Jet Powered air engine 

I think you should research dynamic sealing, one of the way to seal moving pieces without adding friction (the piston for example) is to make multiple grooves (like o-ring groove) along its length, this create sudden pressure increase and decrease as air tries to leak which create very high head loss and stops most of the leak (it still needs a small leak to maintain the head loss, but it's drastically better than just a cylinder in a tube)

For flat sliding surfaces on 3d prints, I usually print the faces with an offset and apply an adhesive-backed PTFE sheet to each side

The alternating filament color to ensure we can clearly see how all the parts work together is truly brilliant and perfectly executed. It's the understated attention to detail you only get from someone like James.

that this sort of valve timing is designed to let pressurized air in at the top of the stroke (both directions) and then the actual power you get isn't from the air pressure of the fans, but rather mostly from the air expanding to normal pressure after the valve has closed. With long strokes, this would require a rather high initial pressure and most of the time your fans are blowing against either closed valves or both sides of the piston (depending on the valve timing, which looks like it may overlap. stacking the fans together to feed a pressure vessel and from there into both cylinders will have the fans blowing less against the closed valves, but you still need to look at the expansion ratio of the cylnders, not just the area. How much volume is there in the cylinder when the valves close vs how much volume is there when the valves hit the other end of the stroke (and how far before the end of the stroke does the valve for the other direction open) what you really want for this sort of low pressure engine is a valve system that will be open for most of the stroke, closing just before the end of the stroke (when the other valve opens) but the valve design you are using is designed to let in just a small volume at the very end of each stroke. The system you are using will let you spin it in either direction, as the valve action is strictly controlled by the position, where what you want is one that's open based on the direction of travel of the valve (or a cam that opens one side for the entire stroke, then opens the other side for the entire stroke)

Measure the static pressure you can generate with a length of tube with water in it by the height change with it off vs on. With this in Pascals and your piston surface area you know the force, then you can do the lever and gear ratio maths to work out the torque at the wheel and propulsive force. You can measure the force required with a spring gauge, or push against a weigh scale. Then you know if it's going to work before you build it.

I’m pretty sure you build the best 3d stuff on RU-vid for how often you upload. Yeah there is cool projects on other channels but your consistency is unmatched

Those edfs have very low static pressure. which results in very low torque. Which is why it's not moving with the tiniest bit of friction. A different type of propeller, or even impellor might work better, like a turbo charger. An other option is to have an air compressor system provide the air, they are good at providing static pressure. And yes you also want as big of a bore as possible to get the most out of the little static pressure you get. then reduce the stroke, and you also don't have to reduce it as much with pulleys. as it will run with lower rpm. Expansion engines like this are all about the pressure, air speed is less relevant, a steam locomotive runs on 200 to 300 psi after all. which is like the opposite of what edfs are designed for.

What if you used a "cone" shaped seal on either side of the piston. When pressure is applied it will expand out to meet the cylinder wall, but when there is no pressure it wont be touching, reducing the friction. Tom Stanton used one in his recent air powered engine.

"You can't suck yourself along with a vacuum cleaner"- james bruton, 2023

I love how you show your failures. I learn a lot and the videos are just as much fun. Plus when you get something right, we can feel confident you're not faking it for the camera. Bravo!

Your honest commentary is refreshing.

Redesign? Sign me up! How about using a fan for each cylinder, but couple the outputs to a single fan powering the exhaust side? It should take some dialing in, but a mild vacuum on the exhaust side should boost performance of both cylinders. With luck, that'll be a bigger gain for power use and weight than doubling the input pressure. Keep it up, James. Perfection's impossible, but improvement is always achievable, you just have to keep finding ways to do it. The trick is to know when to quit until you notice the next possibility. Keep an eye out, then grab them when they pop up. It's simple, but not easy, without practice. Keep encouraging him, y'all.

I HIGHLY recommend you look up player piano wind motors since they use suction to power a rotating axel used to scroll the roll along. They use bellows and external pistons and are quite strong!

What do you do with all the 3d-printed parts after a project is done? I would love to see a series where you build a filament recycling system.

Also, for your valves, instead of making them square, consider spool valves (cylinders). This minimizes the sealing surface/friction creating areas. Picture a smaller inlet/exhaust control cylinder stacked at the end of your larger drive cylinder with two tubes directing air to/from the drive cylinder and to/from the control cylinder. A single drive rod extends from the spool inside the control cylinder, through the drive cylinder and out to the connecting rod. The fan is mounted coaxially on the end of the control cylinder and internal passages ports the output to the control cylinder inlet. The spool valve directs the inlet air to one of two passages that direct the air to either the forward or rear drive piston port. The opposite drive piston port is directed through the alternate passage back to the spool valve, which in this position is directing the air to the exhaust port. When the drive piston reaches the end of its stroke, the spool valve has been repositioned to swap which passage the inlet and exhaust air are directed to, reversing the direction of the drive piston. Phasing is controlled by where the connecting rods are on the crankshaft. Your current design had two cylinders at 0 and 90 degrees. I would love to see a design with three at 0, 120, and 240 degrees. By having the fan, control valve, and drive cylinder all aligned coaxially there should be enough room. By designing the exhaust ports correctly, using manifolds to direct the exhaust, the spools can be mounted on the free ends of the cylinder rods, further reducing friction since they won’t need their own driving rods. Remember the simple per square inch rule of hydraulics/pneumatics. 1 pound per square inch on a 10 square inch surface yields 10 pounds of force. To seal and lubricate your cylinders, consider a light coating of a thick grease such as lard. Or maybe a thin lubricant such as sewing machine oil. I don’t know which would be better. Maybe start with oil, and if that doesn’t work, switch to lard? Edit: just had another thought. Stack belt gears so you can have a 1:x, a 1:1, and an x:1 ratio just by moving your belt. This is the same way the top end of variable speed drill presses work for example. This would allow you to experiment to find the best speed conversion.

acrylic pipes would be a nice touch to see the insides working.

The single cylinder engine also stalls at the end of the stroke which could indicate that the flywheel is too light. It should have the momentum to move the piston across top dead center and activate the switch valve completely. Hope you can implement this in a later version of this. Thanks for the content, James!

You could make the fans pull the air from the engine and out the back so it sucks in the air powers the engine and the air thrown out the back gives you extra force

One giant tube is certainly worth a shot. Maybe you could use PVC pipe and it'll take less time to put together or print parts for it. Try using a heat gun to reshape some PVC as well.

Your projects are great, always appreciate it. Entertaining and informative

"Suck yourself along". Pure gold.

Calipers and something like a lathe would allow for much more consistent tolerances. I say "like" a lathe because you could probably build a jig that would be quite sufficient for working with 3D print materials and the sanding tools that you already have. The key is carefully controlling the movement of the part relative to the sander, holding it at a fixed distance, measuring, then advancing. Applying pressure to a freely moving part by hand is quite variable and not precise.

Adding grooves to the piston will reduce blow by. When the air tumbles in it forms a cushion of air. More grooves = more better.

Awesome sanding music in this one!

I wonder how a similar mechanism would do with hydraulics vs air, since fluid isn't compressible. You'd obviously need gear pumps vs fans. A Stirling Cycle would be neat too! Thanks for sharing James!

I fucking love doohickeys and thingamajigs and this channel never runs out of new ones

Might I suggest Tom Stanton's air valve seal designs to maybe help in sealing the air leaks without adding a lot more friction? A sort of flared edge of rubber or tpu that "breathes" with the motion of movement. Flaring out when the pressure is behind it sealing it in, moving it forward, and when it reaches its peak and the air escapes, another flare on the other side would expand and do the same until it makes it to the other side peak. His videos do a lot better explaining it all, but the concept would work well here.

Oh! i love Matthias! I've been friends with him since WAY back.

James I love your projects their always interesting. has a mechanical fitter on Automobile's. you always have a great way of explaining principles. On a side note, the vacuum cleaner you showed at the start of the video. my farther use for his sand blasting cabinet extractor. (my mothers old one that still works, she wanted a new one).

Really looking forward to the large cylinders.

Super neat project

5:21 In cryocoolers that need to seal high PSI helium use either a regular Buna oRings or Spring oRing fit into a groove cut into the displacer (piston) and then glide rings are placed over those. If you offset them correctly the glide rings protrude out enough that the oRing is forced to compress and the whole system seals. But you don't need five 9 gas so you could just cut a Grove in your piston, put an oring in there, grease it up and that should seal

James I believe your cylinders may be large enough after all, but you need more gear reduction. This will lower the force and increase the airflow in the cylinders. But perhaps you’ll have to balance your camshaft before you go to higher rpms.

Have you ever tried PTFE Film tape to line the spaces between moving parts? I've found that it makes it nearly frictionless.

Awesome engine. Instead of an EDF, try a turbo charger, so you get more air pressure instead of air flow/speed. Then it should be powerful enough for you to ride it, even with the small cylinders.

I think oil would definitely help as well as an increased surface area on the piston. Also your sliding rod going through the piston is a factor in the surface area

Hey James. Are you sure one of the reasons it’s slowing down as soon as you stand on it isn’t because you’re putting one foot and therefore pressure on the cylinder holder? It does look like you’re bending it a bit. Oh yeah, and look into dynamic sealing too!

Such a cool machine!

Edfs move air, but they don't actually create pressure. Same reason you can't just strap one on the intake of your engine and make boost. You need an impeller not a fan.

I’ve been screaming at my screen saying “stop standing on the cylinders” because it’s most likely bending the cylinders just enough to cause the pistons to bind up against the housing

This feels like a problem BEGGING for a Tom Stanton collaboration!

A good rule of thumb in most gas engines is to aim for a "square bore" or equal bore diameter and stroke. Might work out for this situation too?

This is exactly why I have legos. Make your own vacuum engines. So much fun.

thick oil might help seal the piston when moving. Also, you could just look around for a smooth seal so that it doesn't add much friction. The world of 'seals' is literally endless.

The flywheel also needs a counterweight to cancel the reaction to the piston flying back and forth. Look at any steam train wheel and you'll see that it has a huge weight attached opposite the crank arm. And is there any way you can lubricate the cylinder? I feel like if you smear enough grease on it, that might even close up some of the leaks too.

Add at least bearing grease into the cylinder chamber, it would improve a lot, then ruber seals if wanted for high pressure.

Thanks James another fun video :) Have you considered a little counter weight on your fly wheels to smooth out the rotation?

If you want high pressure, you want to use centrifugal or cross flow fans - EDFs are not really designed for pressure and are pretty inefficient in static mode (since they normally operate in aircraft travelling at speed).

Single crank compound: one set of pistons, but instead of the piston going straight into the cylinder it splits off into a top cylinder and a bottom cylinder.

Have you looked into labyrinth seals. They are used to greatly reduce leakage on parts where friction or hazardous environments make other types of seals impractical. They use their shape and tight but tolerance to reduce flow not stop flow.

Bro, I haven't even started the video yet and I'm laughing. My brother even asked me like WTF am I watching... The opening line killed us both 😂😂😂

You could find plenty of information online about steam engines that would help, as it is exactly the same concept as an air engine. You could use valve gear (aren't any gears involved, fyi) to vary speed and compound the cylinders. Additionally, the valve (Controlling the admission of fluid to the cylinders) are typically cylindrical to my knowledge and attached to the pistons. You could also maybe include lubricators, (maybe use a kitchen oil? I'm not entirely sure about how well that would lubricate but it may be worth testing). The best type of steam engines to research for this would be marine or vertical cylinder steam engines.

Simply superb!

"I thought the exhaust of a low compression cylinder, would be louder than TWO 80 AMP JET FANS!" -This Guy

Complimenti..un gran bel progetto..

You need to make the volume of the sylinders so that the airflow per second is the same as the amount of air that the fans supply.

Perhaps put a counterweight on the flywheel to offset the weight that is off-centre where the connecting rods and pistons are mounted. That's what is causing the back and forth vibration and will be wasting a considerable amount of power too.

Getting flash backs to covid times where I spent my spare time outdoors sanding 3d prints...

use a spray bottle of rubbing alcohol as coolant when sanding pla. helps a TON!

A few CM more radius could improve your cylinders - however, you'd be best to measure the static pressure (above ambient) created by your fans and to calculate the force this exerts on your cylinders. I think your static pressure is just very low overall. Sadly though, to get chuff sounds, you'd really need a contained pressure vessel that quietly produces your pressure - those ducted fans will always overpower the cylinder exhaust sounds. Next up, Walshaerts?

- A single stack bladed fan does not produce much thrust, because the air behind it is spinning in a spiral tangentially with the blade rotation. If airflow is blocked the outflow will statically sit there spinning tangentially and not do much of anything. - This is why the noise of a vacuum cleaner becomes higher pitch when airflow is blocked. It takes much less energy for air to sit there spinning tangentially with the turbine blades, than blowing through the pipes. - Stacking a second fan directly behind the first won't do much because the exhaust air is already rotating and the second fan won't add much to the air pressure. - You need fixed stator blades behind the spinning blades to increase thrust. Flat blades facing the other direction will work, but ideally they should be cupped like a Pringles chip, so that air exiting the stator is rotating in the opposite direction of the spinning blades. The gap between the rotor and stator should be very small, millimeters. - Using a stator behind the rotor will also then allow a second stacked spinning rotor and second stator to produce more combined thrust. - The spinning vortex projects out into space in front of the turbine so a stator at the inlet can further arrest the spinning air and generate more thrust.

efd's are generally more suited for high air flows, there's probably a lot of back pressure just escaping back out. You could try a blower style fan that could possibly make higher pressure with same airspeed. Im also not sure if stacking the fans will necessarily help, but give it a try cause who knows

A collab with Tom Stanton would probably be a good idea. He's probably the best expert at air cylinder engines around

You can 3D print an O ring groove and use different thickness of O ring to tolerance the gap, then a little light oil to lubricate and it'll be super low friction AND low leakage.

Love your videos!!!!

Rather than dry fitting the 3d printed parts it might be good to lubricate them. In the past I have used beeswax very effectively. The nice thing about beeswax is that it both fills the small gaps and forms a somewhat hard lubricating layer. Should make the fitting go a bit smoother...

When working a University, a Professor started a design inquiry: make a device that can use vacuum to suck bugs off a distant (3-10 meters) wall. Not all the bugs, just the one you want! She approved $250 and 3 days labor for us to come up with a working design. It took me a full day to explain to her that it was not insubordination to explain how air pressure works, and another two days convincing her we weren't stealing her 'genius idea'. BTW, her specialty was psychology and human memory systems.

You basically built the equivalent of a steam engine cylinder. That's not a bust tho, I like it!

Lack of pressure you should be running at least 3.5 bar to get good results. Someone already mentioned dynamic seals so I won't mention it again.

Excellent! I noticed the base bending a little bit when you stood on it, would that be enough to add more friction in the cylinders?

Would love to see you do a project with Project Air!

Maybe using some sort of air pump to compress the air would be better than the EDF

Bro how many sponsors does one need? Bro dont even need to work anymore to live

Wouldn't this be a 1 stroke engine, not a 1/2 stroke? I thought a 2 stroke gets power every time a cylinder goes down. On this one the cylinders get power going both up and down. Please correct me if I'm mistaken.

you inspire me.

I highly recommend a white lithium based grease on your parts to further reduce friction

Idk ifbyou put any lubricant on anything but a thicker lube would help with friction and might help seal up some leaks

0:24 Peter: I've got an idea!

You could ask the king of air-engines on YT, Tom Stanton. If anyone should know how to optimise a 3D printed air engine it would be him

It sounds like there is a lot of air bypass on the cyliders. you could use some heavy weight oil to seal around the piston and the rods. that will decrease the leakage without changing teh friction.

As you know the input energy could you measure the output energy and work out the efficiency. What power are you getting out of it.

working with a vaccum seems fine, but when pushing air into the system with EDF's, what are you doing about the exhaust? all that extra pressure must be slowing the thing down if you have nowhere for the pressure to escape. Also, maybe smaller valve openings could help? seems like a lot of time is spent with equal pressure on both sides of the piston as the valve traverses from one side to the other.

I'm not sure how it would work for 3D prints, but the steam-era way of getting good surface sealing on sliding joints is to lap the parts together. I've done it with brass, and with a fine modern lapping compound you can get near gas tight seals with surprisingly low friction.

Nice work. Ever thought of playing with shunt valves?

Try adding some silicone oil to the moving parts and cutting or reprinting the piston with 2 groves so you can put an O-ring on it also the piston can be 0.5mm smallerto will help to stop binding

Hi, after watching many model steam engine repair videos, the slide valve timing is critical for peak performance. Cheers

One trick if you were working +ve pressure is steam engines use the pressure to push the valve against the cylinder/intake/exhaust - effectively self sealing.

Great video.......I Love NEW TECHNOLOGY .......😍

hold up. Is that metal bar on the smaller wheel connecting to basically a switch. As in, when that metal bar moves, it changes the airflow from one side of the cylinder to the other side of the cylinder? If that's the case, then because of the smaller wheel, to the pistons wheel, it seems like you're trying to start off in 5th gear. I could be wrong, but that was my hunch from the instant I seen how you were (I assume... I'm loaded) switching air flow from one side to the other.

You really needs a counterbalance like crankshaft. That'll help with efficiency

Do you recycle the printed parts for these projects? I'm just curious how much waste some of these designs produce.

Try adding a slight nozzle to the end of the edf to create a big more pressure

A second crankshaft on the opposite side of the chamber would make use of both the push and pull as it moves.

I love Your Ideas and they are genius, How about new turbo air engine with Vacum half and Runk for ground, so it could vacum easy, please read my comments ❤

paste wax is your friend! try rubbing the moving parts with paste wax, it will help the parts slide and should help seal up leaks.

more pistons on a single push rod , or have the pistons in 180 degree faze (so they push and pull together, if they get stuck you might need a bigger flywheel) , two or more pistons are going to have more Surface area but only when they work together

How did you do that accelerated time lapse part?!?! It's a great effect! (The sanding of the yellow part by hand)

Thumbs up for the groovy music alone! :-)

Have you considered using lithium grease on the internal cylinder pieces, to help against friction and help seal?

edfs are more optimised for volumetric throughput than for pressure. centrifugal fans are best for pressure.