bit wrapped in sandpaper to make the holes nice and smooth. This reduced the weight by about 20% and after a few coats of paint it can be laminated with the covering film to stop air flowing through the holes because leaving a bunch of large holes drilled in the tail of an aircraft would be a bad idea. I then added control horns and attached the control surfaces with packing tape as it's lightweight, flexible and sticks well to the covering film. Now I want this tail to be controlled via cables to keep weight away from the rear of the plane so I used some spring steel wire as a torsion spring within the control surface. This causes the surface to be sprung loaded in one direction which should keep the control cables in constant tension. Now you may notice there's only two tail fins here instead of a separate elevator and rudder and that's because I've chosen to go with a V-tail design. I then printed this odd looking part that the tail boom will mount to and I can attach servos here to control the tail which works really well once the cables are attached. This central hub also mounts to the bottles with a flexible printed ring and a small bracket on the tail boom to complete the plane's fuselage. Now all it needs is the wings to be attached to these points and we have a completed plane. Let's take it out for a test flight. Right we're at 60 psi but I can hear a leak out of the rear bottle cap. It's not a good sign because I've pressure tested this already. Okay quickly release that. Here goes nothing. Check the controls and start the engine. That's brilliant and can I land it? Can I land it? It worked! It worked! It looks like the propellers snapped off though. Let's go and inspect. Right so the crankshaft snapped but that is definitely the longest air-powered plane flight I've ever achieved. So now that the record flight time for a 3D printed air-powered plane has been increased to 31 seconds, I took to Instagram to ask how long you all reckon it will fly for and the majority of you chose more than a minute so I can't give up just yet. To increase the flight time I fixed the leak in the bottle cap by reprinting it with a larger o-ring and it was ready for another test. Don't you know pump it up? You've got to pump it up. Don't you know pump it up? And 100. One for good luck. Please run properly. Oh the battery's come off. Come on we can climb, we can climb. Come on. Yes! Oh that's so much better. It's still going like really well. Oh my god it's working. Mate it's still doing well. It's nuts it's air-powered. I know what you meant about yaw and roll though. Yeah. Mate it's still climbing, the wind's coming from our right isn't it? It's coming from my back I think. Okay I'm trying to turn it, it's coming down. The battery fell out, can you see it hanging out the side? And I've stolen it. Oh no I've stolen it. Oh we have to do another one. For the next flight I decided to add a small onboard camera but due to my excitement of the previous flight whilst tightening a bolt I accidentally pierced a hole in the bottle. You see that? It's just hissing, just hissing out of that hole. Fortunately I had some spare bottles so it was an easy fix. Though throughout this period of adding the onboard camera and bottle replacement I forgot to do one last vital check and that was whether I pushed the tail boom back into its mount properly because this means the control cables are too short causing a lot of up-elevator trim. Three, two, one. This may look like the end of this plane but after a quick depressurization I was surprised to see how little damage there was. The only damage apart from this 3D printed bit here as far as I can see this wing's fine just this wing here which a little bit of packing tape that'll fix that. So I fixed the wing with a small patch of tape and reprinted a new central hub which I honestly reckon saved the plane when it snapped and by the following day it was back up and flying. Though I've yet to beat the 1 minute 22 second flight record again mostly because it really struggles in the wind. I have had a few flights over a minute but I think it needs some weight shaved off before I get near the two minute mark. Even if I can reduce the overall weight of the plane by about 10% that will be enough to add another bottle which will help extend the flight time. Thanks for watching!