[Robert MacCurdy] at MIT wants to change how people think about hydraulics. Using fluid can be very useful in systems like robots, but it is often the case that the tubing that carries hydraulic fluid is not an integrated part of the overall design. [MacCurdy] and his colleagues have modified a 3D printer to allow it directly include hydraulic components as it prints.
The idea is simple. The team started with a printer that uses a liquid ink that is UV cured to produce solid layers. The printer has the ability to use multiple liquids, and [MacCurdy] uses hydraulic fluid (that does not UV cure) as one of the print materials. Just as you can use a 3D printer to build structures within other structures, printing the hydraulics allows for complex closed systems that use the UV-cured resin as mechanical parts that can transfer pressure to and from the hydraulic system.
The group has printed several structures including a gear pump. The pump uses a pair of cogs in a tube that move fluid that is trapped between the teeth. This pumps the fluid in one direction. The gears mesh tightly to prevent flow in the reverse direction. If you have access to a UV resin printer like this, you might be interested in the source paper, so you could try to duplicate their results.
Hydraulics show up in many places in the non-3D printed world. Examples include power steering and even amusement parks. With the MIT technique, designers have a new tool to use when 3D printing moving systems.
AWESOME MATERIAL MAP!!! :)
That is brilliant, hydraulics are very powerful, and can be silent. This new method allows people to fabricate small mechatronics systems that are more “insect like”. It will be interesting to see how people solve the multiple protected fluid lines inside joints problem.
Would distilled water not be an option?
It is an option, but is it the best option given it could leak?
Why would water cause a problem if it leaks? You could also use vegetable oil if you like.
Silicon oil is ideal because it is inert, is an insulator, and is stable. Veg oil oxidises and makes a sticky mess, many of them are also a bit acidic. I have a few cheap Chinese pedestal fans that were probably oiled with veg oil deliberately, it means the fan stops working after a short amount of time, long before the mechanical parts show any wear and tear, but long enough to not be covered by a warranty. :-(
“Boris van Galvin” said distilled water was an option, and then “Dan” asked if it was the best option “given it could leak”.
I agree – silicone oil is the best option.
Silicone oil is very difficult to jet cleanly as discrete droplets out of most inkjets though. Even harder if the viscosity is high.
Have you tried it? Some silicon oils seem particularly thin and slippery to me, in fact I am sure some are very close to water in their viscosity. What about Octamethyltrisiloxane? Have you even researched such compounds?
Siloxanes won’t jet easily. I have done it. The problem is you either need a completely separate cold nozzle for the silicone, and use a peristaltic pump to feed the silicone, Or you need to use a silicone oil with the same viscosity as the molten plastic you are using.
The big show stopper with silicone oil is- if even the tiniest trace of silicone goes where it’s not supposed to- the plastic layers above that area won’t stick to the layers below and it ruins the print.
Hopefully the working fluid doesn’t eat the plastic. Sounds silly, but it’s a real possibility.
silicon oil?
Silicone oil would be better in this application, and is available in a wide viscosity range.
Silicone oil must be a lot cheaper now than it was when I last used it industrially 20 years ago. At that time, reclaimed silicone oil was $10,000 per US gallon, and new silicone oil was $15,000 per gallon. Certainly too expensive to be used for something like this.
Is there some reason why plain old 30-weight hydraulic oil wouldn’t work?
Regular hydraulic oil should work fine, but would need to be tested for long term material compatibility with the other wetted parts of the print. At the low pressures and temperatures involved, nearly any fluid could be used, even vegetable oil, which is finding use in more food processing and animal containment / handling equipment these days due to it’s safety.
Silicone oils are more expensive than regular hydraulic oils, but are inert at these temperatures. Silicone oils are finding there way into more hobbies as the price gets lower and lower. RC car shock / struts are often filled with with silicone oil for damping, and a variety of viscosities are available for the racer to adjust the damping as needed.
DOT 5 brake fluid is about $10-12 USD per pint, give or take.
Brake fluid will dissolve some plastics, so it would need to be tested with the plastics used in the print.
I’d prefer something like ATF fluid over 30W oil.
Add some printed conductive traces, use an electro-reholitic fluid to make integrated valves.
Electrorheological fluid requires high voltage to aglomerate, but that sounds feasible. Another possibility would be a magnetorheological fluid (ferrofluid), and use small solenoid coils to act as valves, or even pumps. Actually, you don’t even need a separate pump / cylinder / valves with the ferrofluid idea.
Sure this might be possible- but practical? A quad headed extruded system to make decent at best built in hydraulics. This also means hydraulics basically need to be redesigned from the ground up using these materials instead of using off the shelf units.
They are MIT people, they probably have a simulator half written for it already, and if not, they easily could write one with all the relevant physics built in. The paper (.pdf) indicates they are already using some related design tools to optimise the structures.
I doesn’t need to be practical, it’s just a demonstration by a group at MIT as part of their post-doc work under NSF grant.
It is pretty neat, and has some limited applications.
That’s my point- why dump hundreds of thousands of dollars into research into something we *know* has very limited applications? For someone else to implement this, not only will they have to spend likely tens of thousand dollars for the hardware, but the equivalent of thousands of hours in experience.
I don’t believe people should be spending this kind of money and manpower to do things just because they can. And of course, because they are MIT, no one even questions it. It’s so easy to convince people what you’re doing is useful when you have a name behind you that it’s basically manipulation.
I missed the bit where you demonstrated that it’s applications were limited. I can think of a few civilian and military uses where it would be very useful. I can even think of ways that it’s particular strengths can be used to create designs that are able to fail more elegantly when damaged, now that has huge potential.
…because space. boom
You would need to look up the NSF grant number to see what application they had in mind. Apparently the NSF thought it was unique enough to give them money.
I hate this kind of attitude. “I don’t have multiple applications for it already in my mind, therefore nobody should research it to find potential applications.” It’s the main thing that slows the progression of science in the world today.
He’s just mad because he doesn’t get grant money from the NSF. Luckily he has no say in what research is conducted.
Just because there are no obvious uses for it today does not mean there will be no uses for it tomorrow. This may be the missing part that makes some other “useless” technology feasible.
How about we just spend trillions of dollars researching whatever we felt like? Just because someone down the line *might* think of a way to use your tech does not justify you building it. Being selective about research does not “slow the progression of science”, infact being unselective does. There are thousands of research projects out there with direct applications which probably will never see funding because they are not from MIT. If you think every whizbang new technique is worth the half million it cost then you need a reality check. This is why all funding is harder to get than it ever was, more and more projects are dead end money pits.
Why do I think this is such a bad idea? First, because it’s 3d printed. This makes large scale production of any of these hydraulic devices extremely unlikely. The printer used is also probably entirely custom with only a couple people in the world who know how to use it. Also, materials are extremely compromised to be able to be extrudable compared to traditionally produced hydraulics. 3D printing is for prototyping- so why develop something that is specifically built not to be a prototype?
“If I can see the heavens it is because I stand on the shoulders of giants”
All of the DARPA, NSF, and SBIR grant / contract money spent each year in the US produces small advances here and there that allow more advanced things to happen. If the US government didn’t offer the money to explore new technologies, then only universities and private corporations would be able to pay for some projects.
This article describes something that a well equipped person could do at home. If this work had been done by a company, you would not see a paper describing the method of manufacture because they would be more interested in making money on the product.
Good thing you weren’t around when they invented the “LASER”.
3D printed microfluidics are already revolutionizing future life-saving medical research.
It takes very little imagination to guess at potential bio-tech uses for this tech.
The idea of printing the enclosed areas with hydraulic fluid pre-loaded is freaking ingenious!
Microfluidics and hydraulics are very different. Everything about microfluidics happens at the scale of a 3D printer while most hydraulics are large scale with emphasis on material strength. Microfluidics are literally just an arrangement of wells and channels in a polymer, the chemistry does the real work.
I look forward to being able to print fully functional devices. The above comments are interesting, involving ferofluids and valves and such, especially the lack thereof in a magnetic system that moves thing magnetically. (far more simple in structure)
I think this has applications far beyond simple bug-bots and such, as it could be used extensively to build everyday objects. They wouldn’t be the same as an assembled part, but you might be able to build the same functional device with different internal technologies and techniques (eg. replace the electromagnetic release on my laptop’s pop-out CD drive with a pneumatic part, with the drive containing pneumatic systems to drive all it’s parts instead of electric)
I recommend skydrol. ;)