3D Printed Pneumatic Actuator... It Actually Works!
ฝัง
- เผยแพร่เมื่อ 14 ธ.ค. 2023
- This video is sponsored by PCBWay:
For high quality and fast PCB & CNC manufacturing visit: www.pcbway.com/
Signup for a free Onshape Account: Onshape.pro/MichaelRechtin
Patreon: / michaelrechtin
CAD for this project: cad.onshape.com/documents/68a...
Learn more about Formlabs and the Form 3+ here: formlabs.com/company/innovato...
Follow me for more updates on projects:
Instagram: michaelrechtinyt
Twitter: / rechtin_michael
Hardware List: (Affiliate Links)
1" PVC Pipe: www.mcmaster.com/4740K64/
M4 Threaded Rod: amzn.to/46MmHgw
24x20x2 O-Ring: amzn.to/3GCIaxH
7.8x1.9 P8 O-Ring: www.allworldmachinery.com/ite...
8mm Rod: amzn.to/485G0mF
Air Fittings: amzn.to/3RhqbSD
Air Line: amzn.to/3RxuElj
Solenoid Valve: amzn.to/3RBt2ag
Links to stuff I use (affiliate links):
Bambu Lab X1C: shrsl.com/29zix-2yn7-1cwx9
Creality K1 Max: amzn.to/3uTuXy0
Creality CR-10S: amzn.to/3BuAiM2
Ender 3 S1 Plus: amzn.to/3xv7zW8
Elegoo Mars 3: amzn.to/3UwVvh1
Creality LD002H: amzn.to/3QXhpqJ
Creality Wash Station: amzn.to/3S1NTBi - วิทยาศาสตร์และเทคโนโลยี
The contraption you made to flick the switch back and forth is glorious, you basically made a very powerful and elaborate useless machine.
It's not useless, it had a purpose.
@@SplarkszterGoogle useless machine
In a world full of microprocessors, what could be better than a mechanical analog solution!
It is certinly not useless. It tests for reliability.
On the contrary, not useless at all. the product was knowledge and experience.
I work for a pneumatics manufacturer who shall remain unnamed, and this is incredible! Couple of notes here though (2 and 3 are probably unnecessary, but fun regardless):
1: Don't use NPT threads for this, as they're a conical that's designed to form a seal when the crests of the male threads deform into the female threads... Look for G threads or BSPP (or honestly for small cylinders there are M5 or 10-32 fittings) which will be parallel and seal with an O-ring. This will put much less stress on the print and they don't need any Teflon tape!
2: Many air cylinders use restricted air flow to act as a cushion instead of a mechanical spring, which I'd love to see working here. Essentially you need a shoulder on the front of the piston that reduces the area that air has to escape through for the last inch or so of the stroke. It slows down your cycle time, but the smoother acceleration is generally better for all your parts.
3: While O-rings are fine for the static seals, you'll get better performance out of V-shaped or Z-shaped seals for the dynamic seals. The V or Z cross section creates a one-way seal (this is why pistons are designed with 2 seals), but the benefit is that the more pressure you put in the chamber, the harder it seals against the tube and the less leakage you have. They're also generally lower friction and handle debris/scratches better than O-rings. They're pretty readily available online, or if you wanted to be cheeky you could order a seal repair kit for an actuator and just design your parts around the kit. Probably defeats the purpose of doing this as cheap and simple as possible, and you've clearly shown that the O-rings work more than well enough.
Again, super interesting project and I can't wait to try and replicate it myself!
lubricated o-rings should be fine for low PSI...hydraulic stuff at thousands of PSI need MUCH more sealing force...and backing washers to prevent extruding the o-ring past the grove.
+1 For the BSPP or G threads. I can even model matching threads in my 3d prints. The taper on NPT just destroys 3d prints. I designed vacuum pods for CNC work.
Would you be interested in designing a better version of this incorporating your proposed changes? It would be very appreciated I think :)
My father would have been sooo envious. In the 1950's he had a one-man contract precision engineering business in the basement of our home.
He converted a couple of lathes to fully automatic, to produce thousands of air-line fittings. Most of the mechanisms involved sequenced pneumatic cylinders, which could be a bit cantankerous at times, but when everything was ticking along it had a rhythm all of it's own. Sheer magic!
Try running the threaded rod the full length of the endcaps, so there's no layers to split
Good thinking... maybe even add flat steel or alloy discs as the first and last items in the stack so the plastic layers are all in compression.
Mind you, if it can stand ten thousand cycles in half an hour as it is, it probably doesn't need anything unless increases in pressure or speed are introduced.
I love the analog switch counter mechanism you came up with... So good!
Print the end caps at 45° or 90° from the build plate so the kinetic energy isn't perpendicular to the layer lines. Should be way stronger. Great video!
I noticed a few things you can improve. I have several years experience dealing with cylinders. First, you are making a bomb, be careful. Design your cylinder to work in 60-90 psi, that’s what the pros suggest. More than that is just a waste of air and money. I would get hold of a catalog from a company like Parker, they have lots of little built in problem solvers that adding in 3D printing should help keep make it better. An oscillating circuit is possible as well. This would get rid of the switch, I see it works but there are more graceful solution, like a physical or magnetic limit switch. All in all it is a great start. Hydraulic will be much safer but don’t get any of that fluid under your skin, it is devastating. Same with pressurized air, don’t feel around for leaks if you can avoid it. Google injuries in you have a strong stomach. Good luck.
I have made several slightly pressurized liquid containers and adapters, and have used Varathane, a thin urethane-based wood hardener to seal layer porosity. Just pour the Varathane in, apply a slight pressure, then removed the excess and allow several hours to dry. I applied pressure by just blowing into a fitted tube. You will likely find a few seeps when the pressure is applied, but once dried I had no leaks. I also replaced tapered pipe with straight threads with oring seals.
This is a really great idea!
One of the best projects I've seen on TH-cam so far. Excellent work!!
Banger video. Love it! Will def be looking to incorperate one of these in my future projects
My final project for engineering degree was to build an ENTIRE 3D printed pneumatic linear actuator: even the sealings were 3D printed 😂 and I did it with a very low cost ender 3. The cylinder was able to resist up to 4 bar with no losses.
I really enjoyed your project too 😊😊
The video quality is superb! You're gonna be huge one day.
Brilliant, Michael! Fantastic design and work! 😃
Stay safe there with your family! 🖖😊
And happy holidays!
Awesome! Great ingenuity and craftsmanship!
Excellent design! Thank you for contributing to our 3D Printing community!
I am working on a similar project right now for a fully automatic can crusher. great video! looking forward to see the hydraulic portion.
Your design relies on the tensile strength between z layers in the end caps to contain pressure. If your tie rods were longer and the held the printed z layers in compression against pressure, it would be significantly stronger.
Good point
This was really cool man, keep doing this kind of stuff! Great watch.
This is quality maker content! I'm definitely interested in the hydraulics, but I'd also love to see that pneumatic one run in that tester, to just see what fails
Also: Not sure it'd work with hydraulic fluid, but bumpstops or polyurethane disc bushings might be a better option (vs. springs) for energy absorption inside the cylinder - Amazon probably has them close to the price of springs and would eliminate the metal wear/spread load on internal cap faces
Also also : Gorilla glue on the NPT fittings works great to seal the threads without having to tighten them down enough to accomplish sealing between the thread faces. Use it all the time on cheap chinese npt fittings - and it locks the fittings when it dries so you can clock them in the most convenient directions. Epoxy works better but cleanup is much easier w/ the Gorilla glue - need a wire wheel on a bench grinder to cleanup the epoxy
love it. cant wait to make this. GREAT JOB
Your test was awesome to watch
Awesome job, very well done!
The testing setup in awesome!
DIOOOOOSSSS
Tu mayor fan!!!! Increible!!! INCREDIBLE VIDEO MEN!!!! Icant wait to see the next one!!!!!!
Very nice project testing. I also would like to see a resin printed version with testing result to see how would them hold up between FDM and Resin printed!!!
Nice!! Love it!!
Nice I love your videos keep doing it
Amazing video! Well done.
Awesome project! Did you have to melt the outer layers of the pla to get an airtight seal between layers?
you could move the flange for the threaded rods to the very end of the end caps, so that the threaded rods can take all the tensile loading. then when the piston crashes into the end cap, the layers are compressed together instead of pulled apart. then the only failure mode for the printed part would is shearing of the layers themselves, instead of relying on the relatively weak inter-layer adhesion.
“Didnt expect to break here first”……literally the thinest part…..at a mounting point……with layer lines in parallel with the mounting bar lmao
that thing is crazy lol, great vid
This contraption tests other stuff too - for instance, that little toggle switch is still going at 10K as well.
You can create an air pocket which will do the same as your spring. I recall they call that an air cushion.
First of all, great vid. I just have 1 question: why do you use rafts on the FDM prints??
Build a vacuum chamber..then soak the FDM parts in polyurethane, vacuum it down and let the polyurethane soak into the layers and inside the print...100% infusion, or even that slow setting 2 part epoxy....
Also...why not have a glue on cap with a screw on nut, this way to service, you just unscrew the nut like on a hydraulic cylinder
Instead of using a mechanical switch, use optical sensors, and an arduino....very very very cool!
my favorite part was when you zoomed in the circular saw, it made me laugh
this is amazing video
Nice work….incorporating cushion stops on your piston and cylinder ends would control speed at the end of stroke and adding pneumatic speed controls in each line would allow you to control flow / speed independently in each direction
If you want the great layer adhesion and preferably well sealing materials, I'd suggest trying to print with a copolyester filament (pet, petg, pctg, cpe, etc etc), and turning down the fan speeds as much as possible. PCTG also had a very slick surface, something to look into
PETG with little to no cooling fan (in a heated chamber) works SUPRISINGLY well..and the layers just seem to bond like glue
Air pistons use air to dampen the impact on the end caps. There is a chamber in the end caps and the piston has a cylindrical extension on both ends that go into those chambers, basically air springs are formed.
Very smart! Thanks.
Very nice experiment. Maybe you should print the end-caps by positioning them on the printer's plate such as the "grain" is in the same direction as the force applied to the piece.
Very satisfying
I also 3d printed a full size car. The part that is 3d printed is the tire valve caps. Lol.
Cost effective compact pneumatic actuators for applying pressure to material(stock) or parts(simple or complex) in a custom fixture for CNC machining is something I'm working on instead of custom flat head screw based vise blocks which is what I've been using for machining simple and complicated parts that are hard to fixture.
Projects like what you did in this vid I find very exciting because there's an interesting future in small scale cost effective machining with decent output levels or cycle times and things like cost effective custom or part specific pneumatic actuator clamps(side or top clamp) for rapid parts holding is one of the keys to making that happen, vacuum fixtures and palletizing are also highly important.
I'd be interested in seeing you try to design a linear actuator that isn't very tall and a small footprint that can be bolted to a fixture plate with 6mm or less of actuator travel but can apply enough sustained clamping pressure for machining. Shouldn't be to hard to get enough pressure for clamping wood or plastic. Clamping composites (G10, carbon fiber), aluminum, copper and capable engineering plastics I could see needing a higher clamping force.
great stuff
Printed tpu seals would be a nice touch.
EXCELENT TEACHER
How do you determine the oring crush factor?
Pro tip: rapid kick your epoxy. You have a 3D printer. Set bed temp to 80 and put a box over it and throw your curing epoxy stuff in there. You might fully cure it in 4-6 hours instead of 24. Not only is it faster, hot curing epoxy makes the final cure tougher and improves adhesion.
BTW when you have 1:1 epoxy, have you ever thought what is in those components? Like when you're dealing with pure chemicals, you don't need more than a few % of the crosslinking agent, and well one part is the epoxy monomers, so what's in the other half? Well there you find resin monomers as well, some crosslinking agent, fillers (often fumed silica) and usually some plastifiers. So the "hardener" part of 1:1 epoxy also often makes it softer and tougher. So you can control properties of the resulting epoxy by adjusting the mixture by up to about 30%, more "resin" part gives you a harder and more chemically resistant outcome, while more "hardener" part gives you a tougher and more flexible one. Very useful when you're gluing things that are particularly hard or particularly soft and you don't want to introduce stress via glue joint, or when instead you're casting and need particular mechanical properties. However you gain the problem then that it will take an eternity to cure (or will outright fail) once you start adjusting the mixture. No problem, hot curing allows to force it to cure regardless.
I know, nobody cares, sorry.
I feel like PETG would be perfect for the caps as FDM print. The layer adhesion is very good and when you torture it by slowly overtightening, it all just deforms a little and doesn't just crack. I actually sort of despise PLA and try not to use it for any purpose that isn't decorative. I also print HIPS a lot, and i don't hate it, i don't know why everyone treats this material as a joke. My issue with PLA is that every single print i did several years ago that was anyhow loaded has failed in the interim, with discolouration and crazing. Initially it's super strong but that's not worth much if i can't rely on it. I have not had any issues with my PETG and HIPS prints.
How do you identify which part is which when they're both clear?
@@anon_y_mousse Hopefully your manufacturer has given them a helpful label like "resin" and "hardener" rather than "A" and "B". Otherwise... the resin part doesn't have a smell. The hardener is the stinky one. The hardener is also usually a little more yellow and usually a little less flowy. They also have a density difference, but it can come out either way, though hardener is usually heavier.
@@SianaGearz Thanks.
And i thought... not a circular saw, band saw, vise for the material, but... a pneumatic actuator from scratch of course. Elegant solution 😂😂
5:27 the confining thread acts as a stress riser, that why it fail. you had like 1 mm of 3d printed layer supported by the thread and nut. this meant that you had a almost pure tension/shear force against a really small cross section of interlayer plastic holding it. the solution didnt need the springs, you just had to print the caps with so the screws were fixated farther in the cap, o even extend the cap. the problem here is like stress propagation in soil. you have an max angle that you can effectively transmit support between the force, the medium and the reinforcement. the only way that the current design could work is just beefing it with steel so the good awful geometry induced stress riser can be overcome by the steel marginal resistance (the resistance after being fatigued by infinity), or make a almost infinitely rigid material (carbon fiber? diamonds?, both would suck against fatigue or impact resistance tho), so the angle of transmission of force can be about 90° (eliminating the possibility of stress riser altogether ).
or you can just extended the thread a bit further back in the caps an let the plastic work in a strong mode: compression and let the steel deal with the tension in the machine.
Should try printing tpu orings and buffers instead of springs and rubber.
Instead of using springs, why didn't you extend the structural bolts to be longer, going over the full front and back end? The front piece broke at the ends of the bolts near the front, so if you extend that, it wouldn't break...?
I think the spring solution is better, because it's addressing the amount of force/time transferred to the printed piece, vs. just trying to make the printed piece resist more force. Even with longer structural bolts, the middle of the print would still be shock loaded and probably begin to separate on the inside, away from where the bolts are clamping
Neat video, I see one flaw in the design. The print layer orientation. You keep putting the Z orientation to the most strained parts (the hook part at the end, and the part that was "hammered" all the time. I bet if you changed the print orientation of those two you could get better results.
6:21 Nikola Telsa would love that lol
I am going to try the top cap in TPU.
whats the max working pressure before it blows up :)
I wonder if a flexible filament could be used to replace arteries, even if only as a temporary measure.
GiGi Dior would LOVE this!
hmm...I see a log splitter in my future now lol
Your idea gave me an idea, a boxer robot who loves punching people😂😂😂
If you fear leakage on one of your prints your best bet is to slow down the print speed to let all layer properly fuse together 😉
What brand is the switch? Good advertisment for them😅
I'm just thinking that you should put something between the rod and the springs to avoid wearing it prematurely 🙂
beautiful, now if only it could push up an umbrella, when folded, and then retract to open the umbrella, and force down a lot of air.
6:12 the stupidest and most genius thing ive ever seen. its hilarious. You earned yourself a subscriber.
can you make a high speed plane, like you mq9 reaper video but focus on speed. maybe even with some afterburners. and also next video upgrade you rc car a bit more too!!! Love the video!!!
So I'm curious on another experiment/possibility:
If you pressurize both sides, could you lock the actuator at the halfway point?
Yes you could. The weak link would be the seal around the shaft. There are types of rolling seals that can eliminate leakage at this point as well.
Hii budy , can you please make a vdo on pneumatic Without big size compressir bcz it will help alot to make a Battle bot. Like Flipper battle bot , they using this type of pneumatics
Any chance of getting a Google drive with stls of your projects? Onshape is good, but I don't use it or know how to, in a fusion guy, lots of others are too and it'd be nice to just have easily printable STL files ready to go if we want to reproduce this. ❤
You can export from provided link, just do some research on onshape to fusion 360 export
Hey bro I want the exact dimension of your project diy cnc laser cutting machine
I know you said you could print the entire thing, yet this specific device used in the video is stated by the title as "3D Printed Pneumatic Actuator" this device is still just a Pneumatic Actuator with some 3D Printed Parts.
Honestly you should hydro pressure test this. The last thing you want is a pneumatic cylinder blowing up in your face at 100psi.
I made an actuator kind of similar to yours, I was testing it with compressed air (computer duster) and just like you it hit too hard and the end broke and blew off, Flew right into my LCD screen and broke it. lol
Bro i am waiting for 3d printed penultimate and you drop a video
I would like to see what happens if you light hair spray inside it kind of like an internal combustion engine, maybe you could make a combustion engine that combusts on both sides of the piston but just disregard the crankshaft and have a cool boomtube instead lol
The way your eyelet mount failed is the same way suspension fails on electric scooters with these shocks on them Look at Dragon Lightning V2 with the silver shocks, they're basically butter.
Should have made bufferchambers on bothe sides, then the piston wont ram the endpositions both sides.
You know well how it started
0:09 The cylinder is already on the table before he goes to grab it in the next shot....
I've 3d printed o-rings with very soft tpu, just to let everyone know is possible!
I can think of something else that can be put on the end of that thing. 😜
I've seen something similar in adult movies
Not sure if that hydraulic pump is going to be big enough
Need a saw to make a saw
Print surface is choosen wrong. Just turn the parts 90degrees and you will see piston durability increases minimum two times
Ehem you know what is much simpler ? Circular saw.
I don't understand how resin would be more durable. In my experience, resin has always been more brittle and fragile compared to any standard fdm material
And now make a pneumatic engine
Dude this rules
Please apply for patent before it’s too late!
epoxy does not dry, it cures.
STOP. DO NOT use a motorized pump to test to failure. Just use a hand pump. You can get them affordably at Harbor Freight, they're used for auto body work and metal punches. They generate pressures in the multiple thousands of PSI, you'll get the desired results.
The incompressible nature of water (and oil) and the low elasticity of some of your parts means that pressures will rise VERY quickly; you can't control a pump motor finely enough to prevent an instant blowout. With the hand pump, you can go very slowly indeed.
Did you really just use a table saw to cut a piece of pipe for a pneumatic actuator that you were going to use to saw wood?
🤔 car ideas...
make a engine out of these
Nice project.
Could activate the switch sooner so you don't over extend so hard.
I would also use flow control valves to have a smooth controlled extend and retract motion with no damage.
th-cam.com/video/2JUz1vKNFYA/w-d-xo.html
Not sure if you could print flow control valves though. But you could add a restricted exhaust.
Hmmm 4:00
So no one will say it?