The Band Should Slip Off But It Does The Opposite! 

SNATCH BLOCK!.... er... uh, I mean. . NOT A SNATCH BLOCK

That explains it.. It don't remember what for, but a few years ago, I designed the concave thingy for a homemade belt system, and I couldn't understand why the belt was, always flying off. I had to put on guards to prevent it... It was a mess. Now I know. Thanks! :)

That concave pulley clip really answered all lingering questions in one shot for me! Great work as always Steve!

2:12 "Our intuition is wrong." Learning physics summed up in one sentence.

You've just turned everything around that i knew about conveyors. I work in the metal recycling industry and usually we have to track conveyors to keep them in line. Usually we tighten the side to make the conveyor move opposite of the side we tighten. Then we'd get some conveyors that just didn't seem to wanna do that. They'd move opposite the way they should and everyone was always confused about it and just chocked it up to basically being a odd quirky attribute of that particular conveyor.

The instant a video starts talking about pulleys I'm already waiting for "SNATCH BLOCK"

You can never miss "Destin", when you say snatch block

I love how all the science communicators on youtube reference each other in their videos

I've worked in manufacturing for years and have observed this effect on lots of conveyors, belt drives, belt sanders, etc. And this is the first real explanation I've heard. Thanks for clearing up something I've been mystified about!

When I did my degree (mechanical engineering, a long time ago) we spent half a lecture on this. The lecturer (a tenured professor) actually explained it in terms of velocity and angular acceleration of elements of the belt. I was more confused after than before. This deflection-based explanation is way better, and makes me wonder if the comparative angular acceleration explanation was just totally wrong. Also had no idea elastomer elasticity was from entropy... awesome

Anthropomorphising a rubber band is streching it a bit! ;)

Hey thanks for this! I work on a plant where we have long conveyor belts. We crown the centre of our head pullies to help centre the belts and prevent them from running skew. Been using this method for years but now i understand why it works!!💪

There is a pulley on my car that always bothered me that it's crowned, always felt like it was gonna be a problem one day. Guess I was wrong lol

Hey Steve! I've followed you for years, so it was super exciting to see our belt grinder in the video Tom sent you. I definitely didn't expect that when I started watching. Obviously, belt tracking is a phenomenon we spend a lot of time thinking about in belt sander design, so it was nice seeing someone explain the crowning effect scientifically. Home builders often run into problems when they add a tracking adjustment because the interaction with the crown isn't necessarily intuitive. The belt seems to ride to the high point of a crown, but it also looks like it rides down the slope when you tilt a wheel to adjust the tracking (as it does in the video Tom sent you). We have our own working theories on how crowning interacts with adjustable belt tracking, but I'm sure I would learn something and see it in a new light if you ever looked into it. Eric

Amazing!!! I have been working on a homemade conveyor and tracking was proving a problem. Based off this video, I wrapped layers of tape around one of the pullies (to make it convex) and it immediately started tracking perfectly! (And when I manually disrupt it, it returns to center.) Thank you!

1:37 I was about to make a cheeky comment like, "What if we won't permit you?" I felt clever until you showed you're more clever...

"Heat is just molecular jiggle" It's kind of unsettling how simple and accurate this is.

Hey now, this type of pulley system looks familiar ; P It's exactly what stabilizes the band on a Van De Graaff Generator.

That was strange for me. When I first saw it, my intuition told me the band should ride up on the pulley, but at the same time that _felt_ unintuitive. Excellent explanation. Also worth noting: Take two tires and connect them with an axle, but leave them free spinning. Now, accelerate the system. Next, slow down one tire but not the other. The system will turn in the direction of the breaking tire, because there is more friction on that side. Now, you have a band across a curved pulley. When you stretch the band, the side that is further from the opposite end is pulled tighter and thus has more friction. This will turn the band in the direction of the tighter side. In short, there are probably two effects causing this behavior. The stretch widening the band on one side is one effect, but the asymmetrical friction is probably also causing the band to turn toward the higher portion of the pulley as well.

Something else interesting (and entropy related) is when you stretch an elastic band, it gets hot. If you let it cool down and then retract it, the same area (middle mostly) that heated up will actually turn cool. Yes, it works like a refrigerant.

I explore ghost towns and mining camps, and have seen a lot of old mines and mills which used belt-driven machinery. I have often wondered why the pulleys were slightly crowned. Now I finally know! Thanks!

I was expecting Destin to jump and say "snatchblock" and he is there...less than a minute into the video... SNATCHBLOCK!

I saw the title and the thumbnail and was like: "Duh! That's how Bandsaw blades are kept in place" (learned that from Matthias Wandel) Woodworkers ASSEMBLE! :D

"Just for completeness" is so satisfying. Thank you

Simple yet very effective principle. I first experienced this when I made my own lathe. I needed a set of transmission pulleys and I decided to make them myself. I first tried flat ones. After few trials, I realized the belt climbs(!) the pulley and thus convex-shaped ones self-align the belt. Without any flange, it was very easy to turn with my lathe in a short time with minimal material loss. Still it worked really good. The belt slipped off only when the chuck stall accidentally with too much cutting force.

I really like that every time you anthropomorfize anything ( a virus, an electron...) you say it at loud :)

“Snatch Block!!!”

Steve Mould: "Heat is just molecular jiggle." 2021 Definitely gonna remember that one😁

The explanation and model with the beaded strings regarding entropy was pretty cool! Haven't seen it explained so easily before.

I fix old audio equipment as a hobby. Old cassette decks, turntables, open reel tape decks, etc. Crowned pulleys are common, and I’ve always known THAT they work, but now I understand why. Thank you!

I suspect the spring demo would work better with a triangular lattice ... because hexagons are bestagons.

Just a recommendation, in the discretization you should use triangles instead of squares. This avoids the rotation of the individual elements at the nodes and produces a more accurate deformation of the whole system

Always get a good laugh out of Destin's snatch block clip

Yup - always travels the the highest point... I learnt about this working for my Father - he engineered many large industrial band saw machines for the Aluminum industry where the blade was riding a crowned wheel (with no lip to hold the blade on). It reduces the overall stress on the blade = more cuts per blade / better productivity etc.....

0:12: No, when I think of a pulley, I think of any kind of wheel that redirects the force that is applied to a string. This apparatus has four pulleys, two fixed and two loose, and it is called a tackle, which in this case reduces the necessary force to 1/4 of the weight of the load.

5:14 are those the same metal beads you were using when you discovered the groundbreaking mould effect?! The greatest effect known to science.

Steve, you are by far one of the best educators on youtube. I love a lot of the popular ones, Tom Scott, Kyle Hill, Destin and Derek Muller, etc. But every video you make I find myself going "Ohhh! Of course!" at some point in time. And your curiosity and passion for things that we otherwise just kinda take for granted is infectious. Thanks for what you do.

"SNATCHBLOCK!" Hahahaha, I died with those clips. loved that episode.

Mathias is nuts! I once saw he building a tool that was precise in the nanometers. You'll never need to be more precise than a milimeter in woodworking. He is just that perfectionist

The jiggling chains was such an intuitive visual model, amazing video yet again

"And just for completion..." Oooh yeah. I like the way you try not to leave loose ends. Your videos are well-packaged.

Rubber band chain analogy is the best part of this video. 5:35 pretty much explains everything!

I hear that Continuum of Springiness are releasing a new album this year.

It was so polite of you to allow me to reject your anthropomorphism of the rubber band! Much appreciation!

This video helped me. The blade on my cheap, poorly made bandsaw kept slipping off due to some combination of misalignments that I couldn't quite sort out. I noticed the wheel holding the blade was flat and I remembered this video and so I used a knife and some sandpaper to reshape it to a crowned profile shape and voilà it fixed the issue!

With your explanation of rubber bands + the video of Richard Feynman explaining them, i think i finally get it

This seems entirely intuitive to me because I worked with belts and pulleys on newspaper presses and delivery systems for a long time. It was a common practice to move the belt position a bit by adding a small piece of tape to one side of the pulley. Or shift the the entire web of paper passing between press units by putting a small piece of tape on a roller that is acting as a pulley guiding the web. I always imagined it had to do with changing the diameter of the pulley on one side, making the belt 'climb' to the high side but I never thought a lot about exactly why it did this. I imagined it pulling harder on one side of the belt. It sounds like I was making it into a sort of crowned pulley, or altering the crown of a that type of pulley.

This entire channel is basically the "why" game kids play, I love it

The differential tension is not the reason the belt climbs. The belt in contact with the faster-moving surface is pulled forward faster (& the other side is pulled slower). This causes the belt to angle itself on the crowned pulley. You can demonstrate this by wrapping a paper strip around a cone, initially starting with it perpendicular to the axis. The strip will spiral toward the fatter part. Key to the centering is the friction between the belt & the pulley. This is put to use as a safety mechanism (especially in the old factories with central engine driving overhead shafts, belted to the individual machines). If the driven torque gets too high, the belt will slip, and the centering function lost. The energy of tension will drive the belt off the pulley, disengaging the load (such as when a lathe tool digs into the work). The belts act like individual "circuit breakers" for the machines.

I LOVE LOVE LOVE when you guys feature each other's work.

I used to tell my students that you can actually feel the entropy change in a stretched rubber band as a temperature change, by using your lip as a thermometer as that part of your body is much more sensitive to changes in temperature. So fun to see a room of students stretching and relaxing rubber bands and claiming "I can feel it warmer/cooler!!"

SNATCH BLOCK! Also, Bandsaws use crowned pulleys on the wheels.

"When you extend it, it wants to be short again." "You are anthropomorphizing an inanimate object." "Ok, I'll put it in terms you'll understand: A shorter rubber band is energetically favorable."

Its a physical visible description of the bernouli principal (in reverse) the slower part of the elastic wants to move faster, while the faster part wants to be the same speed as the slower, so it all pushes towards the slowest lowest energy part.

Mathias is amazing, i feel really Glad to see you recognizing his work.

Matthias Wandel's video is so old it's Mouldy now! :D But I do remember his explanation back then and it was fantastic! He basically tele-fixed my bandsaw! He is the most underrated youtuber that has ever existed, more people should be aware of his works

"Heat is just molecular jiggle"

I hate when reviewers, science channels, etc say they avoid covering stuff other people have already covered. I watch that channel because I like the way they explain things or their presentation style or because I wouldn’t think to learn about that topic otherwise, etc. like I never would have thought to look up an 11 year old video about crowned pulleys but it’s an interesting thing. I wish youtubers would just talk about stuff they find interesting rather than worrying so much about what other youtubers are doing.

Something you glossed over in this explanation is that this centering works even on a flat belt with very little tension. Hard to put this into words, but I would say that the end effect of angling the the belt off from its 'length' alignment causes the belt to climb toward the crown of the pulley. For many flat belt machinery tasks the belt is quite laterally stiff (does not flex along its width dimension) and can be under very little tension. Often times a tractor does not need to line up its drive pulley perfectly parallel to the driven pulley to keep the belt in place well enough. Twisting the belt is done to reverse the pulley rotation as needed. The broad surface of the belt provides enough friction to turn the pulleys even with low tension. One rule of thumb I have heard is 1 hp per inch (of belt width). This power ratio works even with a slack belt, but it increases with tension.

I don't know if my explanation is correct but I think the reason the band doesn't fall of is the following: the part of the rubber-band that is closer to the middle is stretched more, therefore it has to cover more distance compared to the part of the band further away from the middle. This creates and internal stress in the band that pulls it towards the middle.

As a woodworker, I learned about this when I bought my first bandsaw. The wheels are "crowned" and we adjust the angle of one of the wheels to center the part of the blade we want to ride on the crown.

Steve: Snatchblock My brain:SNATCHBLOCK!! A few Milliseconds later... Dustin: SNATCHBLOCK!!

I pictured Destin getting a good laugh out of this video. You’re both so great at presenting a topic/concept.

These are often used in vacuums on the brush and are a great example!

So technically gifted with words, thanks for giving my brain a break when you say "not doing the thing" at 4:00 😆

5:18 This entropy has a name, it is called "mixing entropy". When you stretch the rubber, it unmixes (due to becoming ordered in the direction of stretching) and the heat will then remix it back.

4:57 That explanation kinda of fits most of what happens in our universe, if you think about it.

This was absolute class. You can really tell that Steve has *way* deeper knowledge of all of this than what he shows in the video and yet makes it easy to understand for everyone. I'm impressed. Chapeau

Many years ago Mr. Wizard did something similar. If you connect two identical balloons with a valve, the air flows from the less inflated to the more inflated balloon. The rubber mass is more dense in the less inflated balloon, and the contraction force is greater

Funny, I've thought about this before in the context of the leather belts that were used up until the early 20th century in industrial or farm machinery, and never created a satisfying answer for myself. Until I looked at the thumbnail for this video and saw that the belt was a rubber band. I just never thought of the leather belt as being elastic. If you think about the friction increasing as the band moves toward the center it makes perfect sense.

I was always wondering, when I saw old pictures about machines with transmission belts, why they wouldn't slip off easily. Because I couldn't see a rim on the wheels.

Always good to see Destin. Love his channel too.

"Atomic Jiggle" is my new favourite word for thermal energy

Nice explanation of the curve of the belt, it's very similar to slip angles for tires in car steering. Another application are in long continuous metal coiling lines uses crowned rolls and additionally rolls that change angle to induce the "up hill" climb for steering.

There is an easy way to see most of what is going on, something that is very familiar to most mechanically inclined people. For the convex pully, do your calculus. View it not as a single continuous pulley, but a series of sliced pulleys of varying diameter. It's actually 10, 20, or 100 pulleys of different diameters. And the wide rubber band is really 10 thin rubber bands that are linked sideways similar to the springs. Look at that stack of separate pulleys of different diameters from the side and it becomes much easier to see how if there's a small sideways force you could easily end up climbing onto that next pulley. Most of the force to pull you onto that next higher pulley actually comes from the ROTATIONAL force not the sideways force. It's the same as when you have a very tight belt or tire to put onto a pulley or wheel. You shove it on slightly down lower on the pulley until the edge grips, and then you rotate it and the rotation gives you the mechanical advantage and stretches and pulls the belt or tire onto the pulley or rim. This is something most people will be intuitively familiar with doing if they have worked on cars or tires or similar items. Because the surface is continuous, even a tiny sideways force can cause it to move toward the larger side, because the rotational force then grabs it and pulls it out that way, stretching it onto the larger diameter and pulling it off of the smaller diameter.

I always knew these were self-centering (since seeing them on my hoover), now I know why!

We have been crowning wheels for bandsaw blades for years, amazing how engineering works

As someone who has built conveyors for years I will say that crown pulleys are very important in keeping the belts running true.

3:12 _"When you turn the pulley, it contacts a piece of band that is higher than the band already on the pulley."_ Yeah, I don't get it. That describes the obvious part, the "what is happening", but not the "why". (Also, what does "higher" even mean in this context?) If I say "when you turn on the switch, the lamp emits light", have I explained electricity? No, I have explained only what a switch & a lamp do, but not why they do it.

I respect you so much for always trying to give credit to the other people who have also made videos about whatever the topic is. It's something not many people do on RU-vid.

5:00 I heard "elastic bands are made of rubber witches" and was like "what?"

Cheers for the blinkist recommendation Steve. I sometimes skip over sponsored content but I found blinkist really useful over the past few months to digest content :)

I would have imagined that the more stretched out part experiences a different amount of friction due to more contact area and higher pressure, resulting in a net force, or something like that.

Look up "Bow Rolls" used in paper making. Parallel / cylindrical roller bodies, with extended journals each end. The roller's supported on a self aligning bearing each end, and hydraulic (or other ) actuators impose force upon the journal extensions. You can bow the roller so it's maximum deflection is central or offset to either side. it takes the wrinkles out of the paper web and tracks it, too...

When I was a mechanical apprentice they made us do the maths for belt drives. I can't remember it now but you had to know the tenstion on the tight and the loose side . They told us that belts move up on pulleys. It there are several diameters on the same axle, each diameter curves up to the step / shoulder to the next pulley. They made us calculate for V belts as well. Never used it. Thanks for explaining it.

Why couldn’t i stop watching this 🤔

For the convex barrel it is understandable. For the concave can you try a wider belt to see if it will center due to each side pulling away from end to end. It will be interesting to see the resulting shape.

0:21 lol I remember that episode of smarter everyday

Adding a bit to the model. Ruber is a thermoset, meaning the spaghetti is held together at specific points in the chain. Those anchor points also help with the "memory" of the elastomeric molecular chain in the very specific case of a rubber band. Great explanation of heat and movement!

Thank you for requesting permission to anthropomorphize the rubber band (and offer an alternative) because it’s literally my number one pet peeve in science education!

2:24 even if you don't add nothing new, what you actually add is exposure. At the very least if you don't want to republish existing content maybe a way to share it would be great!

New nerd-meme: "SNATCH-BLOCK!"

heat=jiggle, i will remember this forever now

I love crowned pulleyes. I am on a high school robotics team and we use them so often to drive our belts. Though I prefer the explanation in terms of the difference of linear velocity with the same angular velocity as it is still consistent with non-elastic belts

One minute in and this actually makes perfect sense. The part of the rubber band towards the center of the orange is greater then the outer edges. Causing the rubber band to be pulled by the "Grater Force". For example I was driving down the freeway after a snow storm. the freeway was just wet. I came to a particular part of the freeway where some slush had accumulated on the side of the lane. As my tire hit that slush the friction on that side was greater pulling my car off the road. Same with this rubber band because the one side is more stretched the friction of that side is greater causing it to be pulled to the center until all sides are equal in friction.

Next merch item is a shirt that just says "wants" in quotes. It's basically his catchphrase.

Entropy in layman terms: That what you experience when you open a drawer full of sorted wires after a month and find them entangled

So, do we assume that if you use two spheres in the pulley system it still goes to the centre? What about if you use a sphere and the concave spool?

I've seen this type of rubber belt/crowned wheel arrangements on cylinder and platten presses and never thought about how that works. I saw the thumbnail of this video and immediately realized that the belt finds the high spot on the wheel because there is more tension on the belt the higher the diameter, giving the belt more traction and tension to go in that direction.

Huge fan of Matthias's builds, especially his Pantograph routers