Peano-HASEL actuators that mimic natural muscle
HTML-код
- Опубликовано: 3 янв 2018
- The Keplinger Research Group at the University of Colorado Boulder has developed soft electrically powered artificial muscles which feature characteristics remarkably similar to those of natural muscle. These devices, termed Peano-HASEL (Hydraulically Amplified Self-healing ELectrostatic) actuators, can generate powerful linear contraction, operate at high speeds, and demonstrate high optical transparency. Additionally, these soft devices can simultaneously function as actuators and sensors that are capable of precise and life-like movement while self-sensing their position. Peano-HASELs are constructed from inexpensive thin-film plastics, the same material chip bags are made from. Currently, actuators can be made in the lab for less than 10 cents per device, using industrially compatible fabrication methods.
Highlight video of HASEL technology: • HASEL artificial muscl...
Publication in Science Robotics: doi.org/10.1126/scirobotics.a...
Learn more about how HASEL works: • HASEL actuators with m...
Publication of fundamentals of HASEL in Science: doi.org/10.1126/science.aao6139
Keplinger Research Group: www.keplingerresearchgroup.com/
College of Engineering and Applied Science at CU Boulder: www.colorado.edu/engineering/
Mechanical Engineering at CU Boulder: www.colorado.edu/mechanical/
Video produced by Tim Morrissey: / timmorrissey
Wow, what a fantastic idea! This blows EAP actuators out of the water and looks like it could be made for a fraction of the cost. Excellent work 😊
Thanks, Kaleb, Glad you appreciate the work! Please let me know if you have any questions.
This is pretty impressive! I can't help but notice that this might work great as the fluid component of a McKibben pneumatic muscle, as it wouldn't need an external air tank, and is much faster at actuation than compressed air.
Yesss this is it! Thank you! I thought the force between plates in capacitor will not be high enough unless the distance between plates is very small ~20μm (@1kv). Hydraulic system inside and the resulting amplification solves the problem. Any chance to reduce the voltage to ~50v? 10kv is not very convenient. You could design much more efficient amplification than the one presented in the video.
Hi Piotr,
Thanks for the great questions and comments. Glad you appreciate the work. Yes we are actively working on lower voltage but 50v is pretty extreme but is a great goal. Do you work in the industry? If so I would love to chat with you directly about this work. Please reach out to me at info@tgmorrissey.com if interested.
Hi Tim, I am really impressed with how you are reaching out to people commenting here. I haven't read the paper (yet!) but is this technology something that people could build at home? What are the chances that "open source" plans might be released to increase the number of people/groups experimenting with this technology?
I'm also curious, if you could have two different circuits in one cell as a means to change the muscle's deformation (one circuit for linear actuation like seen here and another circuit to cause the "donut" actuation)?
If you are handy and like building you could certainly make a simple HASEL in the home. The power supply is slightly more complex but some people are familiar enough with HV power supplies are more than capable of making this work in the home. There is an inexpensive power supply on amazon www.amazon.com/Vipe-3-6V-6V-Step-up-Voltage-Generator/dp/B01MQMSV95
These are not robust but are fun to play with, be careful!
Check out the papers and let me know if you have more questions. Best of luck!
Very interesting! I have been following the field of EAPs for over a decade now, but I have not yet seen them move out of the research stage. Ive had some EAP-fluidic coupled ideas myself but they always proved impractical; but the simplicity and effectiveness of your approach is striking at first glance! Are you aware of any attempted commercialisation of your actuators; or are you aware of any research challenges that have yet to be overcome? There is no shortage of applications I can think of, if it works as well as it seems.
Hey Eelco, Thanks for the note. We are commercializing the actuators through a startup, ArtimusRobotics.com, feel free to reach out to me if you want to discuss more.
@@tgmorrissey Thanks for the link; i subscribed to the mailing list. Looking forward to what you guys can realise in terms of relevant technical specifications!
@@eelcohoogendoorn8044 For sure! Question for you. What specs are most relevant for you and your application?
@@tgmorrissey Some questions that came to mind are longevity under repeated stress (seems like some areas of the membrane are under quite high strain in the concept implementation; and this may not be that easy to design around),. Also the overall attainable energy efficiency including going from battery to the required voltage would be good to have insight on. My apologies if your paper answers those issues; I dont currently have access to it.
@@Daniel-oz5dm Contact info here: www.keplingerresearchgroup.com/
Okay, so this is going to be a strange question, but how effective would this be for large scale robotics - say... on the scale of about ten meters for a humanoid robot? My main thoughts are reinforcing the bag against the stresses and loads and optimizing the fluid. It'd probably be more costly, but that makes sense considering the application. This could be some crazy stuff if applicable on a large scale. From my understanding of other artificial muscle technologies they can be very strong but are limited in their speed (though this could be to reduce the cost of the experiment and not indicative of the actual capabilities of the technology). This seems to eliminate that problem while also being strong in its own right. I've always had a desire for real life mechs (like from BattleTech or MechWarrior) to be possible. Of course they don't really have much usefulness on the battlefield - they're just awesome. And if this can be used on large scale robots, that kind of technology could be possible.
How efficient are these actuators?
These actuators can be up to 19.4% efficient according to publication in Advanced Science "An Easy-to-Implement Toolkit to Create Versatile and High-Performance HASEL Actuators for Untethered Soft Robots".
Hydroelectric energy source could afford the cost
Can any dielectric liquid work?
Yes, any dielectric liquid can work, you might run into compatibility issues between the elastomer and liquid, luckily there are literally hundreds or thousands of possible liquids. So glad you are interested in our research. How are you so familiar with liquid dielectrics.
Tim Morrissey I actually found out about dielectric liquids from reading about your project. It's a really exciting idea given cost to performance. I'm highly hopeful it'll lead to lower cost robotics. Looking forward to more and thinking about looking into this myself.
Absolutely rad. Glad you find it interesting. Feel free to reach out with any other questions and stay on the lookout for more work in the coming months.
I read into your paper, any suggestions for low cost dielectric liquids to start with? FR3 isn't very easily obtained given my location and budget. I've been thinking of trying to make muscles using mineral oil, aluminum foil affixed to thin plastic which is welded together to form pouches of various geometries, and a high voltage power supply. Think it's a good idea? Or would you recommend I start with other material? This is really exciting stuff and I'd love to use it for robotics.
TJ. Awesome that you are looking into this more. FR3 is just one of the many liquids that could work. The vast majority of oils are non-conductive and should work to some extent. We actually used canola oil purchased from the local grocery store on our first attempts which should help you get started.
Be careful with the high voltage power supply!
How heavy are those weights?
Hi! This is really awesome, and I am really excited by just imagining all the applications that your research could have. Do you have access to a 3d printer? If so, have you ever heard about inmoov? It's a full size android that is completely open source and it's parts can be downloaded from thingiverse.com. I'm not promoting it at all, but if you are interested you could adapt that existing design to use this actuators and bring the future even closer to us
Thermoelectric Contraction Muscles are better they are totally silent operation, extremely strong, lightweight, compact size and work at normal levels of voltage between 12 to 180V not 8000-15000V what results dangerous.
They are Not better for a lot of applications as the lifetime as Well as the cycle time leave to be desired in comparison I believe