Unipolar Carbon Nanotube Muscles: Ten to Thirty Times Human Muscle and Two to Six times Diesel V8

Researchers describe creating powerful, unipolar electrochemical yarn muscles that contract more when driven faster, thereby solving important problems that have limited the applications for these muscles.

This polymer coating converts the normal bipolar actuation of carbon nanotube yarns to unipolar actuation, where the muscle actuates in one direction over the entire stability range of the electrolyte. This long-sought behavior has surprising consequences that make electrochemical carbon nanotube muscles much faster and more powerful.

The advances provide electrochemical unipolar muscles that contract to generate a maximum average output mechanical power per muscle weight of 2.9 watts/gram, which is about 10 times the typical capability of human muscle and about 2.2 times the weight-normalized power capability of a turbocharged V-8 diesel engine.

Above – This scanning electron microscope image shows a coiled unipolar muscle made from carbon nanotubes and coated with poly(sodium 4-styrenesulfonate). The outer coil diameter is approximately 140 microns, about twice that of a human hair.

For more than 15 years, researchers at The University of Texas at Dallas and their collaborators in the U.S., Australia, South Korea and China have fabricated artificial muscles by twisting and coiling carbon nanotube or polymer yarns. When thermally powered, these muscles actuate by contracting their length when heated and returning to their initial length when cooled. Such thermally driven artificial muscles, however, have limitations.

Electrochemically driven carbon nanotube (CNT) muscles provide an alternative approach to meet the growing need for fast, powerful, large-stroke artificial muscles for applications ranging from robotics and heart pumps to morphing clothing.

Powerful Formula
Wang said the team also discovered that unipolar behavior, without scan-rate enhanced strokes, could be obtained when graphene oxide nanoplatelets were incorporated within the yarn muscle using a biscrolling process that UT Dallas researchers created and patented.

“Use of this guest to provide the dipolar fields needed for unipolar behavior increased the maximum average contractile mechanical power output from the muscle to a remarkable 8.2 watts/gram, which is 29 times the maximum capability of the same weight human muscle and about 6.2 times that of a turbocharged V-8 diesel engine,” Wang said.

Science – Unipolar stroke, electroosmotic pump carbon nanotube yarn muscles

Pump it up
Carbon nanotube yarns can be used as electrochemical actuators because infiltration with ions causes a contraction in length and an expansion in diameter. Either positive or negative ions can cause this effect. Chu et al. constructed an all-solid-state muscle that eliminated the need for an electrolyte bath, which may expand the potential for its use in applications. By infiltrating the yarns with charged polymers, the fibers start partially swollen, so the length can increase through the loss of ions. It is thus possible to increase the overall stroke of the muscle. Further, these composite materials show a surprising increase in stroke with scan rate.

Success in making artificial muscles that are faster and more powerful and that provide larger strokes would expand their applications. Electrochemical carbon nanotube yarn muscles are of special interest because of their relatively high energy conversion efficiencies. However, they are bipolar, meaning that they do not monotonically expand or contract over the available potential range. This limits muscle stroke and work capacity. Here, we describe unipolar stroke carbon nanotube yarn muscles in which muscle stroke changes between extreme potentials are additive and muscle stroke substantially increases with increasing potential scan rate. The normal decrease in stroke with increasing scan rate is overwhelmed by a notable increase in effective ion size. Enhanced muscle strokes, contractile work-per-cycle, contractile power densities, and energy conversion efficiencies are obtained for unipolar muscles.

SOURCES- University of Texas at Dallas , Journal Science
Written by Brian Wang, Nextbigfuture.com

20 thoughts on “Unipolar Carbon Nanotube Muscles: Ten to Thirty Times Human Muscle and Two to Six times Diesel V8”

  1. I think he is taking a simplistic view, and instead of considering AI he is talking about rigorously programmed Van Neumann machines instead. The whole point of an AI is that they are not bound by infinitely detailed catchall programming to account for every linguistic and moral eventuality. If we understand what a human is, then a sufficiently advanced AI can also do so easily. The laws were meant by Asimov to represent the unbreakable *principles* by which an advanced AI operates, not a set of instructions written in 3 lines of Basic.

    I notice he also offers no suggestions for improvements.

  2. Wheels for locomotion, but this with soft robotics for arms, so you have a continuum of force—and not just a raw hydraulic yank— save that for last.

  3. And you get an economy of scale thing, the more jobs you fill with android, the better it is to just make a few more androids for a few more jobs. If everyone is going to get a humanoid servant bot anyway to operate their kitchen and clean the house, then an awful lot of humanoid bots are being made. Mass production should be enough that the $30 grand thing starts to look feasible even before you use them to staff any factories.

    Once these things are everywhere, would it really make sense to tool up a separate factory, manage its whole supply chain, and train up a whole different set of neural algorithms to run a spider bot, even if that is technically the best configuration for agricultural weed pulling?

    For that matter, some of our current labor saving tools (and their supply chains) could just be retired. Why give your android a chainsaw and a leaf blower when you can just give it a handsaw and a rake? It's not like it will get bored if the job takes longer.

    The earliest use of the word "robot" in English came from "Rossum's Universal Robots" (robot meant "laborer" in Czech). Much like in the original play, androids could well be the standard issue, universal labor unit.

  4. It depends on cost. If they are relatively cheap say $30k, and very capable, then it makes sense to use them to operate legacy human equipment: limo drivers, heavy equipment operators, some factory work, because you don't have the cost of replacing equipment or retooling your factory. Some farm work, bellhops, busboys, ushers, waiters/waitresses, dishwashers, ship crew, rail-workers, flight attendants, and amusement ride operators/amusement park cleanup people all seem like good fits. 
    And places that could not afford highly skilled human entertainers would get androids if the androids can do that: musicians, dancers, acrobats, jugglers, martial artists, sort of mascots: Ronald MacDonald, Colonel Sanders, that can entertain the kids, or just make things more amusing.
    And the human form is good for jobs where there is a wide variety of actions: yard work, janitorial, eldercare, bartenders, cooks, assistants for the disabled, stunt doubles, lifeguards? and mechanics? And there are people who assemble random things that come in boxes like furniture, that is ordered by a company. Then there will be super athletes and combatants people will pay to watch. And at this point, robocops are looking pretty good. Bouncers and correctional officers too. You can send androids into burning buildings to retrieve pets and such where it does not make sense to risk fire personnel.
    And I am probably missing half of the jobs.

  5. This already applies to modern industrial robots.
    I was working with one this morning. And when you get it to move at full speed you are looking at nearly 100kg of steel arm whipping through the air at the same speed as a well swung golf club, to finish in exactly the precise location within 0.5mm (if you want more accuracy, slow it down a bit).
    Getting into a fist fight with one would be like walking into the propellor of a plane.

  6. So the androids are going to be able to slam us around like rag dolls, the way they can do in most movies.

    One bionic arm won't save you, Will Smith.

  7. I suspect androids are really the best choice only for a super limited niche of roles where the robots have to look like humans (OK. S3xbots. We were all thinking it.)

    In almost every other case a different configuration would probably turn out to be superior.

  8. People deceiving the machines can be a problem, that can turn robots into assassination machines or something.
    But the machines deceiving themselves is where you get humanity itself being destroyed.

  9. Comparing these to Diesel engines, or muscles is an apples to oranges comparison. They are both "prime movers" which oxidize fuel, and output work.
    A better comparison would be with electric motors which convert electrical energy into work.

  10. This is one of the technologies I have long thought was extremely vital for the future. Of course, as others have pointed out, it has to be viable in the real world. I also suspect, that too is insufficient. It probably won't really take off until the patent expires. So optimistically 5-10 more years of science and getting a factory built and robots designed, and mass manufactured. Then 20 years for the patent to expire. So we are looking at 30-35 years for manufacture at the 50m androids/year thing to arrive. And, like I said, that is optimistic. While under patent, maybe they can get to 5 million units a year.
    In the less optimistic direction, the military may become the first customer and be willing to pay a bunch. With the high prices, customers will be limited, and if the government deems this important to keep to ourselves, that will further reduce customers. I would estimate less than 1 million units a year. We would be lucky to get an android in 50 years.
    We sill might get androids (and related forms) before this using stepper motors and linear motors, even pneumatics/hydraulics. Though, the power relative to weight will probably not be great. But if battery technology takes off, like an order of magnitude better energy density, even these clumsy systems would be quite attractive.

    For manufacturing, mining, farming, fishing, and timber, I suspect androids would not be a good fit. Large hydraulic/servo purpose-built machines would just do these things.

  11. He makes some good points. The 3 laws are maybe impossible to translate into language AI can understand, or maybe even humans. The edge cases will derail them at times. Well then, better keep an off-switch handy, probably a remote one too.

  12. In the story, and in life, those laws will fail. Probably not very workable. They are highly abstract, and people are good at deceiving machines.
    It will be like your dog. You are responsible for what it does. You tell it who to obey, and to what extent.
    When it is smarter than you will it still obey you? Probably. Smart people work for idiots every day.

  13. Nothing is ever "real" until it becomes commercialized.
    Apple had the contract with that company that would manufacture nearly unbreakable synthetic sapphire screens for iPhones but they weren't commercially sustainable.
    Today the best you'll get is gorilla glass.

  14. yes, I am depressed because progress is so slow. For me it's way too slow. I will be satisfied when speed of tech/sci progress will be 50-100x faster than today. I hope that advanced narrow AI's and after than AGI, ASI will make this happen soon

  15. Cool, but I remember similar type of developments about which you even wrote here 10 or more years ago, they've never moved beyond lab, sad….

Comments are closed.