I can say from personal experience that a steel screw dropped onto a tile floor has a coefficient of restitution of at least 1, and this coefficient value is proportional to the rarity and cost of the screw.
@Tim Elsen In a theoretical world where 1 isn't the upper bound, it would bounce higher than the height it was dropped from, if the coefficient of restitution was 2, and you dropped a ball from 10 meters, it would bounce up 20 meters. However this theoretical world is what you are transferred to when you drop one of these rare screws and the screw bounces for eternity somewhere in your workspace.
I don't think you fully appreciate the results of your "research". The sample screw has a post impact path at an angle to the original line of travel. This requires more kinetic energy than the system has before the collision. You are describing a type of hot fusion. I say it's hot fusion since that's what you become after the screw disappears. And, to distinguish this phenomenon from plain old hot fusion it's spelled Fyousion.
As an engineer we call the balls ball bearings and the assembly simply a bearing. As a bearing consists usually of an inner and outer bearing run with the bearings working between being ball bearings needle bearings tapered bearings ect.
@cr Maybe, but they are pretty spherical in general , that wasn't my point in the first place, in your logic there aren't any soccer balls because they aren't perfect spheres, i mean hell nothing is ever a ball, it's impossible to get anything to be a ball, just very close to it. So we use the word "ball" for conveniece. And if you measure ball bearings from roller bearings with something like a micrometer, you will see that they are very very equal on all sides, the taper is indistinguishable to the human eye. I think you got confused with tapered roller bearings or maybe some other type of bearings, but your typical roller bearings which are angular contact roller bearings and deep grove roller bearings pretty much have balls inside of them.
@Call_Me_Madu Um, no, roller bearings have cylindrical (or sometimes slightly tapered) rollers in them, never balls. There's even 'spherical roller bearings' which have barrel-shaped rollers in them (so-called because the envelope of the outside of the barrel rollers forms a virtual sphere, so they're tolerant of misalignment), but not a ball anywhere in sight.
One think you should look into for further improvement is to place the rig on an anvil or some other very large hardened steel thing. fun fact, blacksmith anvils are commonly tested for their quality by bouncing a ball bearing and seeing how high it bounces the ball back up. The higher the bounce, the better it will be to forge on.
When the frame rate makes the bearings appear to hover is so cool! Bouncing in and out of the plane of focus is really cool too. Well done. You look like a week off is in order.
Would you be interested in using some of the sound files and measurements to calculate the gravitational constant? Not only could it be a relatively cheap way to calculate g, I am curious if it could measure the gravity gradient by performing the bounce tests at various heights.
You're going to have to assume that the collision time (the time the ball bearing spends in contact with the surface) is constant, which might introduce some error. A second source of error will be due to the sampling frequency of the audio stream. You could probably get a pretty nice value out though.
If you are testing number of bounces again, samplers like the one in Logic Pro x can easily tell you by loading in the audio file and using transients/peaks as the metric that decides chops/divisions. then it will may tell you how many it was able to locate (dependent on the software it should show up pretty apparently, I know logic’s quick sampler does)
That was genuinely fascinating. I found myself dying to see it tried in a vacuum, until you hit that very subject. XD I think you would have gotten a more dramatic difference with the smaller balls in a vacuum if, as you said, air resistance was the main factor limiting them. I went looking for the Caltech paper to see what the final alloy was and found the simulations they ran to find the best conditions fascinating as well, and slow-cooling molten alloys in a centrifuge at 60,000 g kind of broke me for a second.
I think you might have stumbled upon the secret to the euler discs. The sound the ball makes when it's almost done is similar to the disks when they are almost done. Idk worth looking into though
Hello Steve! Wow, what a great video! 👏 I'm a researcher, and I'm coincidentally finishing my PhD in this research field. We used to call it Bulk Metallic Glasses (BMGs), or just metallic glasses, as you mentioned. In all these years that I've been studying this field, your way was one of the most interesting to present the public that I've ever watched. I'm so happy to see something I study being presented so well to so many people due to the high visibility of your channel! It made my day! You commented on the possibility of someone sending you spheres also made of metallic glass (and yes, you are correct, it would be the ideal way to maximize the bounce time). In the laboratory where I work, we usually produce several amorphous metal alloys, the biggest difficulty is only in the spherical shape, since to get an amorphous structure, as you said, we need a certain cooling rate, and that's why we suck our samples to a chilled mold. I am already in contact with my advisor, so that we can study the possibility of producing this type of spherical samples and send them to you! Hope this message reaches you! (Help me with a like guys!)
Perhaps someone else has mentioned this but I noticed that no where in the video did you mention the sound itself as an energy leak. When you created the vacuum test, the sound was quite apparent meaning that each bounce sent a compression wave through the metal disk, then the base metal cylinder, then the brick, then the metal plate, then the air gap, then the microphone. That has to be a major loss of energy, possibly more that any from air resistance.
You are correct sir. In terms of entropy the base is at a state of rest and the ball is at an energy level higher than it's resting state. Energy will go from higher to lower. So the impact energy transfer will be many times higher than the energy transferred to the air. The ball shaped also contributes to this.
I am amazed by how we could hear the bearing hitting the plate in the last trial with the partial vacuum. Our ear is sensitive enough to detect the vibration through the plate and substrate metal to the brick which vibrates and in doing so compresses the air and sends waves of compression to our outer ear which channels and amplifies these waves to our inner ear and so into our brain so we can 'hear' the vibration. Incredible.
you should do one on spinning tops. I have a brass one with a ruby tip that I hit 14 minute spins with on a slightly concave mirror. I was always trying different things to get better times. Super fun.
Love the bounciness! Would you be able to explain in another video why the balls tend to bounce back to the center of the trampoline? I have an idea, but after I lost our argument last time, I won't share it just in case!
@Andreas No, that's not it. A 2D random walk eventually ends up in *any* point in a lattice, not just the starting point. In this case we can clearly see the ball tending towards the center instead of moving fully randomly. The surface must be concave (and that's what Steve Mould said here in the comments).
I believe it is based on the angle of deflection of the base surface. Consider a 2d model of the circular plate as a single beam. If the bounces perfectly central the slope of each side of the beam are equal resulting in a perfectly vertical rebound. However if the ball lands off centre then one side is shorter than the other, resulting in a slightly steeper slope on the shorter length side. This will result in a bias to deflect the ball away from the shorter side to wards the centre of the beam. Comments please.
I believe it is based on the angle of deflection of the base surface. Consider a 2d model of the circular plate as a single beam. If the bounces perfectly central the slope of each side of the beam are equal resulting in a perfectly vertical rebound. However if the ball lands off centre then one side is shorter than the other, resulting in a slightly steeper slope on the shorter length side. This will result in a bias to deflect the ball away from the shorter side to wards the centre of the beam. Comments please.
so what if you had a metal made from a single "perfect" crystal of metal? would that then become a superior bouncy surface than a standard metal that contains imperfections?
My initial thought about the breeze block was that the rough surface would make it increasingly hard to level everything perfectly so that the balls don't wander off. I wonder how difficult that was. My second thought was that the porous breeze block would probably absorb a lot of energy compared to something more dense.
This is cool, that material must be incredibly hard ont he rockwell scale. You can get a similar, if not as good, effect with heat treated metals with very high hardness. A tungsten carbide disk brazed to a larger steel billet would probably be really close if you wanted to homebrew one.
What happens if you use two surfaces? It would be interesting to see two experiments one where the top surface is stationed above the balls initial drop. Then another experiment with the surface placed into the balls path upon its return upwards.
Would be interesting to derive the cubic/quadratic ratio and test it at different pressure to see which one fit's the theoretical value the best or worst.
Steve, did you consider that the magnet ball release technique might have influenced your 1 minute plus run? I mean intrinsic temporary magnetism or similar?
I just got this video recommended by the YT. Really good video. Production value, recording quality and the writing and delivery are spot on! I would have loved to see that vacuum pump used for a demonstration between the extremes of your experiment (smallest and biggest bearing ball (I just learned that, thanks!) of 2 or more materials). Just because I am intrigued: can someone tell me if you could calculate the "lost" energy from the loudness/pitch of the sound it makes? Also, to get an amorphous metal ball, maybe challenge some chemistry KZcliprs like Nilered to try and make it! I can imagine it might be fun for them and also a great video for us!
have you tested the bounce time with air AND with the glass cover on top? Maybe this is a factor, why in vacuum it doesnt bounce that much longer (similar to the change to put the device on a brick)
The ruby bearing is especially wear resistant. These are used as the point [of contact] in performance tops that are designed to spin as long as possible.
Hey Steve. Not sure if you'll ever see this, but do the bounces of the ball bearing follow the 80/20 rule over time? As in, do 80% of the bounces occur in the final 20% of the time?
If it helps, the outer and inner parts of a ball bearing are called races. The reason they call it a ball bearing is to denote the type of bearing inside the races, as there are tapered bearings and rod bearings as examples of alternatives. Awesome video, thank you for sharing.
Hi Steve, Have you made a video about Nitinol? It is called a "memory metal" because, although it can deform plastically, it can also be gotten to return to its original shape. I'd love you to explain how that is possible. Thanks.
Looking at the AISI steel bearing size test, the 5mm one seems to be an anomalous as the bearings on both the larger and smaller side of it do not bounce as well, yet the 7mm one fits well with the 8 and 6 mm and is almost as great as the 5mm. I think this would be obvious if you plotted the size versus time. Watching this reminds me of the superballs from my youth.
Yes a great illusion. Thumbnail gave the illusion that a ball bearing could bounce on amorphous metal 259 times yet only it was only 45 times. Grand ! Amazing ! Warms my heart too ! Just loved it when I wasted 18 minutes to have the illusion shattered.
Tim is great. I can honestly say I've enjoyed every single show of his. I like how a lot of my subs know each other, and work together. They've even flown from the US to do a show with Steve.
There is an alternative audio software to audacity called reaper which allows you to detect and at the same time count so called "transients" sharp amplitude peaks in audio signals. The function is called "dynamic split items" if you wanna try it out. Reaper is free for a 30 day trial and after that still very affordable at only around 60 bucks or so. So feel free to try it out the next time you need something like that :)
Excellent video. When you pressed your thumb on the artist's eraser around 10 min in it triggered an olfactory memory and I could smell the eraser from my childhood. Fascinating.
Cool video, LiquidMetal Technologies made a video 11 years ago with a stainless bearing bouncing on their material for 81 seconds. Its cool that there are more companies researching amorphous solids. It would be awesome to see an amorphous bearing against the amorphous plate.
Great vid! Watching it made me think of a bounce test, with a steel ball bearing on perfectly flat anvil. The blacksmith in question was extememely impressed on how long the bearing bounced, as well as how much he could conserve in an arm swing with his hammer. It totally sits in line with the results in your vid. As we know, energy cannot be lost, but transferred into an alternative form - sound, heat, light, friction, deformation, etc. I would love to see the results with a hardened vehicle on a flattened, hardened surface (due to the atomic gaps being smaller) as well as on a hardened amorphous (if it is possible) material. Is it physically/atomically possible to revolutionise the ol' hammer? Food for thought!
If you can find a disc of AM-III amorphous glass of a similar size and a ball bearing of AM-III as well, I think you might be able to beat your record.
A long time ago there was a distinction between roller-bearings and ball-bearings. The rolling elements of roller-bearings were tiny cylinders rather than tiny balls. I believe there was another type where each rolling element was machined to be two cones joined base-to-base, pointy-ends outwards. The curved surfaces might not have been the cones you get by taking a gore out of a flat surface, but, rather, curved a bit inwards or outwards. The bigger elements, the circles inside which the bearings roll, could have had quite a complex geometry since ordinarily a cone will not want to roll STRAIGHT on a flat surface, and if forced to do so will have a lot of friction from the forced skidding. I'm not sure how all of that would've been worked out.
Plasticity changes a lot with the change in temperature. It would be great to see how frozen metals compare to room temperature metals in this test. Steve probably has liquid nitrogen next to his bed, so he could make a test and let us know how it worked.
@Jarek Ferenc It's all about hardness, which can change dramatically with temperature. A very cold thing can become very hard and brittle, and could allow for much cleaner bounces because of this. Make it cold enough and the bearing heavy enough, and something will shatter instead of bouncing, and I expect this to be a very elastic collision right before that transition. There are a whole series of experiments about bounce height and material hardness and strength you could do here.
@NO WAY My gut feeling is telling me: a bit. A very tiny bit. Cooling down to -40°C will not increase the yield point significantly. The BMGs exhibit the yield point (if they undergo any plastic strain) at some 1000 MPa, which is incredibly high, so the stress at impact is usually within the elastic range. Thus, lowering the temperature to -40°C will have a negligible effect. BTW, if iron-based BMGs are considered instead of zirconium-based ones (probably they were used in this film), they would withstand the compressive stress of 4000 MPa or even higher. Making the amorphous plate 2 mm thick is possible even if it is iron-based.
That was the variable I hoped he was talking about! And then he pulled out multiple _sized_ bearings, and I was like, Ok, sure, they all have different stiffnesses and air resistances, so you'll get interesting results, but I wanted to see liquid nitrogen and blow torches!
Wonder if you could calculate the speed of the shock wave travelling through the base material. If it is sitting on a material significantly stiffer, then the reflected shockwave could form a constructive interference pattern in sync with the ball bearing contacts. Would provide the most benefit at the beginning of the bounce I presume. What about temps?
to count all but the audio rate bounces, you can use audacity’s sound finder feature (i think that’s what it’s called, and it’s in one of the last three menus). it creates labels which are numbered in a separate label track. you can also slow the audio down and then run the sound finder, to give it an easier time finding the bounces.
Sorry if you cover this later, but while you were talking about the different materials, I wondered about how Amorphous Ball would interact with Amorphous Disc??... Very Interesting stuff btw🤔
As a mechanic without an engineering degree, I'm curious if temperature would have an effect? Would heating or cooling either the materials of the bearing and the impact plate or the air within the glass column make a measurable change? Also as a mechanic who deals with bearings in the field and not a lab, would impurities (finger oils or dust) affect results? I've dealt with bearings that a dirty finger print would affect it's performance, although that is probably more a rotational force.
Running tests with the system at different temperatures would be interesting. A heated ball would have more energy to start with. But, the electrons in the balls metal could be more free to move around. If it's cooled, the electrons are more tightly held but over all the ball has lower energy. Or, heat transfer with the air would cancel out any temperature change. A cool line of inquiry no matter how it turns out.
Steve, Little tip of advice from an audio engineer. If you ever find yourself needing to count the peaks of a audio waveform again. Save your time and eyes, Look up Dynamic Splitting. You essentially set a threshold above the noise floor and it will detect any transients (peaks) above your setting and then splits them all. Resulting in each transient being its own audio "file". Highlight all files and see how many you have selected, or a good Slicer will tell you how many transients are detected before actually splitting for you.
Maybe there is also a sort of resonat effect between the ball and the ground plate, that can be used to improve the duration (theoretically): The ball has its own vibration, it wobbles (11:40) and it does it at a very specific frequency (resonance frequency). the same applies to the base plate. If both have the same resonant frequency and both oscillate synchronously after the initial beat, this could possibly have an effect. Or even if the resonant frequency is different, but in such a way that they influence each other positively at the moment of collision, I think the system will oscillate as follows: When the swinging ball hits the also swinging base plate and that happens exactly at the moment when both are compressed (the vertical expansion of both is at the minimum ), or even a little before reaching the minimum (as in the combustion engine: just before the dead center is reached), so that both make contact while each is still in the contraction phase, so the impact energy is still something for both squeezing a little bit more. It's like a swinging spring that gets an extra impulse just before it reverses direction.
Will Stelter showed a method of testing anvils similar to this, the frequency of bounces toward the end is mesmerising. It's as if the bearing is momentarily floating
From what I gather they were made as an Educational demo. So not available commercially in this format. The company (as a spin off) still manufactures items from this material and other similar ones.
Great video Steve, interesting subject! However, at 9:41 you mention 'the energy is now lost'. Is it really though? ;) Also, I like the mention of Grand Illusions! Been following them for years, it's always fun to see Tim showcase some quirky toy.
You might increase the stiffness by chilling to a low temperature. To avoid moisture condensation, that would necessitate running the experiment in a dry gas (nitrogen and argon are probably the easiest) or vacuum.
The thing that came to my mind when I was watching this was it similarity to Euler's Disks. I would look into the the technology and study behind those because they are just as dependent on material sciences and conservation of energy and you might find some useful information about this topic.
it would be interesting to retest the smallest ball bearings you had in the vacuum chamber to see if there is a large gain with them as you had already selected for the ones that were least effected by air resistance
One other kind of ball bearing material I'd like to see tried is CBN (Cubic boron nitride), being the 2nd hardest material there is after diamond.. Quartz could be interesting as well.. unfortunately I haven't found any sources for them. CBN is often used in machining very hard metals, it's pretty amazing stuff
@Rx7man In theory, the stronger the crystal lattice the harder the material, the problem with ceramic materials it's that it's not structure of single crystal defines their properties but also how crystals are packed in the material. I googled for microscopic structures of CBN and ZrO2 and the CBN has much more empty spaces in structure which can lead to friction between crystals when bouncing and loss of energy. If we could create perfect monocrystalline balls of both material then CBN would be a winner.
@Stewart Ross I think the density would be minor in comparison to hardness.. density would mitigate the air friction a little but not much more.. try lead and gold.. heck, try mercury :)
When I was watching Tourette's Guy on KZclip 15 years ago, I never imagined I'd eventually be watching some guy on here spend an incredible amount of time and energy and considerable expense into seeing how long he could make various bearing balls bounce. I mean ball bearings bounce.
Does the temperature of the materials impact their restitution coefficient? Would make the experiment more difficult for sure though, making sure to keep other variables the same.
the way to get a better coefficient of restitution is to size and shape the anvil so that it reflects all of the sound energy back to the ball before the ball leaves, rather than having that energy come out as the sound we hear. also remember that "super balls" are hollow, again to ensure that the elastic energy goes back into its bounce rather than to a vibrational mode of the ball itself (but good luck getting a hollow bearing ball !).
Even if the ball did bounce an infinite number of times, the duration of each bounce decreases geometrically so the total time taken is limited, kind of like Zeno's "paradox".
I wanted to be a materials scientist when I was little, until I got a little older and realized that my days would be spent doing things like this all day long. It's much more enjoyable to watch you do it for us, lol.
Is there a known value of Young’s modulus for this metal? You mentioned there was a difference between stiffness and elasticity but I thought that they were the same or similar
That amorphous metal might be useful in musical instruments like guitars for lengthening the duration a string vibrates, while maintaining its amplitude for as long as possible. That material could be used for the Bridge Saddle and the Nut. 🤔🤓
So, the amorphous steel just sounds like hardened steel from what you're describing. If you want an amorphous bearing you should take one of those 52100 bearings, heat it to about 1500 degrees freedomheight, hold it there a minute, then quench it in oil. Vegetable oil or peanut oil is just fine.
Your voice is so soothing. Just the tonality and your speech/cadence, it's really calming. Also, this was so much new information. I've also never seen anything so strange and satisfying as these tiny metal/ceramic balls bouncing way more than they have any business doing. Even the sounds are immensely satisfying, the pops from each bounce. I wonder if there's some kind of application for this ultra-bouncy metal, in harnessing kinetic energy. "Every bounce charges my phone 0.003%" lol. And man, I wish KiwiCo had been around when I was a kid. That subscription service sounds exactly like something I'd have loved as a kid, and I'm positive I'd have learned quite a lot from it.
@12.06 could the sound generated echo from the lower contact surface and shoot up and give the force additional energy to keep it running for so long? Which brings to question? do you have any sound density/speed, data for resonance etc... that would be interesting to see.
I created a jig to test COR in university as a project. After testing a number of materials, was very surprised(as was the class when I demonstrated) that cast acrylic had an extremely high COR. I don't know why but everyone assumed it would dampen the bounce.
The 5mm balls used for 'bb gun' ammunition are very bouncy on a hard surface (floor tiling is best but also very bouncy on wood laminate flooring). I think they are also some sort of cast plastic. Used the property for a 'door trap' to make sure no-one was sneaking into a room of mine before - count the balls load them into a pot that is released if the door is openend; because they bounce all over the place it's hardly likely they would all be found and replaced so you know if the trap was sprung (count the balls in the trap without setting it off). Unfortunately I forgot I'd set it one time, and ended up with balls bouncing everywhere on entry. 🙄
Yeah, you can get acrylic balls at TAP that bounce like really noisy superballs... they also don't really warm up as they bounce. Unless they crack or chip, they really don't stay deformed. PLA on the other hand... "sploot"
It feels like a glitch in the matrix.
The sponsor is KiwiCo: Get 50% off ANY crate kiwico.com/stevemould50
for drop line
Please calculate sin, cos, tag angle and difference and track pattern
Can this metal glass be used to make pistons?
@Krabby Patty i
100th comment
That was really cool. This has made me want to try and make one of these... If I succeed, I'll send you one!
Yoo you’re the same guy that turned vinyl gloves and vanilla into hot sauce. Hey man 👋🏻
@Alex Hu it’ll probably be 4
Amazing to see you here, love the channel
Makes me wonder if you could make a ping pong paddle surface out of amorphous metal.
Can you collaborate with electroboom to make something dangerous?
I can say from personal experience that a steel screw dropped onto a tile floor has a coefficient of restitution of at least 1, and this coefficient value is proportional to the rarity and cost of the screw.
@Tim Elsen In a theoretical world where 1 isn't the upper bound, it would bounce higher than the height it was dropped from, if the coefficient of restitution was 2, and you dropped a ball from 10 meters, it would bounce up 20 meters. However this theoretical world is what you are transferred to when you drop one of these rare screws and the screw bounces for eternity somewhere in your workspace.
1 is the upper bound!
@Lyxe It was a joke, but I did not understand it either.
It can't be more than 1
I don't think you fully appreciate the results of your "research". The sample screw has a post impact path at an angle to the original line of travel. This requires more kinetic energy than the system has before the collision. You are describing a type of hot fusion.
I say it's hot fusion since that's what you become after the screw disappears. And, to distinguish this phenomenon from plain old hot fusion it's spelled Fyousion.
As an engineer we call the balls ball bearings and the assembly simply a bearing. As a bearing consists usually of an inner and outer bearing run with the bearings working between being ball bearings needle bearings tapered bearings ect.
@cr Maybe, but they are pretty spherical in general , that wasn't my point in the first place, in your logic there aren't any soccer balls because they aren't perfect spheres, i mean hell nothing is ever a ball, it's impossible to get anything to be a ball, just very close to it. So we use the word "ball" for conveniece.
And if you measure ball bearings from roller bearings with something like a micrometer, you will see that they are very very equal on all sides, the taper is indistinguishable to the human eye.
I think you got confused with tapered roller bearings or maybe some other type of bearings, but your typical roller bearings which are angular contact roller bearings and deep grove roller bearings pretty much have balls inside of them.
@Call_Me_Madu Um, no, roller bearings have cylindrical (or sometimes slightly tapered) rollers in them, never balls. There's even 'spherical roller bearings' which have barrel-shaped rollers in them (so-called because the envelope of the outside of the barrel rollers forms a virtual sphere, so they're tolerant of misalignment), but not a ball anywhere in sight.
@Reid Prichard Well i guess each to their own, but having 2 things having the same name confuses me a lot so i like to seperate them in my own way
@Call_Me_Madu hmm, when I hear "roller bearing" I think of cylindrical/tapered roller bearings.
I've never been so sincerely interested in a bouncing ball before
It would be even more exciting to watch paint dry... on the molecular scale... timelapsed
One think you should look into for further improvement is to place the rig on an anvil or some other very large hardened steel thing. fun fact, blacksmith anvils are commonly tested for their quality by bouncing a ball bearing and seeing how high it bounces the ball back up. The higher the bounce, the better it will be to forge on.
That’s for the same reason as this. Highly crystalline metals are brittle, which is bad for an anvil.
When the frame rate makes the bearings appear to hover is so cool! Bouncing in and out of the plane of focus is really cool too. Well done. You look like a week off is in order.
Ah! You should've tested the smaller ball bearings again in the vacuum tube, as they were most affected by air resistance.
That’s exactly what I was thinking!
Just thought exactly the same!!!
Thanks for saving my time I thought that’s where he was going with it
I think the square cube law would still come into play
Damn, I think I walked in on nerd con
Would you be interested in using some of the sound files and measurements to calculate the gravitational constant? Not only could it be a relatively cheap way to calculate g, I am curious if it could measure the gravity gradient by performing the bounce tests at various heights.
You're going to have to assume that the collision time (the time the ball bearing spends in contact with the surface) is constant, which might introduce some error.
A second source of error will be due to the sampling frequency of the audio stream.
You could probably get a pretty nice value out though.
If you are testing number of bounces again, samplers like the one in Logic Pro x can easily tell you by loading in the audio file and using transients/peaks as the metric that decides chops/divisions. then it will may tell you how many it was able to locate (dependent on the software it should show up pretty apparently, I know logic’s quick sampler does)
That was genuinely fascinating. I found myself dying to see it tried in a vacuum, until you hit that very subject. XD I think you would have gotten a more dramatic difference with the smaller balls in a vacuum if, as you said, air resistance was the main factor limiting them.
I went looking for the Caltech paper to see what the final alloy was and found the simulations they ran to find the best conditions fascinating as well, and slow-cooling molten alloys in a centrifuge at 60,000 g kind of broke me for a second.
I feel like I learn more and tire less, watching Steve than anyone I have ever listened to. Supremely fascinating! Kudos to Mehdi as well!
I think you might have stumbled upon the secret to the euler discs. The sound the ball makes when it's almost done is similar to the disks when they are almost done. Idk worth looking into though
Hello Steve!
Wow, what a great video! 👏
I'm a researcher, and I'm coincidentally finishing my PhD in this research field. We used to call it Bulk Metallic Glasses (BMGs), or just metallic glasses, as you mentioned.
In all these years that I've been studying this field, your way was one of the most interesting to present the public that I've ever watched. I'm so happy to see something I study being presented so well to so many people due to the high visibility of your channel! It made my day!
You commented on the possibility of someone sending you spheres also made of metallic glass (and yes, you are correct, it would be the ideal way to maximize the bounce time). In the laboratory where I work, we usually produce several amorphous metal alloys, the biggest difficulty is only in the spherical shape, since to get an amorphous structure, as you said, we need a certain cooling rate, and that's why we suck our samples to a chilled mold.
I am already in contact with my advisor, so that we can study the possibility of producing this type of spherical samples and send them to you!
Hope this message reaches you! (Help me with a like guys!)
Bump!
Bump
Bump for science!
Bœrmp
Oooooooooh! Great idea. Some percussion. Xylophonic instrument with triggers and undulating reverbs.
Perhaps someone else has mentioned this but I noticed that no where in the video did you mention the sound itself as an energy leak. When you created the vacuum test, the sound was quite apparent meaning that each bounce sent a compression wave through the metal disk, then the base metal cylinder, then the brick, then the metal plate, then the air gap, then the microphone. That has to be a major loss of energy, possibly more that any from air resistance.
You are correct sir. In terms of entropy the base is at a state of rest and the ball is at an energy level higher than it's resting state. Energy will go from higher to lower. So the impact energy transfer will be many times higher than the energy transferred to the air. The ball shaped also contributes to this.
While he didn't specifically point out the sound as energy loss, he did reference the different energy losses based on the base that it is on.
I am amazed by how we could hear the bearing hitting the plate in the last trial with the partial vacuum. Our ear is sensitive enough to detect the vibration through the plate and substrate metal to the brick which vibrates and in doing so compresses the air and sends waves of compression to our outer ear which channels and amplifies these waves to our inner ear and so into our brain so we can 'hear' the vibration. Incredible.
GrandIllusions is one of my favorite channels. Awesome to see their own atomic trampoline in your video Steve!
The nerding out never ever ends in this one, I love it, very satisfying.
you should do one on spinning tops. I have a brass one with a ruby tip that I hit 14 minute spins with on a slightly concave mirror. I was always trying different things to get better times. Super fun.
Love the bounciness! Would you be able to explain in another video why the balls tend to bounce back to the center of the trampoline? I have an idea, but after I lost our argument last time, I won't share it just in case!
@Andreas No, that's not it. A 2D random walk eventually ends up in *any* point in a lattice, not just the starting point. In this case we can clearly see the ball tending towards the center instead of moving fully randomly. The surface must be concave (and that's what Steve Mould said here in the comments).
Could be just because 2D random walk ends up at the starting point math.uchicago.edu/~may/REU2018/REUPapers/Ombrellaro.pdf
I believe it is based on the angle of deflection of the base surface.
Consider a 2d model of the circular plate as a single beam.
If the bounces perfectly central the slope of each side of the beam are equal resulting in a perfectly vertical rebound.
However if the ball lands off centre then one side is shorter than the other, resulting in a slightly steeper slope on the shorter length side. This will result in a bias to deflect the ball away from the shorter side to wards the centre of the beam.
Comments please.
I believe it is based on the angle of deflection of the base surface.
Consider a 2d model of the circular plate as a single beam.
If the bounces perfectly central the slope of each side of the beam are equal resulting in a perfectly vertical rebound.
However if the ball lands off centre then one side is shorter than the other, resulting in a slightly steeper slope on the shorter length side. This will result in a bias to deflect the ball away from the shorter side to wards the centre of the beam.
Comments please.
@BANYANAMANYANA No, gravity between the ball and the metal surface is tiny. It would be impossible to notice in this kind of setup.
so what if you had a metal made from a single "perfect" crystal of metal? would that then become a superior bouncy surface than a standard metal that contains imperfections?
My initial thought about the breeze block was that the rough surface would make it increasingly hard to level everything perfectly so that the balls don't wander off. I wonder how difficult that was. My second thought was that the porous breeze block would probably absorb a lot of energy compared to something more dense.
This is insanely fun to watch. Great explanations.
This is cool, that material must be incredibly hard ont he rockwell scale. You can get a similar, if not as good, effect with heat treated metals with very high hardness. A tungsten carbide disk brazed to a larger steel billet would probably be really close if you wanted to homebrew one.
So nice of Tim to lend you their very bouncy metal!
it took a lot of balls to make this video. well done!
Balls that varied in size and stiffness.
OUT.
😄
"No balls were harmed in the making of this video."
Big cube isnt happy about this.
I'm with you! My first thought was of amorphous metal ball bearings. So curious how that might affect the bounce
What happens if you use two surfaces? It would be interesting to see two experiments one where the top surface is stationed above the balls initial drop. Then another experiment with the surface placed into the balls path upon its return upwards.
Would be interesting to derive the cubic/quadratic ratio and test it at different pressure to see which one fit's the theoretical value the best or worst.
love this! one of the best science videos i've seen in a while, fascinating aspect of metallurgy i had no knowledge of
Why is glass transparent? I know I can look it up, but you should do a video on it because I enjoy hearing your explanations
I love the bit where the bounces per second matches up with the frame rate and it looks like it’s floating.
Steve, did you consider that the magnet ball release technique might have influenced your 1 minute plus run? I mean intrinsic temporary magnetism or similar?
I just got this video recommended by the YT. Really good video. Production value, recording quality and the writing and delivery are spot on!
I would have loved to see that vacuum pump used for a demonstration between the extremes of your experiment (smallest and biggest bearing ball (I just learned that, thanks!) of 2 or more materials).
Just because I am intrigued: can someone tell me if you could calculate the "lost" energy from the loudness/pitch of the sound it makes?
Also, to get an amorphous metal ball, maybe challenge some chemistry KZcliprs like Nilered to try and make it! I can imagine it might be fun for them and also a great video for us!
I don't know if its possible but could you possibly obtain a bearing made of the same exact compounds as the plate and see how that affects the bounce
have you tested the bounce time with air AND with the glass cover on top? Maybe this is a factor, why in vacuum it doesnt bounce that much longer (similar to the change to put the device on a brick)
The ruby bearing is especially wear resistant. These are used as the point [of contact] in performance tops that are designed to spin as long as possible.
Your videos are just amazing
Hi
Nice
Nice ☺️
Nice
Good
Hey Steve. Not sure if you'll ever see this, but do the bounces of the ball bearing follow the 80/20 rule over time? As in, do 80% of the bounces occur in the final 20% of the time?
I never knew ball bearings could be so entertaining! Great weird video - thanks.
If it helps, the outer and inner parts of a ball bearing are called races. The reason they call it a ball bearing is to denote the type of bearing inside the races, as there are tapered bearings and rod bearings as examples of alternatives. Awesome video, thank you for sharing.
Hi Steve, Have you made a video about Nitinol? It is called a "memory metal" because, although it can deform plastically, it can also be gotten to return to its original shape. I'd love you to explain how that is possible. Thanks.
Looking at the AISI steel bearing size test, the 5mm one seems to be an anomalous as the bearings on both the larger and smaller side of it do not bounce as well, yet the 7mm one fits well with the 8 and 6 mm and is almost as great as the 5mm. I think this would be obvious if you plotted the size versus time. Watching this reminds me of the superballs from my youth.
Grand Illusions is such a gem, it warms my heart to see you guys team up in any way!
Yes a great illusion. Thumbnail gave the illusion that a ball bearing could bounce on amorphous metal 259 times yet only it was only 45 times. Grand ! Amazing ! Warms my heart too ! Just loved it when I wasted 18 minutes to have the illusion shattered.
Tim is great. I can honestly say I've enjoyed every single show of his. I like how a lot of my subs know each other, and work together. They've even flown from the US to do a show with Steve.
There is an alternative audio software to audacity called reaper which allows you to detect and at the same time count so called "transients" sharp amplitude peaks in audio signals. The function is called "dynamic split items" if you wanna try it out. Reaper is free for a 30 day trial and after that still very affordable at only around 60 bucks or so. So feel free to try it out the next time you need something like that :)
Excellent video. When you pressed your thumb on the artist's eraser around 10 min in it triggered an olfactory memory and I could smell the eraser from my childhood. Fascinating.
I don't know why or how I ended up here, but that was really fascinating! Really well explained and super engaging! Thank you for this video! :D
The tuning of the ball material to floor material reminds me of impedance matching for optimal power transfer.
Cool video, LiquidMetal Technologies made a video 11 years ago with a stainless bearing bouncing on their material for 81 seconds. Its cool that there are more companies researching amorphous solids. It would be awesome to see an amorphous bearing against the amorphous plate.
I love grand illusions! Tim's channel is so positive. It's always pleasant when things I follow collide
@fluffysheap or break apart, RIP glass tube
And then bounce back with a high coefficient of restitution
I didn't expect I would have watched bouncing balls for 18 minutes...and yet I did. Great video.
The sound of the ball is lovely, sounds like its very excited :)
What would a knife made out of amorphous metal do?
Great vid! Watching it made me think of a bounce test, with a steel ball bearing on perfectly flat anvil. The blacksmith in question was extememely impressed on how long the bearing bounced, as well as how much he could conserve in an arm swing with his hammer.
It totally sits in line with the results in your vid. As we know, energy cannot be lost, but transferred into an alternative form - sound, heat, light, friction, deformation, etc.
I would love to see the results with a hardened vehicle on a flattened, hardened surface (due to the atomic gaps being smaller) as well as on a hardened amorphous (if it is possible) material.
Is it physically/atomically possible to revolutionise the ol' hammer? Food for thought!
If you can find a disc of AM-III amorphous glass of a similar size and a ball bearing of AM-III as well, I think you might be able to beat your record.
A long time ago there was a distinction between roller-bearings and ball-bearings. The rolling elements of roller-bearings were tiny cylinders rather than tiny balls. I believe there was another type where each rolling element was machined to be two cones joined base-to-base, pointy-ends outwards. The curved surfaces might not have been the cones you get by taking a gore out of a flat surface, but, rather, curved a bit inwards or outwards. The bigger elements, the circles inside which the bearings roll, could have had quite a complex geometry since ordinarily a cone will not want to roll STRAIGHT on a flat surface, and if forced to do so will have a lot of friction from the forced skidding. I'm not sure how all of that would've been worked out.
Plasticity changes a lot with the change in temperature. It would be great to see how frozen metals compare to room temperature metals in this test. Steve probably has liquid nitrogen next to his bed, so he could make a test and let us know how it worked.
@Lukas H. How about the balls though? Especially the steel ones.
@Jarek Ferenc It's all about hardness, which can change dramatically with temperature. A very cold thing can become very hard and brittle, and could allow for much cleaner bounces because of this. Make it cold enough and the bearing heavy enough, and something will shatter instead of bouncing, and I expect this to be a very elastic collision right before that transition. There are a whole series of experiments about bounce height and material hardness and strength you could do here.
@NO WAY My gut feeling is telling me: a bit. A very tiny bit. Cooling down to -40°C will not increase the yield point significantly. The BMGs exhibit the yield point (if they undergo any plastic strain) at some 1000 MPa, which is incredibly high, so the stress at impact is usually within the elastic range. Thus, lowering the temperature to -40°C will have a negligible effect.
BTW, if iron-based BMGs are considered instead of zirconium-based ones (probably they were used in this film), they would withstand the compressive stress of 4000 MPa or even higher. Making the amorphous plate 2 mm thick is possible even if it is iron-based.
That was the variable I hoped he was talking about! And then he pulled out multiple _sized_ bearings, and I was like, Ok, sure, they all have different stiffnesses and air resistances, so you'll get interesting results, but I wanted to see liquid nitrogen and blow torches!
@Josh Peterson ohh yeah! Like a double bounce on a trampoline. Although at that point wouldn't you be adding energy to the system?
Great analysis, Steve. You seem to have set many minds thinking. Mine for sure, thanks!
Nice video. Missed to mention that it is in fact the hardness property what we are looking for rather than stiffness
Cheers for giving bulk metallic glasses some much needed airtime Steve
Wonder if you could calculate the speed of the shock wave travelling through the base material. If it is sitting on a material significantly stiffer, then the reflected shockwave could form a constructive interference pattern in sync with the ball bearing contacts. Would provide the most benefit at the beginning of the bounce I presume. What about temps?
I would love to see how this could work on the ISS. Yes, I know that the BB would need some acceleration to get started.
to count all but the audio rate bounces, you can use audacity’s sound finder feature (i think that’s what it’s called, and it’s in one of the last three menus). it creates labels which are numbered in a separate label track. you can also slow the audio down and then run the sound finder, to give it an easier time finding the bounces.
I've not used it since 2003, i think it's called 'mute'?
I think it's called Beat Finder - haven't used it for a while, so..?
What an amount of work! So impressive!
Sorry if you cover this later, but while you were talking about the different materials, I wondered about how Amorphous Ball would interact with Amorphous Disc??... Very Interesting stuff btw🤔
As a mechanic without an engineering degree, I'm curious if temperature would have an effect? Would heating or cooling either the materials of the bearing and the impact plate or the air within the glass column make a measurable change? Also as a mechanic who deals with bearings in the field and not a lab, would impurities (finger oils or dust) affect results? I've dealt with bearings that a dirty finger print would affect it's performance, although that is probably more a rotational force.
Running tests with the system at different temperatures would be interesting.
A heated ball would have more energy to start with. But, the electrons in the balls metal could be more free to move around. If it's cooled, the electrons are more tightly held but over all the ball has lower energy. Or, heat transfer with the air would cancel out any temperature change.
A cool line of inquiry no matter how it turns out.
What would the bounce rate be based on temperature? Like, how much longer or shorter would it bounce if it just came out of a bowl of liquid nitrogen?
The sound during the vacuum test is a really good illustation of energy lost through the materials when it bounces
Steve, Little tip of advice from an audio engineer.
If you ever find yourself needing to count the peaks of a audio waveform again. Save your time and eyes, Look up Dynamic Splitting. You essentially set a threshold above the noise floor and it will detect any transients (peaks) above your setting and then splits them all. Resulting in each transient being its own audio "file". Highlight all files and see how many you have selected, or a good Slicer will tell you how many transients are detected before actually splitting for you.
Would love to see a newtons cradle made from this anomorphis metal and see how long it goes🤔
And a fidget spinner of this metal as well😉
What happens if you supercool or heat up the bearings and bounce them? Does this affect it’s consistency?
I was thinking you could do sample tests of air vs. no air and then compare the means statistically and see if there is a difference (t-test).
Maybe there is also a sort of resonat effect between the ball and the ground plate, that can be used to improve the duration (theoretically):
The ball has its own vibration, it wobbles (11:40) and it does it at a very specific frequency (resonance frequency). the same applies to the base plate. If both have the same resonant frequency and both oscillate synchronously after the initial beat, this could possibly have an effect. Or even if the resonant frequency is different, but in such a way that they influence each other positively at the moment of collision, I think the system will oscillate as follows: When the swinging ball hits the also swinging base plate and that happens exactly at the moment when both are compressed (the vertical expansion of both is at the minimum ), or even a little before reaching the minimum (as in the combustion engine: just before the dead center is reached), so that both make contact while each is still in the contraction phase, so the impact energy is still something for both squeezing a little bit more. It's like a swinging spring that gets an extra impulse just before it reverses direction.
Will Stelter showed a method of testing anvils similar to this, the frequency of bounces toward the end is mesmerising. It's as if the bearing is momentarily floating
Watching the metals change by weight making a normal distribution, very cool.
Why are the amorphous metals suddenly harder to find?
From what I gather they were made as an Educational demo. So not available commercially in this format. The company (as a spin off) still manufactures items from this material and other similar ones.
Sounds like it would make great armour plating. A ceramic armour is often mentioned in scifi (eg 40k marine armour)
*Send me a direct message right away
*Tʜᴀɴᴋꜱ Fᴏʀ Wᴀᴄᴛʜɪɴɢ...
Sᴇɴᴅ ᴀ Dɪʀᴇᴄᴛ Tᴇxᴛ Oɴ Wʜᴀᴛꜱᴀᴀᴘ Wɪᴛʜ Tʜᴇ Aʙᴏᴠᴇ Nᴜᴍʙᴇʀ👆
Great video Steve, interesting subject!
However, at 9:41 you mention 'the energy is now lost'.
Is it really though? ;)
Also, I like the mention of Grand Illusions!
Been following them for years, it's always fun to see Tim showcase some quirky toy.
You might increase the stiffness by chilling to a low temperature. To avoid moisture condensation, that would necessitate running the experiment in a dry gas (nitrogen and argon are probably the easiest) or vacuum.
The thing that came to my mind when I was watching this was it similarity to Euler's Disks. I would look into the the technology and study behind those because they are just as dependent on material sciences and conservation of energy and you might find some useful information about this topic.
This feels like an Inception thing; if you have one of these, and the ball bounces for ever, you're in a dream, if it comes to a stop, it's reality.
it would be interesting to retest the smallest ball bearings you had in the vacuum chamber to see if there is a large gain with them as you had already selected for the ones that were least effected by air resistance
seems like it would make for a good revisit as it is just retesting in the vacuum chamber so i have me fingers crossed there will be part 2
ok
Excellent point
One other kind of ball bearing material I'd like to see tried is CBN (Cubic boron nitride), being the 2nd hardest material there is after diamond.. Quartz could be interesting as well.. unfortunately I haven't found any sources for them.
CBN is often used in machining very hard metals, it's pretty amazing stuff
@Rx7man In theory, the stronger the crystal lattice the harder the material, the problem with ceramic materials it's that it's not structure of single crystal defines their properties but also how crystals are packed in the material. I googled for microscopic structures of CBN and ZrO2 and the CBN has much more empty spaces in structure which can lead to friction between crystals when bouncing and loss of energy. If we could create perfect monocrystalline balls of both material then CBN would be a winner.
@Fred MacMurray yeah, quite possible, and if they're really locked in you'd think that would provide a stronger crystal lattice.
Good point about density. Makes me think that rebound has more to do with how strongly the electrons are locked into their orbits.
@Stewart Ross I think the density would be minor in comparison to hardness.. density would mitigate the air friction a little but not much more.. try lead and gold.. heck, try mercury :)
i wonder if density would affect results. it would be interesting to try osmium
When I was watching Tourette's Guy on KZclip 15 years ago, I never imagined I'd eventually be watching some guy on here spend an incredible amount of time and energy and considerable expense into seeing how long he could make various bearing balls bounce. I mean ball bearings bounce.
Does the temperature of the materials impact their restitution coefficient? Would make the experiment more difficult for sure though, making sure to keep other variables the same.
the way to get a better coefficient of restitution is to size and shape the anvil so that it reflects all of the sound energy back to the ball before the ball leaves, rather than having that energy come out as the sound we hear. also remember that "super balls" are hollow, again to ensure that the elastic energy goes back into its bounce rather than to a vibrational mode of the ball itself (but good luck getting a hollow bearing ball !).
Even if the ball did bounce an infinite number of times, the duration of each bounce decreases geometrically so the total time taken is limited, kind of like Zeno's "paradox".
I wanted to be a materials scientist when I was little, until I got a little older and realized that my days would be spent doing things like this all day long. It's much more enjoyable to watch you do it for us, lol.
ok
A lot of things are more fun when someone else is doing it.
I am a material scientist and i did this 1-2x in my 20+y career.= You have a great missconception.
This job is not boring hypothesis and testing is the most fun you can get out of life lol
love when the ball matches the frame rate of the camera and looks like its floating...
Is there a known value of Young’s modulus for this metal?
You mentioned there was a difference between stiffness and elasticity but I thought that they were the same or similar
Steel bearing on granite, marble, ceramic flooring bounces really well.
I'd be curious if this experiment was performed in a vacuum! I suspect much longer bounce times... :)
That amorphous metal might be useful in musical instruments like guitars for lengthening the duration a string vibrates, while maintaining its amplitude for as long as possible. That material could be used for the Bridge Saddle and the Nut. 🤔🤓
A collaboration between Steve Mould and Grand Illusions was not one I was ever expecting.
2 great people
It has happened before!
"But the point is: look at it bounce!" Timeless
So, the amorphous steel just sounds like hardened steel from what you're describing. If you want an amorphous bearing you should take one of those 52100 bearings, heat it to about 1500 degrees freedomheight, hold it there a minute, then quench it in oil. Vegetable oil or peanut oil is just fine.
Your voice is so soothing. Just the tonality and your speech/cadence, it's really calming.
Also, this was so much new information. I've also never seen anything so strange and satisfying as these tiny metal/ceramic balls bouncing way more than they have any business doing. Even the sounds are immensely satisfying, the pops from each bounce.
I wonder if there's some kind of application for this ultra-bouncy metal, in harnessing kinetic energy. "Every bounce charges my phone 0.003%" lol.
And man, I wish KiwiCo had been around when I was a kid. That subscription service sounds exactly like something I'd have loved as a kid, and I'm positive I'd have learned quite a lot from it.
@12.06 could the sound generated echo from the lower contact surface and shoot up and give the force additional energy to keep it running for so long? Which brings to question? do you have any sound density/speed, data for resonance etc... that would be interesting to see.
So cool! I'd love to know what frame rate this was shot at. Looks like 24fps. Maybe 30? Not important at all. Just curious:) great video!
I created a jig to test COR in university as a project. After testing a number of materials, was very surprised(as was the class when I demonstrated) that cast acrylic had an extremely high COR. I don't know why but everyone assumed it would dampen the bounce.
The 5mm balls used for 'bb gun' ammunition are very bouncy on a hard surface (floor tiling is best but also very bouncy on wood laminate flooring). I think they are also some sort of cast plastic. Used the property for a 'door trap' to make sure no-one was sneaking into a room of mine before - count the balls load them into a pot that is released if the door is openend; because they bounce all over the place it's hardly likely they would all be found and replaced so you know if the trap was sprung (count the balls in the trap without setting it off). Unfortunately I forgot I'd set it one time, and ended up with balls bouncing everywhere on entry. 🙄
Yeah, you can get acrylic balls at TAP that bounce like really noisy superballs... they also don't really warm up as they bounce. Unless they crack or chip, they really don't stay deformed. PLA on the other hand... "sploot"
Some days I see your videos and am locked into the concepts and the variables involved, today I'm like "ooh what kind of cats are those?"
Now I'm curious why rubber bouncy balls bounce so much better than something like a rock in regular situations