Yeah, I noticed that when I was editing.😒 It was strong enough to bend the carbon shaft, so I was probably wasting a lot of force. Next time I'll get it right...
@IndeterminateDesign yes, I noticed the side load as well. You can either add another linkage to the output so that the shaft stays straight in its path and the linkage translates it or pivot the piston. But it came out pretty decent for a first iteration. I can't wait to see how far you can take this build. I have always wanted a mini hydraulic based RC like machine. Are you planning on selling it at some point?
@@clockworkvanhellsing372 Yeah, that's industrial stuff. Also a servo valve, they are always *insanely* expensive - they are ground to micron precision, and have some pretty crazy engineering work to make them effective.
Micro hydraulics seem like an idea with high potential, but i always got the feeling you had to "know a guy." I am really surprised that robotics haven't used more of it with how easy ratios are with piston vs pump size. This seems way more accessible imo, thanks.
I was trying to figure out the same thing. The cost is quite high for the off the shelf parts. I know some places like boston dynamics have used these, it maybe just a money thing.
I actually print mostly clear resin with my resin printer. They sell clear and I would call tinted clear. I use red and blue clear as well as just clear all the time. If you clean it with isopropyl it tends to be a slightly frosted look, but now they have special resin washes as well as water washable resins that allow you to keep a crystal clear print. You can also just dip the alchahol cleaned print back in the resin and then cure it. The added layer fills in the frosted surface to return it back to clear. I use clear things for stuff like LED lighted things etc.
Mini hydraulics parts for convertible cars are actually pretty cheap, also mini pneumatic cylinders work for hydraulics. Pretty cool though especially for a part that isn't something you can get off the shelf.
Yeah, you can definitely buy something similar off the shelf, but not as small or cheap. My hope was to be able to miniaturize it even further. I’ll look more into pneumatic cylinders. I found a lot of cheap ones on aliexpress and I’m curious how well they work.
Cool project! The 3D models make the parts seem so much bigger than they are in reality, at least to me lol. I really like your modeling philosophy of building the flow paths first and then subtracting them from the solid manifold. Well done!
In addition to tolerances that floating side plates address, those standard herringbone gear shapes will allow a leakage path that effectively joins suction to pressure. If you study commercially available helical gear pumps, you will observe that much like on a screw compressor the geometries are designed with care to minimise the leakage path at the meshing point.
I had debated that. I had read herringbone gears had less leakage, which may not be the case. Do you recommend a spur gear? Any particular pressure angle?
Are you planning on taking it further? Im really excited for it! Would be awesome to actually get a hold of rc hydraulics without giving an arm and a leg!
I do want to develop this more, but I don't have a definite timeline. I want to try several different types of pumps because I think that is the hardest problem to solve at this scale. Once we can generate a couple of hundred pounds of pressure, the rest I think will be much easier to solve.
instead of a gear pump, you shall definitely take an ABS unit from a scrap yard, open it and try using the design features inside of it in order to create a small but powerful hydraulic pump. If you go that path, you should investigate a lot on reducing the pressure peaks as the flow of piston pumps is not linear unfortunately. All in all, I think piston/plunger hydraulic pumps are more fit to be printed using FDM and some postprocessing (epoxy sealing of the printed components)
I agree, I think piston and plunger pumps are the way to go. I had a lot of issues creating efficient valving though, I need more iterations. I thought about the ABS pumps, but size was the biggest factor. I would love to be able to fit this in an RC car of some kind. I think I can make a pump under 2” by 2”.
Great video . I used to machine and maintain hydraulic pumps professionally, this looks like a professionally made system . On another note fdm can manage parts like this , however there will be fitment and size limitation .( Gotta be large enough so that part tolarances are within the available tokerance of the printer ) i gotta get my hands on an resen printer.
That’s awesome. I agree, all the FDM pumps I tried had to be much larger in size. The stretch goal is that I could fit a hydraulic system into an RC car. In hindsight I should have started larger and worked my way smaller.
Control valves *should* have bypass when in "off" position. Extended periods of "off" while the pump is running will cause the working fluid to overheat.
Absolutely. I couldn't think of a way to do this with the disc valves I was using. I copied an existing design used in a lot of RC hydraulics. Next time I want to have individual electromechanical 3 way valves.
@@IndeterminateDesign Have a look at spool valves. They really aren't difficult too manufacture and provide all the functionality you want. Tractors, excavators, etc all use them.
Great project! At some time I was going to try something like this too. Similarly, my idea would be to 3d print the channels and blocks but size the prints so they could be contained in stock profile metal sections. Doing this might help increase the working psi as the pressures could be transmitted from the fluid, through the print to the outer metal casing. Also, research might be needed into which hydraulic fluids can be used which won't erode the 3d print materials. Simple veg/sunflower oil might be all that is needed. My idea was to incorporate 3d printed channels for a hydraulic spool for controls and a hydraulic 360` rotating turntable junction such as used on an excavator but contained in stock metal sections.
Yeah. I think something like vegetable oil would be even better. The CHF11 I used is plant based but really thin compared to a vegetable oil. That’s an interesting idea about encasing the 3D print in metal. What kind of pressures were you trying to achieve? The passage ways in my setup are just 3mm or 1/8” and based on the strength figures for the resin that should handle close to 1000psi. But it probably doesn’t do that in the real world.
@@IndeterminateDesign resin as liquid in tested in a hydraulic press might achieve that sort of pressure. However, solid resin uncontained, is still likely to fail on the layer lines. My project still hasn't even started! However, I'm not expecting large pressures. You might find that you can indeed lift 1kg or more with the identical set up that you have but substitute a much larger bore cylinder. Lifting the load will take more time and more fluid but the system should cope. Possibly the "best" type of pump would be like an internal combustion engine in reverse - as used in construction cement pumps. Such pumps can generate huge pressures and hold the load when the pump is stopped
Good idea, I think you would have more success with a piston style pump driven by a stepper motor, think leadscrew on a resin printer, it would be capable of generating serious amounts of force.
I agree. I used a 2 piston pump as my first prototype, but I lost the majority of the pressure in the valves. I need to try it again with some much more accurate SLA 3d printed parts.
with tesla turbine as pump, you can also 3d print the pump, even simpler better than the two gears you use as the pump now, well two interleaved tesla turbine discs stacks act like the gear pump, direct replacement
The Tesla turbine is already a pump without modifications. It's actually quite hard to print, mostly because of the number of disks required and how thin they need to be.
I would have thought a radial piston pump would work best for 3D prints as the o-ring gives a lot of leeway. I would probably cheat and use a metal cylinder 😂. It does add a lot more valves, though.
I think a piston pump would work really well and I worked on building one for a long time. I did have to use metal tube liners in my test prototypes. It was the valves that killed it. I think with SLA prints I could have gotten the valves to seal better, but FDM was horrible.
what about a FDM surine like pump? so having one of that big suringes actuated by a motor an a theated rod with a nut, that will generete positive/negative pressure fluid to actuate the sistem.
Those definitely would work. I have tried a piston pump which is like a syringe but just resets. I had a lot of challenges with the valves. They need more development.
Try regular spur gears, rather than herring bone gears, most proper hydraulic systems use regular gears, and I think there would be allot less chance of the fluid ‘slipping’ Bach through the pump. There’s really no reason to be using herring bone gears as there’s minimal thrust loads
That’s interesting. Several places I read that herringbone gears have lower leakage losses due to the v shaped engagement. It was said they weren’t used because obviously they’re harder to make. Do you have any papers on this? I printed a set of straight spur gears I may try out here.
No, I don’t have any papers, but in my experience working on industrial hydraulics they’re all straight spur gears I may be wrong and the herringbone is just as good, but it would be interesting to see the difference.
Hello ! I have no knowledge about hydrolics or robotics. But when I see the system I don't understand the advantages of going for hydrolics instead of electric actuator. As the power source is an electric motor, I think that the hydrolics are just a inefficient gearbox. What am I missing ?
In many ways it is just a gearbox. The biggest reason is power density. It’s very easy with hydraulics to achieve a massive mechanical advantage by just changing the area ratios between components. This allows for actuators that are much more compact and lighter than a comparable gearbox. Additionally, a single pump can service multiple actuators. And typically all the actuators aren’t using the full bandwidth of the pump so you have less redundancy.
what do you think about using magnetic gearbox to transmit motion to the 3d printed pump through the housing? (one set of magnets on the gear or abit distanced through shaft disk, and other set outside of housing, each next magnet changing polarity for adequate repell forces)
You just reinvented AC motors but worse. Just use a spinning disk with really strong magnets on both sides if you want to transmit torque without contact?
@@youtubehandlesux ...this is what i was saying, not making a motor inside there, it'd be inificient. But using two sets of magnets as as transmitters of motion.
You need a pivot for your cylinder, you're currently binding it with side load
Yeah, I noticed that when I was editing.😒 It was strong enough to bend the carbon shaft, so I was probably wasting a lot of force. Next time I'll get it right...
@IndeterminateDesign yes, I noticed the side load as well. You can either add another linkage to the output so that the shaft stays straight in its path and the linkage translates it or pivot the piston. But it came out pretty decent for a first iteration. I can't wait to see how far you can take this build. I have always wanted a mini hydraulic based RC like machine. Are you planning on selling it at some point?
You could also just put a slot in the lever arm if you wanted to avoid changing the design too much
Scottish yoke
They make mini hydraulic stuff for remote control excavators. It's neat that you built your own, but I thought this was already a solved problem.
It is still nice to do things as a hobby, as a learning experience, as a challenge.
I think the issue ist cost, not availability. Look at the prices at the beginning.
@@clockworkvanhellsing372 Yeah, that's industrial stuff. Also a servo valve, they are always *insanely* expensive - they are ground to micron precision, and have some pretty crazy engineering work to make them effective.
Yeah, it was more of a learning experience. I also wanted the ability to customize the packaging in the future.
$500 for an excavator hydraulic system.. At the current price, I don't think the problem is solved yet.
Micro hydraulics seem like an idea with high potential, but i always got the feeling you had to "know a guy." I am really surprised that robotics haven't used more of it with how easy ratios are with piston vs pump size. This seems way more accessible imo, thanks.
I was trying to figure out the same thing. The cost is quite high for the off the shelf parts. I know some places like boston dynamics have used these, it maybe just a money thing.
@@IndeterminateDesign to be fair, companies like Moog are far more expensive than they should be.
Wow that is amazing dude I’m also still learning this stuff
I’d love to learn more about your designs
Really appreciate your work
This is so cool! i had no idea you can print crystal clear parts like that tank. thanks!
I had no idea either. I saw the option on JLC3DPs website and I had go try it.
I actually print mostly clear resin with my resin printer. They sell clear and I would call tinted clear. I use red and blue clear as well as just clear all the time. If you clean it with isopropyl it tends to be a slightly frosted look, but now they have special resin washes as well as water washable resins that allow you to keep a crystal clear print. You can also just dip the alchahol cleaned print back in the resin and then cure it. The added layer fills in the frosted surface to return it back to clear. I use clear things for stuff like LED lighted things etc.
Mini hydraulics parts for convertible cars are actually pretty cheap, also mini pneumatic cylinders work for hydraulics. Pretty cool though especially for a part that isn't something you can get off the shelf.
Yeah, you can definitely buy something similar off the shelf, but not as small or cheap. My hope was to be able to miniaturize it even further.
I’ll look more into pneumatic cylinders. I found a lot of cheap ones on aliexpress and I’m curious how well they work.
Cool project! The 3D models make the parts seem so much bigger than they are in reality, at least to me lol. I really like your modeling philosophy of building the flow paths first and then subtracting them from the solid manifold. Well done!
Thanks! Yeah, it’s really small. If it worked, I wanted to be able to fit it into an RC car.
In addition to tolerances that floating side plates address, those standard herringbone gear shapes will allow a leakage path that effectively joins suction to pressure. If you study commercially available helical gear pumps, you will observe that much like on a screw compressor the geometries are designed with care to minimise the leakage path at the meshing point.
I had debated that. I had read herringbone gears had less leakage, which may not be the case. Do you recommend a spur gear? Any particular pressure angle?
Are you planning on taking it further? Im really excited for it! Would be awesome to actually get a hold of rc hydraulics without giving an arm and a leg!
I do want to develop this more, but I don't have a definite timeline. I want to try several different types of pumps because I think that is the hardest problem to solve at this scale. Once we can generate a couple of hundred pounds of pressure, the rest I think will be much easier to solve.
instead of a gear pump, you shall definitely take an ABS unit from a scrap yard, open it and try using the design features inside of it in order to create a small but powerful hydraulic pump. If you go that path, you should investigate a lot on reducing the pressure peaks as the flow of piston pumps is not linear unfortunately.
All in all, I think piston/plunger hydraulic pumps are more fit to be printed using FDM and some postprocessing (epoxy sealing of the printed components)
I agree, I think piston and plunger pumps are the way to go. I had a lot of issues creating efficient valving though, I need more iterations.
I thought about the ABS pumps, but size was the biggest factor. I would love to be able to fit this in an RC car of some kind. I think I can make a pump under 2” by 2”.
I have find that oil pumps from motorcycle engine works great as a hydraulics pump.
Thanks for the tip. I see they are really cheap on eBay. I may pick one up and try to miniaturize it.
Great video . I used to machine and maintain hydraulic pumps professionally, this looks like a professionally made system . On another note fdm can manage parts like this , however there will be fitment and size limitation .( Gotta be large enough so that part tolarances are within the available tokerance of the printer ) i gotta get my hands on an resen printer.
That’s awesome. I agree, all the FDM pumps I tried had to be much larger in size. The stretch goal is that I could fit a hydraulic system into an RC car. In hindsight I should have started larger and worked my way smaller.
Fascinating project. Great learning experience on number of levels. The ability to reduce parts, particularly joints is amazing.
Control valves *should* have bypass when in "off" position.
Extended periods of "off" while the pump is running will cause the working fluid to overheat.
Absolutely. I couldn't think of a way to do this with the disc valves I was using. I copied an existing design used in a lot of RC hydraulics. Next time I want to have individual electromechanical 3 way valves.
@@IndeterminateDesign Have a look at spool valves. They really aren't difficult too manufacture and provide all the functionality you want.
Tractors, excavators, etc all use them.
Great project!
At some time I was going to try something like this too. Similarly, my idea would be to 3d print the channels and blocks but size the prints so they could be contained in stock profile metal sections. Doing this might help increase the working psi as the pressures could be transmitted from the fluid, through the print to the outer metal casing.
Also, research might be needed into which hydraulic fluids can be used which won't erode the 3d print materials. Simple veg/sunflower oil might be all that is needed.
My idea was to incorporate 3d printed channels for a hydraulic spool for controls and a hydraulic 360` rotating turntable junction such as used on an excavator but contained in stock metal sections.
Yeah. I think something like vegetable oil would be even better. The CHF11 I used is plant based but really thin compared to a vegetable oil.
That’s an interesting idea about encasing the 3D print in metal. What kind of pressures were you trying to achieve? The passage ways in my setup are just 3mm or 1/8” and based on the strength figures for the resin that should handle close to 1000psi. But it probably doesn’t do that in the real world.
@@IndeterminateDesign resin as liquid in tested in a hydraulic press might achieve that sort of pressure. However, solid resin uncontained, is still likely to fail on the layer lines.
My project still hasn't even started! However, I'm not expecting large pressures.
You might find that you can indeed lift 1kg or more with the identical set up that you have but substitute a much larger bore cylinder. Lifting the load will take more time and more fluid but the system should cope.
Possibly the "best" type of pump would be like an internal combustion engine in reverse - as used in construction cement pumps. Such pumps can generate huge pressures and hold the load when the pump is stopped
Incredible show of what can be done.
You've inspired me for sure.
Good idea, I think you would have more success with a piston style pump driven by a stepper motor, think leadscrew on a resin printer, it would be capable of generating serious amounts of force.
I agree. I used a 2 piston pump as my first prototype, but I lost the majority of the pressure in the valves. I need to try it again with some much more accurate SLA 3d printed parts.
Awesome build.
Thanks!
with tesla turbine as pump, you can also 3d print the pump, even simpler better than the two gears you use as the pump now, well two interleaved tesla turbine discs stacks act like the gear pump, direct replacement
The Tesla turbine is already a pump without modifications. It's actually quite hard to print, mostly because of the number of disks required and how thin they need to be.
@@doughnut1107 no I think its easy to 3d print
@@gsestream Ok, took me many hours to get 20 disks that didn't fall apart. Your skills must be superior to have it be easy for you when you did it.
Nice! What was epoxy did you use to seal things together? The brass tube into that one block. ty!
Just some good old JB Weld. It seems to stick really well to the SLA resin.
@@IndeterminateDesign Good ol' JB! ty.
I would have thought a radial piston pump would work best for 3D prints as the o-ring gives a lot of leeway. I would probably cheat and use a metal cylinder 😂. It does add a lot more valves, though.
I think a piston pump would work really well and I worked on building one for a long time. I did have to use metal tube liners in my test prototypes. It was the valves that killed it. I think with SLA prints I could have gotten the valves to seal better, but FDM was horrible.
Nothing, but *nothing* is incomoressible. End of!
what about a FDM surine like pump? so having one of that big suringes actuated by a motor an a theated rod with a nut, that will generete positive/negative pressure fluid to actuate the sistem.
Those definitely would work. I have tried a piston pump which is like a syringe but just resets. I had a lot of challenges with the valves. They need more development.
Try regular spur gears, rather than herring bone gears, most proper hydraulic systems use regular gears, and I think there would be allot less chance of the fluid ‘slipping’ Bach through the pump. There’s really no reason to be using herring bone gears as there’s minimal thrust loads
That’s interesting. Several places I read that herringbone gears have lower leakage losses due to the v shaped engagement. It was said they weren’t used because obviously they’re harder to make. Do you have any papers on this? I printed a set of straight spur gears I may try out here.
No, I don’t have any papers, but in my experience working on industrial hydraulics they’re all straight spur gears I may be wrong and the herringbone is just as good, but it would be interesting to see the difference.
What type of shaft seal did you use going from your gear pump to the motor? All the little ones i can find on mcmaster dont seem to hold much pressure
I used double silicone o-rings. I may try some x-rings next time, but I think the real problem was the carbon shaft was still too abrasive.
I would maybe try using mineral oil as hydraulic fluid.
I definitely want to pick some of that up. It would be quite a bit thicker than the CHF11.
Hello ! I have no knowledge about hydrolics or robotics. But when I see the system I don't understand the advantages of going for hydrolics instead of electric actuator. As the power source is an electric motor, I think that the hydrolics are just a inefficient gearbox. What am I missing ?
In many ways it is just a gearbox. The biggest reason is power density. It’s very easy with hydraulics to achieve a massive mechanical advantage by just changing the area ratios between components. This allows for actuators that are much more compact and lighter than a comparable gearbox.
Additionally, a single pump can service multiple actuators. And typically all the actuators aren’t using the full bandwidth of the pump so you have less redundancy.
@@IndeterminateDesign thanks a lot, it makes perfect sense !
what do you think about using magnetic gearbox to transmit motion to the 3d printed pump through the housing?
(one set of magnets on the gear or abit distanced through shaft disk, and other set outside of housing, each next magnet changing polarity for adequate repell forces)
That would be interesting to try. I'm not sure how much torque the motor is transmitting, and how big or how many magnets would be needed.
You just reinvented AC motors but worse. Just use a spinning disk with really strong magnets on both sides if you want to transmit torque without contact?
@@youtubehandlesux ...this is what i was saying, not making a motor inside there, it'd be inificient. But using two sets of magnets as as transmitters of motion.
If the torque is insufficient you can even put reduction gearing inside, no?
Amazing
likely losing some pressure in long plastic hoses
Yeah, that makes sense. Hopefully I can figure out a way to get about 10x that pressure with a version 2.
Comment
Ackshually