Magnetic Air Engine 

You can get all the 3D models (Native Files, STEP Files and STLs) here:integza.com/collections/3d-models

Tom Stanton has been doing a ton of work on refining this kind of air-powered engine design over on his channel, you guys should collaborate! Tomatoes are disgusting!

It's Interesting how I never thought of that but it kinda makes sense if you think about it. It would be interesting to see this design taken further and being used on a vehicle (small car, maybe even RC :-)

Some thoughts for improvements... 1. Reduce the volume of air after the "ignition valve" (the one providing air to the cylinder). As it currently is, the pressure shoots up when the valve is open - compressing the entire volume not just the cylinder. It seemed like it was a sort of loopy hose. If the volume there is high enough, it won't be able to fully depressurize when the piston is at bottom and the exhaust ports are open. If it doesn't fully depressurize, then the piston will have to fight the remaining pressure in order to get back to the top of the cylinder. 2. Rather than use a second valve, consider just using a pressure regulator between your high pressure source and the bottle. The bottle can be a smaller one; but the opening should be fairly open -- otherwise you could just use a CO2 canister that you previously emptied. By using a pressure regulator, you won't need a controller. You could then have a source bank of a few of the CO2 canisters, the regulator, a small bottle, the ignition valve and then the piston... it should fit in a model ship if you wanted to do some gear reduction to a paddle wheel or something like that... 3. Lastly, because you are starting with compressed air, there could be a different way to go about it. Consider a wheel with multiple scoops out of its outer edge. These scoops should be something more like a J profile so that one side was very sloped and the other very "catchy". For the simplest design the outer case would have air input ports and air output ports spaced around. The length of the scoops should be such that no two adjacent ports would be open at the same time. You should consider a different number of scoops (say perhaps six) verses sets of input/output ports (perhaps five) -- so that there should always be an input port open to one of the scoops at all times... thus self starting. The wheel itself would be the flywheel... so you could space some nuts around it for mass. You could use magnets if you wanted to in order to use the controller to adjust the pressure to get a desired speed. The first version could just leave the air pressure on all of the time. If you wanted to get more efficient, you could have an "ignition valve" for each input port. Use a single magnet in the flywheel to make the timing right for all of the input ports. Just some ideas, lots of luck...

A suggestion for increased runtime as well as possible torque would be to use a 4500 psi (approx. 310 bar) compressed air cylinder that they use in paintball. The cylinders are about the size of a 1 liter bottle and contain a large volume of air at an extremely high pressure. You could increase the pressure maintained from 2 bars to 4 bars to see how that works out.

Its always great seeing his face lightup as his machines start to work

Just put dry ice in the bottle and it will constantly keep the pressure up. You could add a safety mechanism to release the pressure if it gets too high.

you can add a pressure regulator to the tank so that the perfomance of the engine stays constant throughout most of the tank pressures

9:54 "That's pretty cool". A truly great pun.

You should probably make a collab with Tom Stanton because he made a similar air engine that he has been perfecting.

I suggested this to Tom as well but I think you could get a big jump in efficiency by implementing a helmholtz resonator around the exhaust ports such that after the pressure spike it cycles to a low pressure for the next cycle's exhaust phase. If you can get down to atmospheric or lower pressure in the cylinder on the up stroke it'd take a lot of waste out of the engine.

what i Love with 3d printing is that you can mould Ideas into shape, and that makes it easier to understand and see where can you improve your product

Relocating that electronic valve to the top of the engine would have yielded a higher rpm and more power per stroke. There're very disruptive waves travelling through that long stretch of tubing, greatly dampening flow. Your magnetic timing is a pretty great idea, though embedding an optical sensor in the sidewall of the cylinder would be much faster to react and more reliable. The only other tip I have is to move the optical sensor mount point lower incrimentally until it counteracts the delay in the flow of the air, similar to adjusting valve timing on a car.

If you put the magnet on the crank wheel, you could adjust the engine timing by moving the hall effect sensor around it - that way you could pre-fire the solenoid valve and get it to go faster :)

I recommend putting the magnet on the flywheel and making the sensor bracket adjustable so you can start the valve opening just a little before the piston reaches the top. If that works out you can use the micro-controller to adjust how far in advance depending on the RPM. You could also use the micro-controller to adjust how long the valve is open depending on the pressure in the tank.

It seems like this would be an easy way to make a silent or quiet motor for a small jigsaw or sawZall. Adding an idling mechanism would increase the run time by keeping it at a constant speed and evening out the weight on the main gear and shortening the top nail would help with efficiency(if you're still using the nail) but giving it a good balance and even another chamber or spring (like a Stirling engine) would be an interesting improvement. I would love to see you make a Jumping Cubli!!!

Brilliant job, I think, increasing your co2 input and more bottles will allow you to support more volume, but other than that it is brilliant. Would also like to see a small force or drag to see if it has the force needed to move an object

As a car guy who loves your videos, it would be very cool to see this design implemented in an inline, multicylinder engine! Much love Integza!

To utilize the full potential of the electrical valve you should, instead of the magnet switch use a sliding contact on the flywheel (like a contact breaker in an old engine), this would allow you to not just open the valve at exactly top dead center, giving you more power and better efficiency, but also give you the ability to tune the engine for power/ run duration by varying the length of the metal strip: Top dead center to ~90 degress of rotation for longer running time | Top dead center to exposure point of the outlets for maximal power Edit: Alternatively you could use a cam & microswitch, instead of a sliding contact, this would be easier to set up, but you'd have to deal with imbalence of the flywheel caused by adding the cam.

Now that you have the cylinders triggering their own pressurization, you could add more cylinders and perhaps a small gear instead of a fixed axel. You could demonstrate the pros and cons of inline vs v shaped engines. My son loves this video. Very informative and safe!

Great video! Instead of using a soda bottle go to a brewing store and look for brown pet bottles, they can handle more pressure and come in loads of sizes. On top of that the same store sell fittings for these bottles and other fittings/lines/tanks for CO_2

If you want to increase max RPM, consider using two solenoid valves in parallel. That way you can alternate between the two, theoretically doubling the rate which you can supply air to the engine.

Improvement Idea: I wonder if you could retrofit a rim/tire to provide both the air reservoir and means of locomotion. If you do the BB8 thing, where an off-center weight is the actual transfer of motion, use your air engine to move the weight forward a bit and let gravity do the work of moving the tire along. If you use your timing circuit to detect top dead center, you could add another signal that detected if the weight is bottom dead center, then AND the two signals together to fire the engine. Or something.

I've built this design (uniflow bash valve) for over 50 years. Use a Reed valve over a hole (inlet) at the cylinder's top! I also build peristaltic engines. Need lightweight higher pressure? Put a chunk of dry ice in your "boiler". The phase-change partial pressure of dry ice is 920 psig!

Use a 1-litre scuba tank and a first-stage pressure reducer. Then you have 200 litres in 200 bars reduced to 8 bars. These tanks are common in the diving community as "argon" tanks for dry suits ( but everybody just uses them with cheaper air). Just fill at any dive shop for under a euro. Any other size tank (0,1 to 15 litres) is usable too.

I would recommend using an air regulator between the bottle and the air engine. I think it will reduce the oscillations in pressure. Also use it with your CO2 recharging system because that was genius. Happy building!

Nice to see you put your own twist on what Tom Stanton has been doing for a while. I feel like with your technical know-how on things outside of the air engine, and his experience with air engines, you two could make something pretty cool if you collabed.

For future testing of pressure vessels, I suggest filling with water first. This way if they do break the explosion will be much less violent, and less likely to injure you.

Awesome video as always! You should time the motor from the fly wheel and if you created a vacuum chamber of some sort ie… mount a turbo on exhaust or something of sorts. and used it on the return of the cylinder to TDC you could potentially double your efficiency. But that being said it’d only be beneficial if the mass was able to be kept down to minimum.

My suggestions: Multiple cylinders. Gas regulator for X bar. Lose the soda bottle or replace with smaller to even out pressure. Parallel valves - If one isn't fast enough, mayby they can take turns (no need for Adruino, just small electronics). More canisters. Good luck and looking forward to the next one.

You should make 8 cylinders connected to 1 flywheel, it can be a neat little v8!

I like this style of video way better than your typical video. Just down to earth, talking about what's happening without the attempts at humor.

Amazing video and a great continuation of Tom Stanton's design. I love this iteration using CO2 cartridges. A couple things : 1. Add a regulator, you will get constant controllable pressure instead of periodically uncontrolled spikes. 2. If the intent is to use CO2 cartridges as the main pressure source, add more in parallel to linearly scale up engine time. You can add more ''fuel'' up to a point where the tubing/fittings mass doesn't balance out the benefits of more engine time. 3. Remove the bottle, simply use the cartridges with tubing, valve and the regulator (less mass). 4. Put the valve closer to the head of the cylinder, it will reduce the volume of gas after the valve and therefore increase the pressure in the cylinder because of the reduced expansion necessary for the gas to reach the cylinder. It will also decrease the pressure drop through the tubing from friction. More pressure = more force = more power. 5. I would love to see more detailed engine performance data out of this, this is awesome for a pressure driven piston engine.

* Tom Stanton enters the chat *

I think you can place the magnet on the flywheel and manually adjust the position of the sensor. Thus you’ll be able to find the most efficient angle of valve opening. Some sort of variable valve timing :)

Great video! Here are song suggestions I have: 1. Remedying backpressure. High exhaust pressure can create backpressure in the cylinder, which resists the movement of the piston during the intake stroke. This backpressure reduces the engine's overall efficiency and power output because the piston has to work against this resistance. Maybe adjusting valve timing, valve size, and exhaust system design would help with that. 2. Consider using materials with lower friction coefficients for the cylinder and piston. For example, you could use materials like PTFE (Teflon) or brass for the cylinder and maybe consider high-quality piston rings to improve sealing and reduce friction. Lower friction = better efficiency. 3. A magnetic sensor and magnet setup is a step in the right direction for sure. Maybe you may also want to check out other valve types, maybe solenoid valves, to achieve smoother operation and better control over the air intake. 4. Maybe you'd like tomato if you had a good cold one. I enjoy a recipe that has been in my family for ages which is essentially a tomato salad. Chopped tomato (bite size), cubes of feta cheese, salt, pepper, and a whole lot of ev olive oil. Mix it up and let it sit in the fridge for a while then eat some as a side with dinner. You can add things like cucumber if you like as well. Worth a try!

You’re genius man,keep it that way!!!

I recommend putting the air valve in the top of the piston to try to limit the amount of waisted air in the lines. It might also help it run at slightly higher rpm.

You should use a mechanical pressure regulator instead of the bottle to have a more consistent air pressure when operating the engine. Also could potentially enclose the crank assembly and have the evacuated air from the down stroke funnel into the crank to push the piston up, potentially increasing efficiency.

Maybe put the magnet on the crank shaft, like a flywheel magnet, to lighten the piston weight while keeping (and possibly tuning) your timing. Also, do you think an opposed piston design would improve efficiency? Also, also: What if you implemented a flexible rubber bleeder valve that sealed when the forced intake air is deployed, but remains open when the piston(s) return to TDC in an effort to remove extra compressive resistance of the piston(s) moving back up?

Hey, very good video as usual! My suggestion is to add a pressure regulator to the bottle so you can just pressurize the bottle, set the regulator to 2 bar and of you go!

Suggestion: You have made a lot of rocket and jet engine and you have mostly used propane for those my suggestion is that you make a jet engine that runs on HHO that is obtained from performing the electrolysis of water and as you already know hydrogen has the highest calorific values so it should give you hell lot of thrust and also try different types of nozzles with it

Use the same magnet idea on a rotary engine, and have three valves for each side of the rotor. It could produce more torque. I’d also love to see you put multiple rotary engines in series to see what effects it may produce. Love your videos man

My suggestion is to create a rack of bottles all connected via a tube with each one having the electro valve mechanism that you made connected to each one!

Here are several tips! All can not be done in the same version but several would work together. Design it like a V engine using atleasr 2 pistons connected to a single axis. Use a smaller magnetic valve for fastet action or push the air in sideways instead from above; this wont force the piston as hard in it's "shut position" the same way. Use springs to dampen the effect of the closed position.

Now, build the 3D prints in aluminum to ensure that the tolerances are correct and use the engine on a small plane!

Since you already have the arduino controlling the bottle pressure, add a magnetic encoder and control the angle to open and close the valve to improve efficiency. You could also measure rpm with this!

that C02 cart with bottle battery was a genius idea!

Your design is really similar to Tom’s but I like the electronic flow regulator. It’s something he didn’t use and opted for a passive design. I think he did it to save weight, but your method allows for multiple fuel canisters and a fast reload capability.

You could use a pressure regulator (en.wikipedia.org/wiki/Pressure_regulator ) in order to keep the pressure constant. You could also improve the piston seal so you don't lose pressure.

Really fun design! If you wished to continue to run this at a higher RPM/higher pressure, you could use a rotary encoder in lieu of the magnetic switch. With how you have made the flywheel, you might not even have to redesign much at all, just cover the empty space outside of the nuts used for weight with a dark paper. I really like the idea of using the compressed air canister as a larger reservoir! I bet there's a way to use the fact that it drops in temperature so heavily without getting too far out into the design-creep mire, but not sure how exactly right now.

You need timing advance to run at higher pressure. One way would be to move the magnet to flywheel as others suggest. Then have the reed switch move, either mechanical governor or an electronic version. Better yet swap the reed switch for a hall sensor and write arduino code. You can detect distance to magnet so timing can be all software. You wouldn't need to move magnet to flywheel in this case, but might still be a good idea.

Instead of having the constant friction of a normal oring or a tight fitting piston, using an expansion ring on the piston that expands when the air goes in and ones the pressure is released through the ports, it's shrinks to its original size and the piston will slide up with more ease. Maybe also play around with the timings of the valve and see if you can find a faster solenoid valve or another valve type altogether, there are some options that open and close in 1ms, not sure the specs of your valve but it's worth looking at if you want to increase rpm while still using the magnetic sensor

The resemblance to Tom Stanton's video about this type of engine is insane, even the same sponsor! x)

I would recommend putting a pressure regulator between the bottle and solenoid, keep the length of tube between the solenoid and engine as short as possible, and moving your sensor to the flywheel. With the sensor on the flywheel you can adjust the solenoid firing time from “top-dead-center” (TDC) for slow speed, and a few degrees ahead of TDC for higher speed. To get really fancy, some testing could result with a gradient equation to get the most torque at every speed.

There once lived a man in New Zealand. I don't remember his name but I watched a ted talk about him. He warmed his cup of tea up by blowing on a whistle submerged in his tea. He passed away without passing on the knowledge he had attained. Could you maybe figure this one out in a future video??..Much love from New Zealand young sir. Love your work..

To remove losses from the ball valve, how about using two pistons and a rotating valve? Send the incoming air to a junction with a cylinder with a rotating valve which is also the end of the crankshaft. Right before the first piston hits the bottom the valve closes the incoming air to the first piston and starts opening the flow to the second piston. The valve is really just two half-filled (D-shaped) cylinders on top of each other, rotating inside a hollow cylinder with incoming air through the bottom, holes on the opposite sides of cylinder wall for the pistons intakes and again, above it holes for the exhaust controlled by the other (inverted) D-shaped cylinder with the exit on top. Easy to draw and make, not so easy to describe.

It would be awesome and it will be more and more lighter. Also, you can try using your hyper jet steam, but controlling the water input(that make the steam) by the pressure remain in the pressure reservoir .Or you can use liquid oxygen( It can be any gas that liquifies at lower than 0*C) in a container, and the just let it to warm up at the room temperature and use it as fuel.

To optimisme the opening and closing time of the valve, try attasching the magnet to the flywheel and move the sensor tangentially and radially to adjust the phase angle end opening time, respectively.

You could move the center of the piston position slightly to the side for a more efficient down stroke and maybe decrease the bottles size for for pressure by decreasing the volume

As an improvement you could tweak the software, so the valve opens for a fixed time. That could improve the runtime even more and it's easy and fast to test.

Could you do a video showing the concept of vacuum buoyancy? It would be interesting to see which shapes and structures can withstand atmospheric pressure while being less dense than air

I like how someone started building some sort of compressed air engine and now slowly more and more of tinkering and engineering channels are infected by the air engine fever. xD It's amazing to see the different approaches, even if the designs are sometimes similar, different tools, materials as well as designing philosophies make every iteration enjoyable.

Really fun video... not sure if it was mentioned already but the thought occurred to me that when you build high performance engines (for cars) you typically tune for better torque and RPM response with adjustable ignition and valve timings. Your design's electronic inlet valve should make this type of tuning fairly simple with an Arduino (retarded timing at low RPM, more advanced at higher). You might also consider modulating the power to the valve and regulating the dwell time to better regulate the amount of gas/air entering the engine. A lot of this would depend on the responsiveness /design of the valve in use but you should still be able to see noticeable gains...

The fit went incredibly hard in this video, especially the sunglasses

This thing is awesome! I had so many ideas while watching. You could attach a throttle to it that just adjusts the point of when the ardunio puts more air in to increase or decrease the minimum pressure. You could use a larger refillable air tank like the kind used for paintball, that would give you a pretty crazy runtime or add a second cylinder offset by 90 degrees to get more torque for the same speed. I’d also love to see what kind of propeller it would spin and how much thrust could be produced.

like a spark plug in an engine you need to account for the delay of your solenoid valve by varying the firing angle. move your sensor from the piston to the flywheel and use the Arduino to calculate the piston position and speed, then add a delay for firing the next valve opening based on the current conditions. depending on the nail to ball size ratio using the solenoid valve should increase efficiency considerably

To reach higher RPMs, you could have the electrovalve open fully once the crankshaft hits the RPM limit of the valve opening and closing

Two hours and 43,000 views! Holy smoke you’ve gone from strength to strength! It’s been amazing to watch you from the very start, and I’m so happy to see your most outstanding success.😊 The beautiful part is you really really have not changed your delivery or format. That says a lot for how good you’ve always been. 🇦🇺🤜🏼🤛🏼😎🍀🍀🍀

I think if you made a similar set up, but instead with super cooled water, there could be some interesting results. If you took cool water then passed it through a super chilled compressor. With the expansion of the water from getting below freezing you'd have a massive pressure difference. You could just recycle the same water as it is used for moving the piston it'd heat up slightly. Although the balancing act of keeping it from freezing would be hard. Just curious if somthing like that is even possible in a construct such as this one. Thanks! Love your videos. Wish I could make some cool stuff like this. I always loved science but as a boy I was really bored of math and thus could use a brush up but if u had practical ways of seeing things applied it makes much more sense.

Have you tried tuning the magnet opening timing backwards for stronger pressures? You could also set that up through the arduino to vary the tuning for different pressures automatically

Same development journey as Tom Stanton it seems :D He 's been working on air pressure engines for a while, using soda bottles to store presssure, and a similar ball valve design. I think you guys should collaborate to make something awesome! Definitely check out his videos, he goes into depth about the process and what he did to make his engines so efficient, purely mechanically!

If you add some kind of heat exchanger, you can improve the gas efficiency, because as the gas expands, the temperature increases, so it expands less. Also it would be better to combine this system with some kind of pressure regulator, so the pressure would be even more stable + I would like to see the final power output of the engine, if you can measure it. Nice video :D

Congrats on bringing electronic injection to air engines! Nice job😊

I’m currently making the version shown at 0:44 so far have the weight and crankshaft but having problems with the pistons length I’m doing this with an electric saw, a bottle cap for the crankshaft, a paper clip for the pistons body, a wood piston head, and the ball valve currently does not have an o ring because I have no idea how to make those or where to find one. For those asking why I don’t use a 3D printer I simply don’t have one and am attempting to see if it’s possible to make it without a printer

Love the Hall effect switch. If I could give some advice, adding variable timing would definitely help performance. Many internal combustion engines run in the 30 degrees advance, not sure how much advance you’d want for an air engine.

You are insane...!! But only in the most creative, inquisitive and daring ways....!!

Build a radial engine! It would have to have a lot of tubing but I really like how they work and I think it could be pretty fun!

You should try to make a Taurozzi Pendulum Engine, that basically has a curved cylinder and piston to reduce friction between those parts. It has been used in 2 strokes to reduce the amount of oil consumed and it amazingly worked.

Great video once again! You should add an elastic chamber before the plastic bottle that, by expanding and compressing, will help to keep the pressure variation constant: it would be something to smooth out the transient, just as a solenoid in an electric circuit or a flywheel in a mechanical system

You have a uniflow engine. As the piston approaches bottom centre it uncovers the exhaust ports letting most of the exhaust gas out, as it starts back up the cylinder the piston covers the exhaust ports and compresses the residual gas which serves to bring the piston to a smooth stop at top centre. Uniflows sometimes need relief ports to avoid over compression. The problem is the valve either stick and ball or magnetic opens and closes the same distance before and after TDC so the first half of the gas is being used to try to push the engine backwards. You don't need this gas to slow the piston down and reverse it as compression in the cylinder is already doing that. If you can adjust your valve timing somehow to open at TDC and close where it does at present you can cut your gas consumption in half and the engine should run faster as you will have got rid of the braking effect of early admission. It should also run with a lighter flywheel as you should need less inertia to carry it over TDC.

You can never go wrong with a good slide whistle foley.

Some improvement ideas would be to adjust your timing on the valve opening. Since you're already using the arduino, you could easily set it up to have a delay. Or, make it so there is a magnet on the flywheel for timing. Next improvement comes from Tom Stantons videos where he improved the seal, I would say, looking into improving the seal as well. Also, try playing around with the stroke and bore of the engine, and if you do some pressure calculations to determine the appropriate amount of stroke/bore for the pressure target you have. It might also be worth while to look into crank offset, it's a somewhat complex topic but could really improve performance and run-time.

I'd say go with a smaller soda bottle reservoir (more compact) and add a bank of CO2 canisters for the energy source. You could also put the flywheel on a gear set to keep the weight low and allow for a little smoother engine for the low RPM. Love the content as always

I would not mind a free 3d Printer to get things stated for my Scouts. A suggestion for improvement: 1 - Magnets on the fly wheel with coil armeture at the base, similar to cumbustion engine push mower 2 - Collect that electricity in super capacitor array to keep the electronics running - (could be a means of using the design to make hybrid vehicles) 3 - recycle the exhausted air to increase the run time

Most plastic quart/litre beer bottles have a plastic liner that allows for a greater pressure reservoir than a soda bottle. An air gunner will tell you (as you saw) that the faster a charge tank empties the more it frosts up and loses power before the charge is empty. You might employ a passive insulator or an active chemical heat source like a hand warmer to encourage better efficacy from the tank.

Hey Integza! You could put a regulator on the inlet to use as a throttle! Awesome video as always.

you can make a 3d printed mechanical pressure regulator (spring and diaphragm)to increase the running time of the engine.

If you are looking for a more constant toque you could have a pressure gauge and an electronic valve at the inlet to the engine. You use the arduino again and make the electronic valve adjust in order to keep the pressure constant.

A bit late, but you could put a short extension on the piston at its circumference, curved to mate with the inside surface of the cylinder and just tall and wide enough to cover a hole in the cylinder side, with a matching hole in the lower art of the the extension, next to the piston face. The curved extension of the piston side is long enough for its surface above the matching hole to shut off the pressure from the gas when the piston is at the lower end of its travel, letting the flywheel send the piston back up. When the holes in the cylinder and piston extension match up, the gas pressure can push the piston back down. That way, the piston does not have to lift up a valve against gas pressure, nor does there have to be a complicated electronic arrangement to open and close the valve.

In the cylinder, drill the holes to let the air escape only on one side. This way you get extra thrust for free!

Use a much smaller bottle as the pressure reservoir and switch the tiny tank for a paintball tank. Also, you might consider a wider piston with a shorter crank distance to increase torque while maintaining pressure usage rates.

Use PTFE or teflon dry lube on your piston, ball, all moving parts. The graphite actually works like fine sandpaper over time creating dust that works against itself over time.

Instead of trying to push air in. Use a vacume to pull the air through. Also the fly wheel. For your weights add magnets. And a copper coil. So when it spins it can generate power for the vacume. Also if this gets added to a vehicle. Allow for the air intake to capture the air from the outside of the vehicle. The vacume only has to run when the vehicle is not moving. When the vehicle is moving just allow the air intake to push the air through the system.

a simple v twin engine with that co2 pressure system might actually produce a significant amount of torque! it'd be fun figuring out a crankshaft and output for the engine, maybe even a centrifugal clutch. super cool as always

Double (or higher) pistons, offset on a crank shaft like an ICE. Multiple cartridges in parallel for longer runtime.

You can put a spring to subject the ball to the bottom of the release valve and do not have leakege when piston is compressing the air

Hey Integza ! Love your videos, especially the ones on air engines ( I have a strange fascination with them, and luckily for me all of youtube also does :p ) I was thinking, since your have an electrovalve, your could time its opening using a microcontroller to regulate the pressure of air in the cylinder, thus making the motor work at a certain pressure wich you could optimise ! You would need a pressure sensor though... Also maybe bringing the valve as close as possible to the cylinder may achieve greater efficiency ( a lot less volume of unswept air volume in the chmber = greater compression ratio I think ) I really look forward to your next videos ! Cheers from France

A pulse engine (PDE) that has both fuel (you'd need to play with proportions of fuel you previously made for it to burn faster into small explosions), BUT it also has a scramjet feature in which it intakes air and combusts it for a speed boost (in short, pulse engine with new optimal fuel, on top of a scramjet feature on it, possibly using your previous turbine projects) - this way you make NEW fuel, INCORPORATE turbines, AND get a new type of engine

@integza check out the di pietro motor, may give you some ideas! Also could wrap up/tape up the soda bottles allowing for more pressure storage and possibly doubling or more the capacities. That could also lead to a bigger boom if it goes so if you do I would do it in layers and make an area that is less wrapped up allowing for a more controlled rupture. Lastly can't go wrong with more magnets, maybe even throw some peltier modules for cooling down or heating up areas...all could be ran off solar. They do have mini electrical bike pumps, used as an unlimited air compressor, that could probably be ran off a direct dc to dc buck converter solar setup to keep weight down. May have to have more than 1 though depending on how long to recharge and such. There's some ideas.