Compliant Mechanisms that Roll Like GEARS!?! --
Vložit
- čas přidán 30. 06. 2021
- #VeritasiumContest
I am posting this video in response to a contest announced by Veritasium challenging educators to explain a counter intuitive concept in one minute or less (www.veritasium.com/contest).
Since the prize money is coming from a UCLA professor who lost a physics bet to Derek from Veritasium, this is my attempt as a fellow UCLA professor to redeem my university’s reputation.
This video is about how compliant rolling-contact joints can be used to counter intuitively create compliant lattices of gears. Compliant rolling-contact joints consist of crisscrossing flexures that are deformed around cams. They achieve large ranges of deformation without increasing in stress and achieve near zero stiffness about the desired axis of rotation while maintaining high stiffness in all other directions. 2D and 3D lattices of many shapes can be designed.
My work pertaining to these lattices, which are called ‘Compliant Rolling-contact Architected Materials’ (i.e., CRAMs), is previously published in the journal Nature Communications at the following link:
www.nature.com/articles/s4146...
If you are interested in learning more about compliant mechanisms in general, check out my CZcams Channel, “The FACTs of Mechanical Design”
/ thefactsofmechanicalde...
and be sure to check out Veritasium’s video on the topic as well:
• Why Machines That Bend...
Acknowledge: This work would not have been possible without the help of my students, Luke Shaw, Samira Chizari, Matt Dotson, and Adam Song!
Donate to help support my channel:
If you’d like to make a one-time donation, you can use the following link:
PayPal.me/FACTsMechDesign
Thank you for your support! It is much appreciated and helps enable me to make more content.
Disclaimer: Responsibility for the content of this video is my own. The University of California, Los Angeles is not involved with this channel nor does it endorse its content.
If I understand them correctly, these mechanisms can't actually rotate continuously like a true gear. They can only rotate within a limited range (
correct
true.. they kinda clickbaited us but i think its still cool what theyve shown here
You could probably use this concept to make compliant rolling joints with a much larger range (a few rotations), but that would involve wrapping the flexible bar around the "gears" multiple times, either in a spiral or like a screw.
yeah this, but i'd still be interested in seeing some possible real life applications of those links.
I'm sure there's some clever ways to stack them if you need multiple revolutions of range, but yeah, gonna have a limit as long as things are attached.
"..can move 360 degrees.."
in all designs you showed, if it were to move 360 degrees it would break or heavily deform. Even the one you showed while saying that is 180 degrees of movement, not 360.
It actually moved closer to 320 degrees or so, in a reference frame where one part is fixed (from near parallel on one side, to near parallel on the other side). What it couldn't do is continuous motion, because the flexing part is fixed to the rollers. If it wasn't, we'd have a belt.
In the one demonstrated the top part moved 180 degrees and the bottom part moved 180 degrees the other way, so they two parts can move 360 degrees relative to each other.
@Project Kaizo Trap what's common core math ? radioactive corium meltdown math ?
@@monad_tcp nah, that stuff is too small of a cancer risk to be associated with common core math
@@aDifferentJT The one at 0:27 ? Nope.
Zero-backlash bands have been around for a very long time.
I saw my first one tearing down a floppy-drive in 1985 or so.
The stepper has a drum, band wrapped around drum.
Band ends were on the 'sled'.
Stepper rotates, band drags sled to and fro.
Very precise movement that rack and pinion gearing would have a hard time matching.
I'd love to see that.
Do you happen to know why they switched to leadscrews? That's all I've ever seen in 3.5" floppy drives. I did, however, see a band in a more recent drive. (Iomega Jaz, I think. I only found the linear stage in the recycling, without the enclosure or any other parts, so no labels.)
Metal-band and roller mechanisms like this have been implemented successfully in a few niche applications requiring ultra-low friction, one of which being early automotive airbag sensor switches. Also, sometimes surface grinders use a band and roller to translate rotational motion of the hand wheels into translational motion of the table, presumably because the force vector is always exactly tangent to the path of motion, whereas with a rack and pinion the force vector is not exactly tangent, and moreover it can vary in direction and magnitude over the range of travel if there are any geometrical imperfections, which would be a problem for a precision grinder since imperfections would show up in the workpiece. I have never seen an example of band and roller used to convert rotary motion to rotary motion, but I’m sure there exist machines which could benefit from incorporating this concept
Metal bands have been used in hard disk drives and floppy disk drives head assembly stepper motor drive since long time.
Which surface grinders? Asking cause ive used quite a few of them
@@Ryknfjor I'm not sure. It's something I've read about but not seen in real life. I recently heard that my friend's Harig surface grinder uses a rack and pinion for manual feed, but uses a hydraulic cylinder for auto feed. This may be a more common arrangement than the metal band that I earlier described, and would still be pretty much as good since the hydraulic cylinder should be capable of much smoother motion than the rack and pinion.
@@Ryknfjor @Andrew Phillip Mitsui's use belts like that, among others. You only see them in the super high end grinders for the most part
@@cylosgarage @Ryknfjor That is cool. Yesterday I was looking at some parts made on a tormach SG and the lines from the rack and pinion are very apparent in the finished workpiece
That first demo of two semi circles in articulation could be cool as a load bearing joint in a robot arm. Could use opposing linear actuators on either end to articulate it like muscle on bone. Only challenge I can think of would be whether the surface friction and rigidity of the join constraint would be enough under load.
it does in a way resemble a knuckle joint
@@DeVibe. the idea is a bearing alternative
@@DeVibe. yes, and preferably with justifications
@@DeVibe. in a robotic arm where the rotation is less then 360° how is this sub optimal compared to a bearing?
This is more of one of those 'toothed hinges' you see on fire doors to ensure that they close properly rather than a gear
One of the best videos in this contest.
Здравствуйте, можете подробнее рассказать об этих механизмах на своем канале? Было бы очень здорово
Love this video. I hope this video wins. You have explained it so well.
i really hope your video wins! You made a high quality video which perfectly shows something that looks like it should not work
Woah. Thanks for the level up!
love it! More please!
Amazing extension of motion geometry
The world needs more videos like this! :)
i need a long version of this
Mindblowing!
Thanks so much sir! I´m already working on a prototype linkage system based on CRAMs
Thanks for the content
That's cool. I would be interested in seeing practical applications.
very cool mechanism!
Thank you brilliant ❤
this is so cool !
Ok I need an extended version exploring the full scope of possibilities here
This would be great incorporated into a steering column. It would be super responsive. Probably would have some furious kickback tho.
The single one may make move obvious sense in being restricted like a knee joint. The multiple one seems a little odd, but I could still see some application if it had springs and dampeners to a mostly monostable position. (Earthquake dampener maybe?)
That makes sense.
This is amazing. I love your videos and wish that you were still producing them!
I had a question regarding these "gears" is there a way to design the flexure such that there is a preload force on the gears? I am imagining the flexures with variable thicknesses such that the gears want to roll one way unless a force resists them.
Instead of printing the flexure "belts" straight and then bending them, you could just produce them in the preferential alignment.
that's beautiful...
I want to print these. Super cool.
Whoa! Mind blown!
Great another way to complex my designs
Whoa thats sick
That sounds perfect for skyscrappers joints
Blazing cool!
cool I am impressed
i remember opening up an old radio and seeing a simmilar systemm of strings controlling the frequency button.
// "Since the prize money is coming from a UCLA professor who lost a physics bet to Derek from Veritasium, this is my attempt as a fellow UCLA professor to redeem my university’s reputation."//
LOVE it. Showing UCLA deserves a second chance.
In my opinion, owning up to being wrong and actually paying for the lost bet deserves more respect than being right in the first place, so nothing to redeem here ;).
Edit: Very nice video btw!
@@tpog1
If he truly had owned up to his wrong, I'd agree with you. He didn't. He paid the bet, claimed that Derek won on a technicality and issued another wrong description on how the Blackbird worked (or didn't).
That's why you don't see anymore than the 5 seconds in Derek's video on the topic of Kusenko's payoff -- he was a seriously sore lose and Derek was very kind not to rake him through the coals.
@@johnborton4522 Oh, I didn't know, thanks for clearing that up! Is there a link where I can see/read his complete reaction or do you just know because you're from the same university?
@@tpog1
I watched the entire exchange on Twitter as it went down, then at the end Kusenko deleted his entire Twitter thread and wouldn't respond to anyone on the topic. I'm sure someone saved it off, but I didn't.
I know that Derek commented both on Twitter and in YT comments that he had hoped to do a video with Alex where he explained how he came to his new conclusions, but Alex would have nothing to do with any of it.
Alex was all into public access to the process and science education until it was demonstrated that HE was the one in the wrong and then he just wanted it done and gone.
@@tpog1
I suspect he has a really hard time swallowing his pride after statements like this to the press.
// ""Thanks to the laws of physics, I am not risking anything," he told Vice in an email. "So, I could accept any bet, however large or small the amount might be."
Everything was created in beautiful and most effective way. Why!
Because the creator is The Affectionate and have the knowledge of all things.
He gave us physics to show us how hard and beautiful his creation is, So we love him for giving us all of it.
Awesome!
Mamma Mia, it's no question a subscription!
Great visuals. Now im thinking how something lke this can be seen in real life application
Thats wild!
this si interesting idea for alien technology - for movie/tv shows. and civilisation who never invented screw or gear. I often think about such solutions to make designs more real and alien but realistic.
The concept is the same as a rolamite bearing or a scrollerwheel bearing, but simpler.
Wikipedia articles:
en.wikipedia.org/wiki/Rolamite (for linear motion)
en.wikipedia.org/wiki/Scrollerwheel (for *continuous rotation*)
I don't remember if those articles link to each other, but they should.
That 3d one is bad ass
Me: “Fascinating!”
Me: mind blown
This blew my mind. How much stress can it take compared to the normal gear type? Limb joints on robot possibilities?
not even close to the same amount, it's just good for certain applications
These linkages may wind up having an application in MEMS-based analog computers, if not already.
Los ultimos segundos del video me recordaron a una proteina. Una proteina esta formada por una mezcla de 20 piezas diferentes (como piezas lego).
Segun el orden de las piezas, la cadena comenzara a tomar una forma tridimensional especifica. Una vez formada su figura tridimensional final, solo hay algunos lugares donde es mas probable el movimiento (segun la energia que requiera separar uno, dos, tres o multiples enlaces). Los modelos computacionales son muy poco intuitivos y no existen modelos fisicos (manejables con nuestras manos) que nos permitan predecir con facilidad estos cambios en la proteina. Las proteinas forman micromaquinas que pueden ser puertas, brazos, cerraduras o llaves, agujas, (porfirina), prensas (atp sintetasa), y la capacidad de construir la siguiente generacion de tratamientos medicos esta basada en la comprension de los mecanismos que rigen los cambios conformacionales de las proteinas. Que metodo podria ser bueno para entender y simular estos cambios...? El que nos permite hacerlo con nuestras manos como en los ultimos segundos de tu video. Excelente trabajo!
These guys are out here building biblically accurate angels haha
How much weight can the flextures hold up without bending?
Wow!
Ok, and what use are they in practice? Anything other than fidget toys?
All the good without mentioning all the bad... Also how is it 360° of rotation?
Novel idea, but is there any practical use? That is to ask, does this provide less rolling resistance than typical gears? I also noticed you suggested that plastic is not an ideal material for this and should use metal instead, but wouldn't metal tend to have permanent deformations, as well as a tendency to break? Or at least, that's my experience with stainless steel spoons and copper wire.
Yes there are many applications. I plan to do a future video about this so stay tuned. As for the material, yes if you wanted to do a practical product you'd want to use metal and probably wire EDM it from a sheet. It would work great and be very durable. The reason I used Teflon is because it cuts well and quickly in our cheap laser cutter. It's also good for prototypes and nice demos but I'd never use what I showed in a real product.
@@TheFACTsofMechanicalDesign Yes of course. I knew you did not use teflon in the final product, but is metal a suitable alternative, based on my experience with bending spoons and copper the metal seems to have permanent deformation. On the other hand, I now remember than medieval crossbows seemed to have metal in the bow and those worked, so maybe metal can work. While assuming metal can work, I am still curious to compare the rolling friction of this compared to standard gears.
Yes metal could be made to work, you'd just need to make sure the stress in the straps never exceeds the yield stress of the material of which they are made. It would require thin straps and larger diameter cams. As for friction, the teeth of traditional gears slide against each other during the entire time they are contacting except at the pitch point. But with the compliant gears, they theoretically shouldn't experience any friction but of course in reality they would slightly since nothing is ever perfect.
@@TheFACTsofMechanicalDesign Hmm alright, sounds like a good product.
@@pinkiepie-yv8tk Metal bands are commonly used to convert from rotary to sliding motion in precision but low-force CNC applications (laser cutters, pen plotters, etc.), just like how you could use a string wrapped around a drum to do that. Others here said they were used in the head actuators of old hard drives and floppy drives. So we have plenty of examples of them being durable. AFAICT, you only need to keep the deformation in the elastic region (where it can recover) rather than going into plastic deformation (which is permanent). Practically, for metal band mechanisms, that just means respecting a minimum bend radius. The minimum bend radius and/or the durability is better for thinner metal, just like how stranded wire can withstand more flexing than solid wire, because any given absolute bend radius will be subjectively a gentler bend from the perspective of a thinner material-the bend radius is larger as a multiple of the thickness.
How can it have 360 degrees of motion? Only 180 was shown.
I wish they didn’t cut off the cool mechanisms at the end, I needed more time to see those!!
This is really cool but also it makes me wonder if it's ever possible to make a compliant "gear" that can rotate continuously
Idk. But the video implied things that are just incorrect and might make a person assume that such a thing was shown.
Mostly thinking about the "360 degrees" comment. And the entire premise that this is a replacement for gears when it's a replacement for joints.
the answer is no,
that would just be a pulley and belt system which is not "compliant"
en.wikipedia.org/wiki/Scrollerwheel
And en.wikipedia.org/wiki/Rolamite for linear motion. I don't remember if those two Wikipedia articles link to each other, but they should.
This mechanism is awesome! Does it have an official name?
IIRC I've seen this done before with strings in an experimental RepRap 3D printer, but
I didn't think about using strips of flexible material instead.
I guess this would work well with PET-bottle-plastic and
post-assembly-tensioning with a screw.
Something to try …
---
Also: Maybe this could someday be used for future diamondoid nanomachinery.
Nanoscale defect free diamond can flex >10% and flex is very minimal in this mechanism here.
Or maybe strips of graphene sheets could be used.
RepRap Simpson? The pictures I've seen did show gear teeth, I guess because they're free to include when you're printing the parts, in addition to the strings holding them together.
Yes, gear bearings make things more rigid but are compatible with this. Could be I've seem a version without gearteeth but IDK.
cool..
This is awesome, I wonder if such mechanism would work 3d printed.... Maybe with petg
It does. PLA works fine. However I printed the bands seperately to screw on later. I think the quality will be higher this way. Will try all-in-one like in the video though.
@@benjaminheindl1069 So you print the band flat on the bed?
hmm i think i need to try this out. In my experience PLA is to brittle for such a big flexingrange and would break
Why is the video broken ?
The advantage of "traditional" compliant mechanisms is their precision and lack of backlash/flex in undesired directions. I assume one of these mechanisms cut out of metal on a wire EDM could achieve significantly less backlash than some traditional gear designs, but is it actually better than existing technologies? Also, would this even work with metal?
I want to see an example of a real-world application where this exceeds the performance of whatever mechanism would traditionally be used in that role. Either an increase in reliability, decreased parts count, easier manufacturing, decreased assembly complexity, etc. Otherwise, some novel application where this excels other than novelty toys.
I guess if this is made out of spring steel, it would make for a very reliable linkage with very minimal backlash if any. I can't see it being used without a connecting link between the two center points though, because the discs are basically held in contact to each other with two nearly parallel tensile forces in the bands. I could also see it being expanded to over 360° with bands that twist around two rods like screw threads, instead of hitting their limit after 180° each.
wow
0:35 how can we add this a motor?
Dude this guy isn't even challenging the other competitors. He is challenging the engineering world itself to think harder on things. Good work mate I look forward to your other uploads.
Atlas copco and eimco pneumatic mukers (underground mining loaders) had the bucket with this system... it is difficult to describe just search for pictures
Pulley would match that narrative better, interesting mechanism though.
Oof, laser cutting Teflon is no bueno. PFAS are no joke.
If I may offer a suggestion, plain old polyethylene is probably plenty good for this application. Maybe UHMW if you need it to be particularly dense and lubricious.
Or POM, cuts nicely, low friction, tougher that teflon
whoa
We have constructive gears.
Isnt this just a belt drive with the two pulleys touching?
Looking at this makes me think these will have a limited cycle life. All your stress will be a push pull action where the strap part meets the round part.
This is ingenious. Put some of these models on Thingiverse and identify the right materials to print with?
"most would suspect" said nobody who has seen rubber belts before.
interesting.
Mind blown... oh I guess all the snot on my hands mean my nose was blown instead. Nose blown.
"can rotate 360°" *proceeds to rotate max 180°*
이거 안드로이드? 인간형 로봇 관절에 쓸 방법이 없을까?
This is how original floppy drives would move the head assembly.
I just turned 360 degrees and walked away
В старых жестких дисках актуатор работал на таком же соединении.
And that's how baby Love Bugs are made.
lasercut teflon??? i sure do hope those guys have good fume exhaust
👍
🤯
Technically not clickbait bc of the question mark in the title lmao. So it's not a lie when I ask "Did I just see a swastika on that guys hand?".
Now let's take a moment to remember the technician that got to inhale the vapors coming off the laser that cut those Teflon parts.
Our laser cutter is closed and vents fumes to our fume hood so we should be fine
Next video should be even faster than this one so we can understand even less 👍
Not really a gear, but those last two seconds almost gave me a headrush.
You said 360 degrees at 0:29, but it was clearly 180 degrees.
Both halves rotate 180, giving a total 360. If you held one half fixed at the top, the other could rotate from being left of it with the handle also pointing up to right of it with the handle pointing up again.
@@Pystro never mind, I just understood what you meant. 👍
Next time, end the video with 10 minutes of satisfying footage of these things. Mwah!
This is a variant of the Rolamite mechanism patented in 1969: en.wikipedia.org/wiki/Rolamite
en.wikipedia.org/wiki/Scrollerwheel too. It's similar but for continuous rotation rather than linear motion. (I don't remember if those two Wikipedia articles link to each other, but they should.)
Interesting o0
excuse me the division for anti-witchcraft activity would like a word about that last mechanism there
back in the day when i made my index card magic wallets
We had to make a wallet like that in woodwork at school in 1970's
Rubik's magic puzzle kinda.
Yo, this isn't fair, just kidding, Great use of an Amazing resource as such
I like the part where the guy gets shot and rolls under the car to escape.
This is cool but not at all new. It is on page 163 in the Handbook of Compliant Mechanisms. It is coined there as a compliant rolling-contact element (CORE).
Yes, I'm an author of a chapter in that book. What's new is that we put CORE's in lattices to propose new shape changing architected materials! More to come about these......
I love that. "It's not new I've seen it in a book!"
- Yes, I wrote it
Power move.