TL/DR: put some vertical bars inside the tube to prevent balls from rolling back down, archimedes screw only works because its tilted and cant just flow down the screw like the balls. you're relying on the friction between the cylinder and balls, which is fine for grain or cereal that dont flow down the spiral. but with smooth atomic units like these balls a vertical bar or "cleat" in the tube would greatly increase the flowrate by preventing the balls from rolling back down. thats what i thought you were getting to with the vertical bars at 0:17 , but it never came in the whole video. please flip the red bits inside so they fill the gap between the tube ID and screw OD and youre golden.
@@holyravioli5795 Only if the screw is turning in the same direction. The important thing is the relative speed between the screw and the casing. The bigger the difference the faster the outcome
The important point is the minimal damage to the items being transported. If the screw moves there is more damage to the items. This was first proposed by a Mr Olds to minimise damage to grains. Turning the screw as well may increase speed but may increase damage to the items being transferred.
just add a thin vertical extrusion on the internal circumference of the tube and it will push the balls physically, preventing them from rolling back down
My thought exactly. Without extrusions like you propose it's impossible to empty the sump, as the balls only rise up due to back pressure from new balls being pushed into the tube, which only works with very light balls like those ping-pong balls, but not with heavier ones such as golf-balls.
@@cheeto4493 I just said it myself in a reply to someone. I'm amazed how far I needed to scroll to find it, and I'm amazed James didn't do this already. Seems SO obvious, and it was even shown in the CAD model. We're either being played with, or James got tunnel vision on this and couldn't see it.
Great stuff. Was hilarious seeing your test apparatus keep getting more and more involved! :D Zero resale value, though :D Keep up the awesome content. Looking forward to seeing your 'great ball contraption'...
I liked how you narrated and built it as if winging it as you go without a plan, and everything admirably just fitting and ending up working together. Literally build as you go. Well done!
I'm pretty sure my grandma gave me a toy when I was a kid that did something similar to this. I remember being a little fascinated that the screw didn't move.
Now I am wondering if this could be made flexible, similarly to flexible shafts in sleeves, making "active pipelines" to transport small grain materials and fluids. Also, what if the design was reversed, so that inner core was smooth and cylindrical, and the spiral ridge (path) was attached to outer wall? For my former idea, that would mean using almost every part of existing flexible shafts, replacing only the sleeve and adding side ports for transported material at endpoints.
FRC Robotics teams have used this idea but made it more efficient with a thin vertical set of bristles on the inside of the tube. It increases friction in a gentle enough way to force the balls up quicker.
I'd forgotten I'd seen the Tom Scott video but my initial reaction was "well if the screw turns and it works, then yeah, turning the housing instead would also work. Because which one's moving is relative."
I think that if there are no more balls in the reservoir, the "pump" will stop pumping. I also think this does not count as an Archimedes screw. But this is a very nice feature for something like a marble machine that uses ping-pong balls. In an Archimedes screw pump, the medium is pumped up due to the fact that is is trapped in a small pocket where the medium cannot flow back down again because the screw is angled. This system uses a combination of balls being pushed up because of the force of the balls entering the scoops combined with a bit of friction from the tube on the balls. This only works while the balls are light enough and have enough friction. This will not work with tennis balls or glass marbles.
So the thing stopping this design from being extra efficient is that there's no groove on the tube for the balls to slot into. If the balls could only slide vertically upwards along the track, then the screw would provide the upwards force as it spun.
I think that adding vertical strips spaced at the size of the balls, on the inside of the rotating cylinder, would push the balls around, which would make this whole contraption much more effective
Aah thanks for finishing the video with the Tom Scott clip.. I had paused the video to search for it since I remembered seeing some video about it.. But I thought it was Smarter Every Day talking about grain bins.. But nope.
As long as there is enough friction between the outer tube and the material being lifted and the incline isn't so steep that gravity pulls it back, it should work exactly the same
The scoops are the key addition, as they force the material into the system. The friction with the sidewall does also help, but it's mostly about the scoops.
As an individual piece you could just attach a slide to the top catcher to feed the bottom but if it is going to be part of a larger system than that works to. Can't wait to see what you come up with.
I'm not a rocket scientist, but here's my theory: The balls are spun and gain energy. They are moved IN and so as with the conservation of angular (something) they will move faster. Repeat as the next one comes in behind. Also, friction from the outside all and the ball will make the ball move "forward" and thus up. Oh, and I love your toys behind you.
The system works but is greatly limited by the friction between the tube and the ball. You only have a point contact there, versus if the screw was turning you would have a lot more surface contact between the screw and ball. It works better (but still not great) for stuff like seeds because they send to not have a spherical shape and thus for their size they have a decent amount of surface contact. Plus the coefficient of friction is higher between seeds and the tube versus plastic on plastic.
13:30 It works because they spin isn't explaining anything. The reason they spin (relative to the screw) is because of mostly 3 forces: 1. The opening at the bottom which are inclined and especially with the pressure if all the balls in the basket, actually strongly push all the other balls up. If instead you had a flat opening, the whole thing wouldn't work. 2. Friction of balls against acrylic cylinder, helping a lot to makes them spin 3. Pressure equilibrium: It'd work with water as well and there it's easier to understand that you'd get the balls in the basket and in the tube to be at the same level (supposing there are no frictions or you just make the whole thing shape to achieve the same) Note that the Archimedes' screw that you referenced in the beginning only works because it's inclined. This mean that due to gravity even stopping it would keep most of the water in the tube. It'd only work vertically with a LOT more speed. This is important because turning an inclined tube isn't the same here. In you turn the tube of an Archimedes screw, without turning the gravity vector (Earth), then you only have the friction of the cylinder. Supposing a perfect screw and cylinder, it would actually not work anymore.
If it was working with the screw rotating , it'd be rotating with respect to the floor the friction if which would presumably help push them, assuming there's a pool , however it isn't here . I think?
So given that the Archimedes screw works and this system works, would the next step then be using a planetary gear setup such that the sun gear is attached to the screw, the planetary gears attached to a stationary motor, and ring gear attached to the outer tube, which would have a opposing rotation for the screw and tube, thereby increasing the movement of fluid / particles?
Ok what I'm seeing is that the base is what is actually driving the balls with mechanical pressure, not the interaction between the screw and the tube. I'd like to see it with a tighter tube providing friction between the balls and the center screw which should then use (in theory), the friction from the outer tube to drive the ball up the screw. Basically, you'd know it works if it can elevate a single ball to the top. Your model (and the commercial one) require a critical mass of material to be present in order to get it to work.
You could do this with marbles, and have the return spiral around back to the hopper. Marbles would rotate one going up, and spiral around the opposite direction going back down. Better than a lava lamp, if a bit noisy. Lol
That moment when you think up a crazy experiment and find out it works, only to then learn afterward that it was already invented and patented. Good job though, I figured it would work because I used to sand wood all the time. Let me explain. When it comes to sanding wood, you can either lay the sandpaper down and move the wood against it or hold the sandpaper and rub it against the stationary wood. Neither process is better or worse, though holding the sandpaper might help smooth out indentations and laying the sandpaper flat might help to sand larger, flat areas at once. The friction, in the end, is the same. Figured it would be the same whether the tube or the screw was spinning. Seems I was right, but for the wrong reasons.
Not sure why some would think this wouldn't work. I guess it's just our perspective that is skewed, and therefore not seeing that everything is still moving the same in relation to gravitational forces.
I think one problem with this idea is that you already made a mistake at the start when you assumed the balls would rise [better] when you turn the screw itself. Because that's what you take as a baseline and even that wouldn't work properly if the screw would go straight up. Only if you angle the whole construction and gravity keeps the balls on one side, they raise when you slowly turn the screw. If the whole construction goes straight up, the balls won't rise but just move around in a circle at the bottom. The only reason this still works somehow is friction. Friction is basically replacing gravity here, but I'm pretty sure gravity would do a much better job, even though gravity will only work as long as you turn the screw instead of the tube around it and you angle the whole construction. But as long as you depend on friction instead of gravity, it doesn't really matter if you turn the screw or the tube.
I'm thinking that this would be a good idea if the middle screw was printed with some compressable material on the surface, so the balls travel up the screw sort of like ball bearings in a bearing.
It seems like the pingpong ball needs to make a 180 degree turn right when it gets inside the tube ... the screw going clockwise would have been smoother. Other than that I just marvel at the stuff people can actually build.
I'd love to see a large scale demonstration of a ball screw like this as understanding how they translate motion can be a bit difficult to get your head around.
Helps best if the balls are colored on the sides so the motions can be represented. Solid white ping pong balls are horrible to tell which way they are spinning. Even drawing lines on the hemispheres and painting the 8 different sections could help immensely, and make it more vibrant.
if you had added traction to the tube i.e. several thin vertical strips of rubber or such (on the inside), it would make if far more effective (I think!)
Paused at 5:53 once I had a better idea of how he was going to build it. For balls, yes, I think this will work, just not as efficiently as if you spun the screw. The reason is because the balls are requiring the horizontal friction from the movement of the clear plastic piece to move them up the ramp whereas spinning the screw utilizes the vertical friction of the screw itself... One works with gravity, the other does not. In Zero-G, I think spinning the outside would work just as well as spinning the screw... Let's continue watching and see if I'm right...
unlike the archimedes screws, it's usefulness is completely dependent on a constant supply of solid, non-liquid, material....if you had stopped restocking the hopper with balls, it would have ceased elevating anything...you can see a precursor to this between each ball pickup when flow pauses and even reverses...which means it could never clear itself of material through the top unless a material was supplied through the bottom
That's a pretty sweet setup! Maybe put a mark on the tube (like a sticker) so it's easier to see that it's spinning. And why not have the balls drop down in the base again, to keep it going indefinitely? Would work great to just keep it running in a science fair environment!
My guess, is that it would still work. Like the difference between a propeller and an impeller. Spinning tubes are constant acceleration, which is just gravity, so at a high enough velocities: things stick to their walls. Then from the perspective of the balls on the wall: the corkscrew in the sky, and the entire universe under their feet: would both be spinning around a stationary corkscrew. So the depending on the reference frame: the screw would still be able to be seen as spinning. I wonder what would happen if they both spun in opposite directions snd were in contact with a fluid: specifically I'm wondering how the speed of rotation would correlate to the amplitude of the fluid incursion 🤔 Because of viscosity and stuff, I imagine right at the boundary: you'd get two opposite whirlpools (for the two rotating opposite direction model) and they'd just peeeeek into the tub before spiing out down the middle, but the whirlpools would probably be eggshaped rather than circular: and if things spun fast enough they could probably be stretched out by maybe almost a single screwthread. So if you then a tilted your archimedes screw relative to the gravity plane of the earth, by about 45° you could probably get some weird fluid rythms going, as slosh resulted in slight overflow, which would accumulate and slosh and overflow. The video of it that I just now imagined in my head: would look similar to counting. You know how when you count: each digit goes a bit slower? It would be like that, but analog. Each thread would be like a digit. Sidenote: nature is fundementally quantum, quantum means discrete. So analog computing isn't truly continuous, but it is fairly close to being continuous, because the molecules in the water are a lot, so it is like, counting with very big numbers. I think that might be fun actually hmm 🤔 pls try it, and pls give me an itty bitty credit pls, if this post inspired you, thank you 🙂 PS: then you could compute by putting the "double winder" [as we'll call it for brevity] on a scale, and subtracting the container weight from the scale, and using that for computation. And if you used analog dials instead of discrete read-out, and you fed those you potentiometers, and sent those into other motors and gizmos, then you'd have an analog computer, but it's primary design flaw would be that it would have to be in a gravitational field, so to use it in space, you'd have to be in one of those space Odyssey 2000 people tubes, the big ones that spin up artificial gravity. To use it while moving it around: you'd probably have to switch to a ferrofluid, and then you'd have new design limitations and would have to make things chonkier so that you could account for and accomodate the magnetic spikes that ferrofluid likes to make. You'd also need gravitational sensors, and accelerometers to actively offset the effects of wiggling and jostling and changing gravitational persuasions, so to use that kind of fluid computer, you'd also have to have electronic scale controllers, and fitting w/ the spiritx they'd have to be analog too, but preferably a discrete type of analog which I'm not sure exists, but that with all that fanciness the test would be to run it upsidedown on the ceiling, and then run upsidedown on the cieling again, but halfway through the count, spin it right side up, and then compare the numbers and calibrate accordingly, and then ship it Ganymede or Io, and calibrate it again without interfering with the earth calibration, and also try odd angles and different count interference times. Then the final test wpuld be to put it on one of those work-put jigglers, and just oscilate it to death, and do your best to give it a good tolerance 😉🤣👍 (Maybe don't bother with the upgraded version, that sounds like a lotta work for nuthin in my oppinion, I just like to design things in my head) PPS: the ultimate analog readout would be a circular arc of variable length. Preferably from zero to 45 deg.
Well THIS guy's no Mark Rober. It's just the friction from the glass providing the force to move items up the screw, Big whoopie-doopie, my mind is not blown.
What if you made a hopper that fed the balls back into the sump? I guess a cardboard box is cheaper for the testing/ prototyping phase, but then you test for flaws indefinitely.
The fundamental flaw of this design is that there is no medium for transmitting the energy to the ball. In case of the spinning screw, it's the grove wall that transmits the energy: while the ball is free to slide along the grove, it cannot slide across. In case of spinning tube there is no such restriction - the ball can slide any direction - so the energy is not transmitted hence the ball not rising. The only reason why it (kinda) works in this video is that other balls play role of the restrictor, thus transmitting _some_ energy but not nearly enough....
If you put spines down the Inside of the tube that can touch the balls but not the screw then it would work much better. But I guess it wouldn't be the same thing anymore. Or just use something with a rougher surface. Friction helps.
With the feeding mechanism you basically pushing them up the screw. The real analogy would be to only turn the zylinder, what probably wouldn’t work very well.
You need to put a couple of lego wheels with motors so the ball are thrown away and not fall in the box. I see it's pseudo random the timing of the balls. Would be an ping-pong throwing machine to practice.
If you take the top chute off, do the balls just fall into the bowl at the bottom and start going up again? Or do they just bounce out on to the floor?
