You might consider offsetting the centerline of the piston from the centerline of the crankshaft. Look at the VW VR6 engine for an example. By offsetting the crank, you can change the amount of lift duration you will get on either side of Top-Dead-Center. You can also use this to get the downward stroke of the engine such that the connecting rod is more vertical during the power stroke, which reduces side forces on the piston while under load. Playing with this could yield some benefits to the design.
A compression release would improve efficiency. What an engine like this needs to make the most of the air supply is a valve that opens precisely at top dead center then closes right before the piston uncovers the exhaust port. Also right at bottom it should have a valve that opens an exhaust port at the top of the cylinder so there's little resistance to the piston moving up. That valve should close just before the air inlet valve opens. All that would make the most efficient use of the compressed air, but the linkage to get the two valves operating would be complex.
What if there was two holes in the side of the cylinder wall near the top, one on each side. Then a moving sleeve on the outside of the piston with two holes in it that correspond to the two holes in the cylinder, except they are offset in height. That way when the sleeve is slid up, one hole opens, and when it is slid down the other hole opens. Now connect that sleeve to the crank shaft 90 degrees out of phase with the piston. That way you have a valve mechanism which switches from intake to exhaust when the piston is on top and on bottom. It would require more precision fits to keep it airtight, but it should be about the same as making the piston seal inside the cylinder.
@@greggeshelman No, I think you want the intake closed well short of the exhaust opening, or you get no expansion benefit...ie. you exhaust high pressure air WITHOUT it doing any work. Would be high power, but lousy efficiency. I agree his design is WAY to far the other way, and the intake valve needs to be open longer than it is. I suspect, off the top of my head, what you actually want is just a bit above atmospheric at exhaust port opening. All of this can be calculated to provide any efficiency/power curves you want, within it's capability, which for this engine is ZERO. ;-) I would have designed it to be double acting, as well. Then you can use a MUCH smaller flywheel, you are not throwing all THAT energy away, and the power will double (more or less) for a given piston sweep.
Maybe the Carbon fiber prop couldn't carry enough momentum at low rpm to keep the engine running? I am by no means an expert though, so I could be wrong...
Definitely the flywheel is too 'small'. To get maximum increase here with minimum additional weight, mount extra weight at the end of the propellor. Must be well balanced though. This would surely do the trick, since the design idea itself is excellent!
That's actually a really good idea. Make the rod coming up out of the piston a threaded rod. Would just need a jam nut on it or something to keep the timing from slipping while the engine was in operation. Or, once optimal timing is found glue into place.
I appreciate seeing the failures as much as the successes. I have always learned more for failure than success. Keep it up, the engineering approach is refreshing.
Those exhaust holes look pretty small to me considering they need to let all of the air out of the chamber in such a short time. Maybe you should try increasing the size of those.
The spring is there to offset the opening and closing of the valve. It is opened at TDC and closes much later because of the spring. The design without the spring is basically symmetrical, so after the chamber gets pressurized the piston does go down but it wastes too much energy to go back up. I.e. the piston gains almost the same amount of energy going from TDC down as it takes for it to go back to TDC. Add some friction and your engine goes puff puff stop. I say replace the center plastic pin with a metal one that moves up and down and rubs against the con rod. The top shape of the con rod should be made like a cam so that it pushes the center pin out. This way you no longer need a spring and you achieve basically the same thing.
You might also want to experiment with the amount of volume above the piston. Because if it is too small then you might also compress the remaining air too much in the up stroke. And if it is too big you waste a lot of compressed air.
I wonder if there is a way to properly design these things without resorting to trial&error. Or at least do the trial&error in some kind of simulator instead of a 3D printer ;)
Your engine needs to compress a small amount of air while the piston is going up so you lose energy there, I think your best bet would be to have a separate valve activated by a camshaft. You could just make a rectangular piece that runs in a groove in the side of the engines are rides on a off center circle connected to the crankshaft, you lose a bit of energy to friction but you would still lose less than you do with your current design. It could also work as an exhaust valve.
It's air, operates cool, and doesn't seem to need much (if any) lubrication. In theory you could have the air push the piston up from the crankcase and get a more compact design. (So pulling on the usual "pushrod" of associated with a piston design.)Then have a slider-block type valve on the flywheel to meter and time when the air is allowed in.
I just noticed you tried to put your face on the video thumbnail like a lot of youtubers do to gain clicks. It looks like it backfired, which tells us a lot about the people watching your videos. They seem to be really interrested in engineering things, not only watch but also doing it themselves. Thats worth an abo from me!
A couple suggestions: 1) a heavier flywheel (prop) goes a long way to making troublesome expansion engines run 2) the crank bearing should be mounted eccentrically in a sleeve in the crank case nose. Turning the sleeve would have the same effect as lengthening or shortening the peg on top of the piston. This is how Telco Shark CO2 motors are adjusted 3) have you considered putting an inflation valve right into your bottle cap assembly? That would allow unmodified bottles which could be quickly swapped out and comfortably run at a higher pressure
I really love your contents. Please make a rotary version too. Edit: I meant rotary engines or different engine designs.(the dorito ones, any other engine types would be cool though) please make a turbofan with compressed air tube running into the compressor blades or something
air vane motor, just like in an air tool , no springs in them! or a piston controlled inlet like in a toy steam engine, the cylinder moves in a wobble to cover/uncover the inlet and exaust
HalfBredNinja I understand what your saying and it’s completely right i was just referring to a rotary engine as in a car. I know it’s actually called the wankle but sometimes referred to as a rotary
Always good to combine the strengths of materials nice project! One mod I would try is the piston piece on the top use a threaded rod or screw so you can make very small adjustments. The top of threaded rod that touches the ball can have a piece of hose to give it a very slight amount of shock reduction and give it a little more compression which will make it run smoother.
Nice Experiment. Also experiment with rotating the motor, try having the head at the bottom, gravity might help holding the ball valve open for longer, or open for ever! The ball valve is closing too early, the spring used to keep it open for longer. One solution could be a second conrod to action the valve with the correct timing Another solutions is a 2 stage valve... the story is: it takes a lot of force to open the valve, it takes very little tho keep it open. It is used a lot with hydraulics
Currently my most favourite youtuber by far at the moment, always watch your videos first! Keep up what you are doing and you deserve more subscribers!
"Back in my day" we had Telco CO2 motors which operated on the same principle. A couple of features they included that you might be interested to incorporate. Firstly, the crankshaft bearings were mounted on an eccentric carrier which was rotated to act as an adjustable throttling device - that is like you sanding the top off your piston. Secondly, the pin is aerodynamically important and (from memory) was profiled to flow the gas radially outwards from the pin axis. I'd also guess your pin is blocking most of the inlet port area which is what's causing the lack of running.
You can separate the piston and tip that open the airflow. You need two shafts and two steps on the crankshaft. It will shift a piston little bit down, when air flow is opened.
Bigger flywheel required. Try a metal bottom-end on the motor or a prop/flywheel combination for the sake of a quick prototype. Single cylinder engines require much higher rotational inertia. Enjoying your thinking processes here.
I bet a V-twin air engine would be awesome and effective! Both cylinders would be angled up so the ball bearing would always seal by gravity plus you’d have much more power! Have you ever considered a pure mechanical way to let air in, for example a poppet valve operated off a cam? Great videos keep it up!
Once you get this bit working, I suggest you try to incorporate a pressure regulating ball-valve in the cap area of the engine, keeping the pressure in the manifold somewhat constant. As the pressure of the air-feed changes, this necessarily also changes the optimum opening-timing of the ballvalve in the manifold, just like ignitiontiming has to change depending on load in a combustion engine. My feeling is, that this system rod and ballvalve system is very sensitive to the pressure it is being fed. OR, you could just "cheat" and make it more steam-engine like by using a sleeve valve, governing when to let air into the manifold. :P
I think that the main reason that this failed was that there is no negative return pressure. Just 1bar or atmosphere at TDC that is then compressed probably to about 3 or 4 bar. It was the return spring that did most of the heavy lifting on the last engine. If you wanted rid of the spring you could use 2 pistons If you made them horizontally opposed they could use the same con rod. These are just my thoughts and I might be wrong. Please tell me if you think so. Great video. "From success we learn nothing."
Maybe make the top of the cylinder a bit wider, so more compressed air can rush in to later pust out the piston. Also try putting in a bigger ball with a bigger hole,so there is a bigger cross section for the air to move in. One more thing, if you make the piston's pushing part by drilling and threading the piston and drive in a screw, it's going to be adjustable ;)
Just a few thoughts, you might want to have a pushrod actuating the valve that is separate from the piston, so that you can adjust the timing. Also, if you made the engine a 3 cylinder rotary, you could design it so that it can't go backwards by nature of its design, and you can use things like cam rings that actually make the engine simpler, although on that scale you might have to go with a planetary gear with little valve pins actuated by an eccentric gear mounted to the crankshaft. Another thing that might help is regulating the gas flow to a specific pressure, which might make tuning the engine easier, and allow you to get crazy with super high pressure tanks, like 150psi+, but allow you to keep using the lower cost materials that otherwise wouldn't stand up to the absurd stresses.
I think the issue is the piston is experiencing as much force in the opposite direction as it is in the forward direction, so the forces are canceling each other out. To go springless you could use a single cam on the crank shaft and have it open the ball valve. Then you would be able to adjust the timing on the engine as well as determining exactly how long the valve is open for.
You have so much knowledge, just do a multi-cylinder Version. 2 or 3 should be enough to get rid of the spring and let it run smother! Smaller holes would counter the higher air consumption.
Are you loosing compression around the piston? I saw the sanded version, just wondering if there's too much of a gap there. Also is the ball re-seating reliably? If that used a spring, maybe the required spec wouldn't be as finicky. Nice design improvements, keep it up.
looks awesome! When you're still struggeling to find the issue you may want to consider to build something for 'debugging'. (thinking of an oscilloscope with some simple contacts or mayb some plug and play hall-effect sensors :) ) Would give an awesome insight into the working engine
My thoughts are that the piston is having to compress too much air on the up stroke. Perhaps increase the volume above the piston at the top of the stroke, or move the exhaust ports further up the cylinder. You will lose power, but it might run...
If you get the new design working, do one with two pistons and two bottles on top of each other so you could have the high pressure-tolerance of the single bottles but the capacity of the glued bottle, and also a smoother running engine.
This idea might be more suited for a multi cylinder setup so you dont have to coast as far. Also, check the water rocket you tube channels. They get the same strength out of double and tripple joined bottles compared to a single bottle. You'll get lots more volume.
Instead of changing the length of the ball valve why not add a retention bar so that the ball only moves enough to add in enough air. That should fix the main problem. Then if speed is still an issue I suggest removing that spring inside of the engine and adding a second piston to the system. At 8:47, when the propeller stops it does a return probably due to that spring storing the energy that the propeller wants to use to keep moving and continue the system. A second piston would take advantage of the smaller gap and allow for better air use with two strokes.
Co2 cartridges are very high pressure, the motor would most likely explode. But if a metal regulater was used the cartridge would last a long time however the regulater might be to heavy to put on the plane
Here is a idea might require some scale changes, increase the size of you piston barrel so piston can only reach 80% of barrel height, replace the ball bearing with longer shaft which has hole/gap with a o-rings on end of the gap in the shaft (this hole allows the air flow to the barrel). When the piston is moving down the cycle then the gap closes, at the highest the gap is open for air. Because your piston has expansion chamber now as well I think it will also allow for higher pressure like using co2 but i don't know what that would do on the plastics your using... Also alternative is to put the spring on the shaft on-top of the piston but change your cycle so that when the piston is at it highest that the air gap is closed but the spring then pushes shaft on the piston down so it can allow air flow. I think this is then similar to blow back action of paintball/bb gun but that action is connected with bolt to the valve action and your action is all contained in one action...
you could add a smaller crankshaft connected to a slider valve by a smaller connecting rod. that, or you replace the ball with an exceptionally wide spring-loaded tappet valve that can allow a lot of air to flow into the chamber while barely having to move by top dead center, and add a somewhat heavy flywheel. i'm no engineer, but that's something i'd personally try.
If you lengthened the blue cap, you might be able to move the valve stem up into the 3d printed parts so that there is no need to puncture the bottles. You might get more pressure that way.
you might consider a more positively controlled intake/exhaust system...cam actuated, rotary disc, etc. one that offers the ability to keep the exhaust open till near tdc. enjoying your videos!
The highest you can safely go with a soda bottle (NOT WATER BOTTLES) is ~85PSI. Also, keep an eye on the seam in the bottle. You should try making a v-twin air engine.
In the US when I was a kid there was a set of toys from AirHogs which were air powered. I had a car and motorcycle that had an air motor and tank you would pump up to get them to run. I remember the motors were done in clear plastic so you could see the internals. They both ran pretty good. Maybe you can find some info on there designs to help with yours. No play existed as far as I remember but this was before brushless motors and RC battery packs were NiCad. LIPo packs and well at that time Cell phones were things of the future. Good Luck.
Two pistons or a double acting piston are the way to go. However a phase lag to the valves is still needed. I’d use a barrel valve on the crank. The exhaust could be ran through it too.
As a tinkerer and inventor, I understand the need to solve a problem, but I fail to see the purpose of a compressed air engine. Having said that, what your design is missing is an adjustable timing system and flow regulation. Being able to adjust the timing of the stroke without disassembly and modification of engine speed would allow for more efficient tuning, while air flow regulation would allow for a more consistent flow of air over the duration of charge and allow for longer run times. Having a regulated air flow would also allow you to tune the engine down to the absolute minimum pressure needed to run, thereby giving you a baseline for performance as you proceed with the design tuning and modifications. it would also allow you to estimate potential run times given a starting air pressure in the bottle.
Have you looked at pneumatic hammers ? They work by accelerating a hammer backwards and forwards through pressurised air and clever porting Seems you could be able to adapt this to an engine, replacing the hammer with a piston and the anvil with a crank
What if you used a larger ball for the ball valve? It should allow more immediate pressure build up. That'd be preferrable to the slow spring method... seems inefficient to fill the valve through most of the down stroke. Someone else suggested larger exhaust ports, and I agree with that as well. And what are your thoughts on making a CO2 adapter to replace your bottle cap adapter? It'd save you a lot of pumping... but of course then you have to buy the CO2 bottles.
The problem with CO2 is twofold. First the pressures contained in a CO2 cartridge FAR exceed the bursting strength of the materials he is using. He would need a pressure regulator to step the pressure down. Second, the decompression of the CO2 coming out of the cylinder makes it severely cold! O-rings fail regularly at these temps (Discovery SRB). Also, most extruded plastics become quite brittle at these temps.
When thinking about how an engine works, you have and intake stroke, compression and combustion, then the exhaust.... in that cycle, the compression makes the combustion more aggressive. In this design, there seems to be the intake (compressed air) and the exhaust, then a compression for when the piston makes its way back to the top. What if you eliminated that compression stage. Maybe check out how steam engines were designed. They use the same ports for the exhaust and intake, but use some fancy valve work to have an intake stage on both sides of the piston eliminating the power loss from compression.
It seems like you're basically making a steam engine with air power instead of steam, there are probably very many steam engine designs that could be referenced for ideas. I think your current design is good, without the spring; but if you had the valve actuator on a separate connrod/crank pin from the piston (rather than as an integrated part of the piston), you could adjust the timing independently, opening the valve when the piston is actually at TDC, and keeping it open until just before the piston opens the exhaust valve. More complexity of course, but I think would immediately result in a decently working engine with all the "pros" you're looking for.
You could "potentially" see better results in terms of timing with a rotary valve. Granted that's much more difficult to design/make BUT the rewards may pay off. That said, a Wankel rotary would definitely be something to see. Great videos man!
You could always do the math and get exactly how long the pin should be. Also, here's my suggestion: make an actual valve instead. Have a small rotating disc that opens and closes the airway, just use a rubber band or a gear to spin it off the crank.
Look into how stationary steam engine work, they're basically just compressed air engines but steam has more energy. you could use an eccentric to time a valve opening and closing the air supply, or use a fly wheel to maintain momentum through the top of the stroke. Or both, both is good. Source: model enginerd.
maybe bleed the exhaust into the crankcase to push the piston back up and add a rotary reed valve to release the pressure from the case as the piston gets close to TDC.
Why_are_there _no_names Useless. Would not make nearly the same in power and would be much less efficient. Not to mention there is no compression, intake, or exhaust stages, meaning a complete redesign would be necessary.
In steam engines the valve controller can be run on the same flywheel as the piston but it needs to be offest to that it doesn't open the valve before the piston is completely closed. Your crank shaft just needs another crank on it, set a little behind the piston shaft, so it will open the valve just after the piston has passed its high mark and started its descent.
Hello. I like your video and efforts very much. When I was young a had e working steamtractor toy. Quite large actually. It had a double working cilinder. It utlitized very clever timing and sliding valve mechanisme to use a smaller bit of pressure to help push the piston up and just through the top dead center in your model. The main 'down' valve actually opened quite late. The piston itself could do open and close in and outlet ports. It requires pressure channel along te side of the cilinder though and the piston needs to pass them and therefor open and close them at the right moment.
Your problem seems to be the ball valve, it doesnt have enough force pushing it down back into the o-ring after the valve has been opened and so the valve stays open for too long. I know the spring is the thing you wanted to eliminate, but you could do with one behind the ball valve pushing downwards.
You might consider offsetting the centerline of the piston from the centerline of the crankshaft. Look at the VW VR6 engine for an example. By offsetting the crank, you can change the amount of lift duration you will get on either side of Top-Dead-Center. You can also use this to get the downward stroke of the engine such that the connecting rod is more vertical during the power stroke, which reduces side forces on the piston while under load. Playing with this could yield some benefits to the design.
USWaterRockets yes that would actually be a very good alternative, good explanation.
That's exactly what I was also thinking about. It would prevent the engine from backfiring - basically putting the timing "later".
A compression release would improve efficiency. What an engine like this needs to make the most of the air supply is a valve that opens precisely at top dead center then closes right before the piston uncovers the exhaust port.
Also right at bottom it should have a valve that opens an exhaust port at the top of the cylinder so there's little resistance to the piston moving up. That valve should close just before the air inlet valve opens.
All that would make the most efficient use of the compressed air, but the linkage to get the two valves operating would be complex.
What if there was two holes in the side of the cylinder wall near the top, one on each side. Then a moving sleeve on the outside of the piston with two holes in it that correspond to the two holes in the cylinder, except they are offset in height. That way when the sleeve is slid up, one hole opens, and when it is slid down the other hole opens. Now connect that sleeve to the crank shaft 90 degrees out of phase with the piston. That way you have a valve mechanism which switches from intake to exhaust when the piston is on top and on bottom.
It would require more precision fits to keep it airtight, but it should be about the same as making the piston seal inside the cylinder.
@@greggeshelman No, I think you want the intake closed well short of the exhaust opening, or you get no expansion benefit...ie. you exhaust high pressure air WITHOUT it doing any work. Would be high power, but lousy efficiency. I agree his design is WAY to far the other way, and the intake valve needs to be open longer than it is. I suspect, off the top of my head, what you actually want is just a bit above atmospheric at exhaust port opening. All of this can be calculated to provide any efficiency/power curves you want, within it's capability, which for this engine is ZERO. ;-) I would have designed it to be double acting, as well. Then you can use a MUCH smaller flywheel, you are not throwing all THAT energy away, and the power will double (more or less) for a given piston sweep.
Maybe the Carbon fiber prop couldn't carry enough momentum at low rpm to keep the engine running?
I am by no means an expert though, so I could be wrong...
Yeah I think there needs to be more mass for sure. Maybe a small steel flywheel mounted behind the prop.
At least it worth testing with a flywheel instead of a prop. Just to study the engine.
When it is mounted on a flying airplane the airflow through the propeller will give it a lot more momentum.
Definitely the flywheel is too 'small'. To get maximum increase here with minimum additional weight, mount extra weight at the end of the propellor. Must be well balanced though. This would surely do the trick, since the design idea itself is excellent!
What 3D printer are you using Tom? It's really good that it produces airtight parts.
you could use a screw to open the valve then you can just turn the screw to increase or decrease the timing.
That's actually a really good idea. Make the rod coming up out of the piston a threaded rod. Would just need a jam nut on it or something to keep the timing from slipping while the engine was in operation. Or, once optimal timing is found glue into place.
I appreciate seeing the failures as much as the successes. I have always learned more for failure than success. Keep it up, the engineering approach is refreshing.
Fail forward
Those exhaust holes look pretty small to me considering they need to let all of the air out of the chamber in such a short time. Maybe you should try increasing the size of those.
i think what the spring does is it lags the air intake, whith this design it is the only way to control the timing
or add additional exhaust ports.
The air goes very fast,so you don't need a big holes
3:55 literally just made a smiley face
The spring is there to offset the opening and closing of the valve. It is opened at TDC and closes much later because of the spring. The design without the spring is basically symmetrical, so after the chamber gets pressurized the piston does go down but it wastes too much energy to go back up. I.e. the piston gains almost the same amount of energy going from TDC down as it takes for it to go back to TDC. Add some friction and your engine goes puff puff stop.
I say replace the center plastic pin with a metal one that moves up and down and rubs against the con rod. The top shape of the con rod should be made like a cam so that it pushes the center pin out. This way you no longer need a spring and you achieve basically the same thing.
You might also want to experiment with the amount of volume above the piston. Because if it is too small then you might also compress the remaining air too much in the up stroke. And if it is too big you waste a lot of compressed air.
K van der Veen good point.
I wonder if there is a way to properly design these things without resorting to trial&error. Or at least do the trial&error in some kind of simulator instead of a 3D printer ;)
Something along the lines of AutoCAD with a CFD (Computational Fluid Dynamics) plugin?
hmm, went looking at CFD options: AutoDesk CFD and SimFlow were the main ones I found. SimFlow has a free evaluation version.
In real cars this space is known as the cc ( cubic centimetres) and the car makers have perfected this ratio so very good point U have there
Your engine needs to compress a small amount of air while the piston is going up so you lose energy there, I think your best bet would be to have a separate valve activated by a camshaft. You could just make a rectangular piece that runs in a groove in the side of the engines are rides on a off center circle connected to the crankshaft, you lose a bit of energy to friction but you would still lose less than you do with your current design. It could also work as an exhaust valve.
The bottle should be the fuselage w/everything fixed to it, Or create a pressurised fuselage to use as a tank? Love the ideas on this channel:)
It's air, operates cool, and doesn't seem to need much (if any) lubrication. In theory you could have the air push the piston up from the crankcase and get a more compact design. (So pulling on the usual "pushrod" of associated with a piston design.)Then have a slider-block type valve on the flywheel to meter and time when the air is allowed in.
Make A V twin motor!and have the 1st piston open the other valve to the #2 postion?
Yeah some kind of twin I think.
good to see you back Tom, hope your holiday was good.
I just noticed you tried to put your face on the video thumbnail like a lot of youtubers do to gain clicks. It looks like it backfired, which tells us a lot about the people watching your videos. They seem to be really interrested in engineering things, not only watch but also doing it themselves. Thats worth an abo from me!
Nice to have you back Tom!
A couple suggestions:
1) a heavier flywheel (prop) goes a long way to making troublesome expansion engines run
2) the crank bearing should be mounted eccentrically in a sleeve in the crank case nose. Turning the sleeve would have the same effect as lengthening or shortening the peg on top of the piston. This is how Telco Shark CO2 motors are adjusted
3) have you considered putting an inflation valve right into your bottle cap assembly? That would allow unmodified bottles which could be quickly swapped out and comfortably run at a higher pressure
Nice work! The fill valve can be put on the cylinder head. It will remove the need of tooling on the bottle.
Always great to be included along your journey. Love how your mind works. True curiousity.
Great to see you back in action!
I really love your contents.
Please make a rotary version too.
Edit: I meant rotary engines or different engine designs.(the dorito ones, any other engine types would be cool though)
please make a turbofan with compressed air tube running into the compressor blades or something
Xdew Gaming&More do you mean radial?
air vane motor, just like in an air tool , no springs in them! or a piston controlled inlet like in a toy steam engine, the cylinder moves in a wobble to cover/uncover the inlet and exaust
+1 for WW1 style rotary :)
AIO inc. a radial engine has pistons and a rotary has “Doritos” both would be cool though
HalfBredNinja I understand what your saying and it’s completely right i was just referring to a rotary engine as in a car. I know it’s actually called the wankle but sometimes referred to as a rotary
Always good to combine the strengths of materials nice project! One mod I would try is the piston piece on the top use a threaded rod or screw so you can make very small adjustments. The top of threaded rod that touches the ball can have a piece of hose to give it a very slight amount of shock reduction and give it a little more compression which will make it run smoother.
OOOHH...finally an air engine video!
this video wasn't a fail, I love to see the development process
You should get a DLP 3D printer for those small parts, looks like it could be a great addition to your tools/machines!
Nice Experiment.
Also experiment with rotating the motor, try having the head at the bottom, gravity might help holding the ball valve open for longer, or open for ever!
The ball valve is closing too early, the spring used to keep it open for longer.
One solution could be a second conrod to action the valve with the correct timing
Another solutions is a 2 stage valve... the story is: it takes a lot of force to open the valve, it takes very little tho keep it open.
It is used a lot with hydraulics
Nice work tom! Very impressive (as usual :P )
would it work better if you had another piston and bottle on the bottom which was 180d out of phase with the current one?
Currently my most favourite youtuber by far at the moment, always watch your videos first! Keep up what you are doing and you deserve more subscribers!
"Back in my day" we had Telco CO2 motors which operated on the same principle. A couple of features they included that you might be interested to incorporate. Firstly, the crankshaft bearings were mounted on an eccentric carrier which was rotated to act as an adjustable throttling device - that is like you sanding the top off your piston. Secondly, the pin is aerodynamically important and (from memory) was profiled to flow the gas radially outwards from the pin axis. I'd also guess your pin is blocking most of the inlet port area which is what's causing the lack of running.
You can separate the piston and tip that open the airflow. You need two shafts and two steps on the crankshaft. It will shift a piston little bit down, when air flow is opened.
The compressed air engine series is extremely cool. I'd love it if you'd like to continue it.
Bigger flywheel required. Try a metal bottom-end on the motor or a prop/flywheel combination for the sake of a quick prototype. Single cylinder engines require much higher rotational inertia.
Enjoying your thinking processes here.
I bet a V-twin air engine would be awesome and effective! Both cylinders would be angled up so the ball bearing would always seal by gravity plus you’d have much more power!
Have you ever considered a pure mechanical way to let air in, for example a poppet valve operated off a cam?
Great videos keep it up!
Once you get this bit working, I suggest you try to incorporate a pressure regulating ball-valve in the cap area of the engine, keeping the pressure in the manifold somewhat constant.
As the pressure of the air-feed changes, this necessarily also changes the optimum opening-timing of the ballvalve in the manifold, just like ignitiontiming has to change depending on load in a combustion engine. My feeling is, that this system rod and ballvalve system is very sensitive to the pressure it is being fed.
OR, you could just "cheat" and make it more steam-engine like by using a sleeve valve, governing when to let air into the manifold. :P
I think that the main reason that this failed was that there is no negative return pressure. Just 1bar or atmosphere at TDC that is then compressed probably to about 3 or 4 bar. It was the return spring that did most of the heavy lifting on the last engine. If you wanted rid of the spring you could use 2 pistons If you made them horizontally opposed they could use the same con rod. These are just my thoughts and I might be wrong. Please tell me if you think so. Great video.
"From success we learn nothing."
Maybe make the top of the cylinder a bit wider, so more compressed air can rush in to later pust out the piston.
Also try putting in a bigger ball with a bigger hole,so there is a bigger cross section for the air to move in.
One more thing, if you make the piston's pushing part by drilling and threading the piston and drive in a screw, it's going to be adjustable ;)
Just a few thoughts, you might want to have a pushrod actuating the valve that is separate from the piston, so that you can adjust the timing. Also, if you made the engine a 3 cylinder rotary, you could design it so that it can't go backwards by nature of its design, and you can use things like cam rings that actually make the engine simpler, although on that scale you might have to go with a planetary gear with little valve pins actuated by an eccentric gear mounted to the crankshaft. Another thing that might help is regulating the gas flow to a specific pressure, which might make tuning the engine easier, and allow you to get crazy with super high pressure tanks, like 150psi+, but allow you to keep using the lower cost materials that otherwise wouldn't stand up to the absurd stresses.
I think the issue is the piston is experiencing as much force in the opposite direction as it is in the forward direction, so the forces are canceling each other out. To go springless you could use a single cam on the crank shaft and have it open the ball valve. Then you would be able to adjust the timing on the engine as well as determining exactly how long the valve is open for.
i really like this concept, i think its really interesting and i would like to see more of it. Keep up the great work!!
Chloroform to vapor smooth PLA. Keep up the cool work!
You have so much knowledge, just do a multi-cylinder Version. 2 or 3 should be enough to get rid of the spring and let it run smother! Smaller holes would counter the higher air consumption.
Are you loosing compression around the piston? I saw the sanded version, just wondering if there's too much of a gap there. Also is the ball re-seating reliably? If that used a spring, maybe the required spec wouldn't be as finicky. Nice design improvements, keep it up.
Science progresses as a series of failed designs, it's good to see stuff that didn't work completely but taught new things.
I'm just fascinated watching these videos!
Tom, consider an adjustable screw in place of the fixed length piston, so you can adjust the timing.
@3:54 The motor were clearly happy about your new idea.
looks awesome!
When you're still struggeling to find the issue you may want to consider to build something for 'debugging'. (thinking of an oscilloscope with some simple contacts or mayb some plug and play hall-effect sensors :) ) Would give an awesome insight into the working engine
My thoughts are that the piston is having to compress too much air on the up stroke. Perhaps increase the volume above the piston at the top of the stroke, or move the exhaust ports further up the cylinder. You will lose power, but it might run...
Never give up mate! You'll make it work! Have you thought about running it with gas? (CO2,..)
I love how the end of your explanation made a smiley face. Could you just have an impulse turbine on the inside of the engine?
You should make a huge RC WW2 plane using this idea 💡
Nice work Tom. :) How constrained is the ball bearing at the top? Could it be bouncing around unpredictably upsetting the timing?
Great video as always! Loved the explanation!
What a happy crankshaft diagram!
If you get the new design working, do one with two pistons and two bottles on top of each other so you could have the high pressure-tolerance of the single bottles but the capacity of the glued bottle, and also a smoother running engine.
This idea might be more suited for a multi cylinder setup so you dont have to coast as far.
Also, check the water rocket you tube channels. They get the same strength out of double and tripple joined bottles compared to a single bottle. You'll get lots more volume.
Four minutes into the video, and the fly wheel on the diagram looks like a smiley face!
Do be afraid to post failures we all learn from it still great content
Instead of changing the length of the ball valve why not add a retention bar so that the ball only moves enough to add in enough air. That should fix the main problem. Then if speed is still an issue I suggest removing that spring inside of the engine and adding a second piston to the system. At 8:47, when the propeller stops it does a return probably due to that spring storing the energy that the propeller wants to use to keep moving and continue the system. A second piston would take advantage of the smaller gap and allow for better air use with two strokes.
Great - even just the "bottle top" is very useful - thanks!
Could you tap a co2 bulb into the bottle in for longer range?
Co2 cartridges are very high pressure, the motor would most likely explode. But if a metal regulater was used the cartridge would last a long time however the regulater might be to heavy to put on the plane
Here is a idea might require some scale changes, increase the size of you piston barrel so piston can only reach 80% of barrel height, replace the ball bearing with longer shaft which has hole/gap with a o-rings on end of the gap in the shaft (this hole allows the air flow to the barrel). When the piston is moving down the cycle then the gap closes, at the highest the gap is open for air. Because your piston has expansion chamber now as well I think it will also allow for higher pressure like using co2 but i don't know what that would do on the plastics your using... Also alternative is to put the spring on the shaft on-top of the piston but change your cycle so that when the piston is at it highest that the air gap is closed but the spring then pushes shaft on the piston down so it can allow air flow. I think this is then similar to blow back action of paintball/bb gun but that action is connected with bolt to the valve action and your action is all contained in one action...
you could add a smaller crankshaft connected to a slider valve by a smaller connecting rod. that, or you replace the ball with an exceptionally wide spring-loaded tappet valve that can allow a lot of air to flow into the chamber while barely having to move by top dead center, and add a somewhat heavy flywheel. i'm no engineer, but that's something i'd personally try.
If you lengthened the blue cap, you might be able to move the valve stem up into the 3d printed parts so that there is no need to puncture the bottles. You might get more pressure that way.
you might consider a more positively controlled intake/exhaust system...cam actuated, rotary disc, etc. one that offers the ability to keep the exhaust open till near tdc. enjoying your videos!
The highest you can safely go with a soda bottle (NOT WATER BOTTLES) is ~85PSI. Also, keep an eye on the seam in the bottle.
You should try making a v-twin air engine.
The old compressed air engine is how I found your channel, a few days ago.
Spooky stuff.
In the US when I was a kid there was a set of toys from AirHogs which were air powered. I had a car and motorcycle that had an air motor and tank you would pump up to get them to run. I remember the motors were done in clear plastic so you could see the internals. They both ran pretty good. Maybe you can find some info on there designs to help with yours. No play existed as far as I remember but this was before brushless motors and RC battery packs were NiCad. LIPo packs and well at that time Cell phones were things of the future. Good Luck.
Two pistons or a double acting piston are the way to go. However a phase lag to the valves is still needed. I’d use a barrel valve on the crank. The exhaust could be ran through it too.
Love your videos and channel mate 🤘
As a tinkerer and inventor, I understand the need to solve a problem, but I fail to see the purpose of a compressed air engine.
Having said that, what your design is missing is an adjustable timing system and flow regulation.
Being able to adjust the timing of the stroke without disassembly and modification of engine speed would allow for more efficient tuning, while air flow regulation would allow for a more consistent flow of air over the duration of charge and allow for longer run times.
Having a regulated air flow would also allow you to tune the engine down to the absolute minimum pressure needed to run, thereby giving you a baseline for performance as you proceed with the design tuning and modifications. it would also allow you to estimate potential run times given a starting air pressure in the bottle.
Nice to see u back!
more than one piston might be a good idea?
Have you looked at pneumatic hammers ? They work by accelerating a hammer backwards and forwards through pressurised air and clever porting
Seems you could be able to adapt this to an engine, replacing the hammer with a piston and the anvil with a crank
What if you used a larger ball for the ball valve? It should allow more immediate pressure build up. That'd be preferrable to the slow spring method... seems inefficient to fill the valve through most of the down stroke.
Someone else suggested larger exhaust ports, and I agree with that as well.
And what are your thoughts on making a CO2 adapter to replace your bottle cap adapter? It'd save you a lot of pumping... but of course then you have to buy the CO2 bottles.
The problem with CO2 is twofold. First the pressures contained in a CO2 cartridge FAR exceed the bursting strength of the materials he is using. He would need a pressure regulator to step the pressure down. Second, the decompression of the CO2 coming out of the cylinder makes it severely cold! O-rings fail regularly at these temps (Discovery SRB). Also, most extruded plastics become quite brittle at these temps.
Martin Dinner I knew the pressure would be an issue, but I totally forgot about the decompression temperature drop.
3:54 Smile!
When thinking about how an engine works, you have and intake stroke, compression and combustion, then the exhaust.... in that cycle, the compression makes the combustion more aggressive. In this design, there seems to be the intake (compressed air) and the exhaust, then a compression for when the piston makes its way back to the top. What if you eliminated that compression stage. Maybe check out how steam engines were designed. They use the same ports for the exhaust and intake, but use some fancy valve work to have an intake stage on both sides of the piston eliminating the power loss from compression.
Yeah, FIVE Weeks!!!!!
Finally back! ;-)
It seems like you're basically making a steam engine with air power instead of steam, there are probably very many steam engine designs that could be referenced for ideas. I think your current design is good, without the spring; but if you had the valve actuator on a separate connrod/crank pin from the piston (rather than as an integrated part of the piston), you could adjust the timing independently, opening the valve when the piston is actually at TDC, and keeping it open until just before the piston opens the exhaust valve. More complexity of course, but I think would immediately result in a decently working engine with all the "pros" you're looking for.
You could "potentially" see better results in terms of timing with a rotary valve. Granted that's much more difficult to design/make BUT the rewards may pay off. That said, a Wankel rotary would definitely be something to see. Great videos man!
Springless Piston probably needs you to make the ball-valve rod separate from the piston again, so you can change its timing relative to the flywheel
You could always do the math and get exactly how long the pin should be.
Also, here's my suggestion: make an actual valve instead. Have a small rotating disc that opens and closes the airway, just use a rubber band or a gear to spin it off the crank.
Thread the top of piston to make adjustable lifter length, use threaded rod. More mass in flywheel, use a real wheel not a prop for testing.
welcome back
My personal thought is that you should be looking at using cams to time the valve (essentially for your exhausting and expanding stages).
Look into how stationary steam engine work, they're basically just compressed air engines but steam has more energy.
you could use an eccentric to time a valve opening and closing the air supply, or use a fly wheel to maintain momentum through the top of the stroke. Or both, both is good. Source: model enginerd.
It’s the R in R&D keep on iterating, us engineers do that. Awesome
That's a very tidy shed!
maybe bleed the exhaust into the crankcase to push the piston back up and add a rotary reed valve to release the pressure from the case as the piston gets close to TDC.
Dude your a genius fair play to you 🤘🏼
No springs, moar cylinders! :D
Maybe you should incorporate an eccentric sleeve in the crankcase; this would allow the piston TDC to be adjusted by rotating the eccentric sleeve
How about an air Wankel?
Why_are_there _no_names Useless. Would not make nearly the same in power and would be much less efficient. Not to mention there is no compression, intake, or exhaust stages, meaning a complete redesign would be necessary.
Oh yes, because it’s of utmost importance to achieve maximum efficiency and performance from your air powered toy engine 🧐
A Wankel wouldn't have enough expansion due to compression of the air to work
A compressed air Wankel would just be a wobbly turbine
BoostedJDM Well if he wants to put it on his airplane again then there's no reason to waste time on it
In steam engines the valve controller can be run on the same flywheel as the piston but it needs to be offest to that it doesn't open the valve before the piston is completely closed. Your crank shaft just needs another crank on it, set a little behind the piston shaft, so it will open the valve just after the piston has passed its high mark and started its descent.
Then you can work on a pressure regulator
You've drawn a happy crank shaft
Finally BACK! :3
Hello. I like your video and efforts very much. When I was young a had e working steamtractor toy. Quite large actually. It had a double working cilinder. It utlitized very clever timing and sliding valve mechanisme to use a smaller bit of pressure to help push the piston up and just through the top dead center in your model. The main 'down' valve actually opened quite late. The piston itself could do open and close in and outlet ports. It requires pressure channel along te side of the cilinder though and the piston needs to pass them and therefor open and close them at the right moment.
Your problem seems to be the ball valve, it doesnt have enough force pushing it down back into the o-ring after the valve has been opened and so the valve stays open for too long. I know the spring is the thing you wanted to eliminate, but you could do with one behind the ball valve pushing downwards.
Yay you're back