I've been working on and off on a bolt sorting system for the last 2-3 years. One of the methods I used for aligning the bolts into the holes was a Subwoofer set to between 10 to 30 Hz.
Long slots instead of circular holes. That way, the shank of the screw can rotate into the correct orientation naturally under gravity. Cool build, also.
I thought that too, had some hesitations but I think I figured them out. Slots would be good. The holes will refuse letting a larger screw through. They get clogged up which is why it’s nice to have many. If a longer slot gets a larger screw captive, it’s forced straight down which is optimal for packing. It’d still be much more “hole” per area than many trying to pack circular holes, so much more screw capability. They’d need a much thicker base though to ensure screws going down passed the filter honestly, not by having the head at a wack angle. But slots would definitely make it much easier for screws to find a hole and then orient themselves. That’d take out a lot of the force requirement to get them to happen to be straight up, might even barely work with current motor.
If you sort by head size, have the screws slide down a slit (vibration motor assisted) which gets progressivly wider (or stepped). This would prevent the system from getting plugged. A coin sorter might be a good inspiration.
I would use rotation and make a tumbling action. Imagine 4 cement mixers or rock tumblers stacked up. Each tumbler is a screen. Don't forget to add a mixing/tumbling blade. Every time I saw your prints I had a chuckle... even the color was bad lol. I make a lot of jank so I appreciated your jank 😊 great channel
I suspect that part of the problem isn't only movement range or speed, but actually acceleration of the moving axis. Since you're using a circular motion with an offset dowel to provide this cam-lever movement, you end up with a more or less sinusoidal function of movement vs time in the axis in which the drawers are moving. Sine waves aren't very "jerky", the acceleration is rather small at any given time. Maybe changing this system for something that provides a more jerky movement function, such as something closer to a triangle wave (i.e. with very strong changes in the derivative of the position vs time), could provide more of a push to the screws! Maybe mounting the drawers with some spring force that drives them to a position A, and a cam that upon falling smashes the axis and makes them move quickly to a position B?
As a fellow maker of custom screw sorting machines I can say that: 1.) A series of fixed width sieves is a practical way of sorting by diameter without getting too crazy on assembly complexity. 2.) Slots work better than holes but the longer the slot the easier it is to flex and not have a fixed width. Beam deflections scales with L^3 so don't get too long. Stiffness of the beam scales with W^3 as well though so increasing the gap (W) between adjacent slots can help a lot as well. 3.) Massage vibrators are cheap and you can get them at walmart although might not have enough power for the combined mass of the hardware you're trying to get shaking. 4.) If you're going to be at ERRF/3d Printopia, I have an optical + electronic based method to demo this year that accounts for sorting by length and head type.
We gave our students the similar problemen. What we saw worked pretty good was using vibration (as otters said) . The teams created that with a motor with an excentric ring, running somewhere around 1000 rpm. This Will create a lot of sound though. With plastic the noise should be less.
I think that a large part of what makes the smaller version work well is the motion on the Z axis, causing the bolts inside to be flung up and down. Another change I'd make it to fit more holes in each layer, by placing them in a hexagonal grid instead of a rectangular one. Although I think the idea to change them into long slots that @airman2468 mentioned is better Edit: another idea that came to mind for how to create the Z motion: solenoids! Place 3+ of them below the device, and let them shake the whole thing violently
A similar principle to a semi manual pharmaceutical capsule filler jig might work. It has a self orientation sliding layer to drop the capsules into the holes below in the correct orientation.
The shaking doesn't get the bolts to separate. You need vibration to get them to jump up. Plus you overloaded it. The feed rate needs to allow for initial sorting, otherwise it will just bridge.
Your inspiration was a vibratory sieve. They also have ones that drain out automatically too, but to the point. Think you need to vibrate it up and down and maybe add an opening funnel to your slots. Great job/project. Eagerly anticipating the next rendition.
Need a vibrator motion to get the bolts through the holes. Maybe a tumbler with different sieve sizes. Change the sieve in larger increments to achieve the sorting
The bolt bucket is an altar to elder gods. When you find a screw, sacrifice it to the bucket. Then, in your hour of need, the bucket gods will provide. All hail the bucket gods.
Very cool build! Could the whole thing be set onto a vibrating table? Like the ones used for casting. Attaching it somehow to something like a vibratory parts tumbler could work. I would think anything that vibrates quickly and violently enough would work.
You could use a weight on a motor to create vibrations, like it's done in game controllers. Fix the motor to the uppermost tray and fix the lowermost tray on a rubbery mount.
I've seen lots of different types of sorting, and all of the industrial ones use high frequency low amplitude. Low frequency requires high amplitude, and no matter what you do, they just shake themselves apart. A 3d printed unit is likely to last even shorter. So, my suggestion is much higher frequency, low amplitude, with some sort of forms/shapes in the bins that will help to turn the screws the right way. But that needs lots of experiments...
Keep the sizing boxes fixed to each other then place the assembly on a separate vibration/shaker unit. The vibration/shaker unit could have a fixing plate which the sorting assembly would attach and detach to. Your sorting can remain modular, but the rigidity of the assembled modules ensures the vibrations and movement are transmitted from top to bottom. If you can modify the vibrations, you can adjust to suit M2 and M3 sorting, vs M8 or larger sizes. Either that or lift the bolts so with potential energy they can fall down through the sorter, (sorter plates in a series of steps, with the treads angled downwards), and the unsorted bolts be collected and lifted back to the top. May still need some vibration/movement on the steps to sort.
i honestly think a handheld shaker at a large size (think the micronics solution for cleaning parts) is still a better option than the motorized version that, seemingly, has allot of friction in the bearings aswell since the plastic will form around the balls due to the pressure, is will create a much larger contact patch, causing allot of friction. this is also why the baskets rotate side to side when moving
I think you need more a shaking up and down motion so you pop screws out of holes that the thread fits but head doesnt and so you realign so they can rotate up and down. I would do a circle that grows like a spiral then has a sharp drop and it gets pucked up by springs.
I feel like it's salvageable if you can add some vertical into your agitation. Where can you introduce a cam that could gently lift then drop the trays? Maybe between the trays themselves so the horizontal action forces them up a little then drops them? Not sure if that little motor would cope with all of that though
Maybe instead of a back and forth motion, how about a rocking motion? Put a pivot pin in the middle of each drawer and push the forward (or back) up and down. This would require less energy, or force, from the motor to get the screws to move. Replace the holes with slots would also help encourage the screw to rotate easier.
Similar to the rotating drum idea but trying to keep as low profile as possible you could consider moving to as close as a ‘folding’ or ‘tipping’ motion as practicable, where rather than a single screen have two screens of the same sift category ‘tipping’ its load into the other. I’d imagine the degree you would need to tip would come down to how much you’ve got in there. Unfortunately the more bolts there are the more you have to try and displace each to get them to fall through.
Looks like your filament might have a bit of moisture in it (the little voids within a single layer line) All you really need to dry filament is a heated bed and a box the filament fits in. A small fan in the box will speed up the processes but its not required. 50-55c bed temperature is what I use for pla. 65-70c is what I use for petg.
To me it looks as if there was a far too high layer height set, maybe to speed things up. Better make them in good quality, it may be a source of problems afterwards if low grade prints are being used.
@@MAButh You can print up to 0.3mm layers on a 0.4mm nozzle with out too many issues. Thick layers do tend to highlight imperfections however. That said, drying the filament won't magically make the part look better but it will fix the voids they have in the layers.
Maybe try using a different shaking mechanism. For example, shake a rigid stack, just like in the manual version, but using an up and down motion with a powerful motor. This motion should be large and with a frequency similar to what you would pull of by hand (1-2hz) with an amplitude of a couple centimeters (yes it's a massive amplitude, but similar to what you're doing by hand). This should shake the bolts through the seive. However the seive would need to be tough enough to sustain the dynamic load and not have the stack break apart during shaking. Tldr: make a mechanical movement similar to what you're doing wuth your hands. (Maybe have a lever that with a small input movement has the stack make a big swing)
What if you make a slider/funnel/feeder that feeds the bolts slowly (similar to the coin sorter where it Slides down a rail). This way it wont be clogged. If you angle the siv the small motion could be enough to move the bolts downwards(might clog thought) Thanks for sharing you unfinished project, dont see that often.👊😝
Reduce the offset for the pin in the slider. This will reduce the torque loading, enabling the motor to run at a higher RPM. This should transmit a higher frequency oscillation to the bolts with a smaller amplitude.
What about something that's along the lines of cylinders that are made of the mesh holes, they would cascade and have funnels under each that pour into one another from above, each tipped ~45 degrees so parts drop down them and stay in but also engage with the mesh sidewalls as rotated, then you'd really only need to provide a slow continuous rotation instead of harsh impulses for agitation. One idea for these 'mesh cylinders' being easy to print, Make them flat sheets then wrapping them around a disc base and hoop that lock them to being a container, maybe a cone shape would even be more suitable somehow? Just some ideas, else a flat sorter needs an agitator that is more on-par to a paint-shaker.. longevity would be pretty questionable for printed pieces. Also I'm totally going to print those bolt sorters and see if they can help tame my bin of screw woes, very clever sussing out the min and max dimensions of certain common screw types and using the heads for the sorting mechanism, I'd have thought that much too uncorrelatable prior.
Great video! It looks like you would need to pull all the trays out starting at the bottom to get the top bolts if they have any length to them at all. Not a big deal but would making the tray a little taller, with the sieve towards the middle, or more spacing between trays allow for longer bolts and easier access? What if you replace the holes in the sieves with long slots? This would allow the bolts to still fall through if they were to small but catch the desired size. That way the bolt does not have to stand up to make it through the sieve. Do you have any thoughts on how you would then take the sorted bolts and sort each size by length?
Keep the modular part but instead of those forks use pillars of springs around a fork with a shoulder joints. This way you could have a motor on the base that shakes the assembly similar to those earthquake simulators for models. Presumably there is a frequency that creates resonance.
How about using a sloping slit that gradually increases the span. You might have to slope the slip pretty aggressively to make the bolts slide by themselves or introduce a shaker to move the bolts along. Same principle as this grading system for oranges: ruclips.net/user/shortsfR3jD5df56A
I saw issues with it instantly. Why not make the screen a bunch of long ovals instead of teeny tiny holes that do absolutely nothing? Also yes. You need a strong and SHARP shaking motion. Pivots don't provide that unless you go too fast.
just make it bigger. tie it to a suspended plank with rope. hit repeatedly from above with your fists. the returning force is done by te plank. this way the number of parts is increased only by 1 with the rope (unless you count the plank). // You've built quite the contraption which is quite mesmerising to watch. Aesthetics aside for practical applications smaller part numbers are usually better. succes
nice design so far, but i guess you have not enough motion in the contruction overall. It could be possible to overcome that with a solution similar to this: ruclips.net/video/4td_U09cSbQ/видео.html They use it for sorting caps for bottles. Only if the solid side of the cap is near the metal backside of the "escalator", the cap has its center of mass oriented in a way that it does not tip over and would have to try again. You could try to use the design and put your sieve cutouts in the back. But i have to admit, i don't know how you could ensure that nothing get's jammed - eventually by using a big vibration motor like this one ruclips.net/video/zOsQ2BirLf8/видео.html
I've been working on and off on a bolt sorting system for the last 2-3 years. One of the methods I used for aligning the bolts into the holes was a Subwoofer set to between 10 to 30 Hz.
That’s so clever and really out of the box!
Long slots instead of circular holes. That way, the shank of the screw can rotate into the correct orientation naturally under gravity.
Cool build, also.
I thought that too, had some hesitations but I think I figured them out. Slots would be good. The holes will refuse letting a larger screw through. They get clogged up which is why it’s nice to have many. If a longer slot gets a larger screw captive, it’s forced straight down which is optimal for packing. It’d still be much more “hole” per area than many trying to pack circular holes, so much more screw capability. They’d need a much thicker base though to ensure screws going down passed the filter honestly, not by having the head at a wack angle. But slots would definitely make it much easier for screws to find a hole and then orient themselves. That’d take out a lot of the force requirement to get them to happen to be straight up, might even barely work with current motor.
I was thinking the same, slots could be something to try.
If you sort by head size, have the screws slide down a slit (vibration motor assisted) which gets progressivly wider (or stepped). This would prevent the system from getting plugged. A coin sorter might be a good inspiration.
We learn more from failure than we do from success. Thank you so much for sharing this!
@FunkyAbigail Couldn't agree more, and thank you!
It's easy to underestimate the efficiency and power behind good old fashioned hand cranking
I would use rotation and make a tumbling action. Imagine 4 cement mixers or rock tumblers stacked up. Each tumbler is a screen. Don't forget to add a mixing/tumbling blade.
Every time I saw your prints I had a chuckle... even the color was bad lol. I make a lot of jank so I appreciated your jank 😊 great channel
I suspect that part of the problem isn't only movement range or speed, but actually acceleration of the moving axis. Since you're using a circular motion with an offset dowel to provide this cam-lever movement, you end up with a more or less sinusoidal function of movement vs time in the axis in which the drawers are moving. Sine waves aren't very "jerky", the acceleration is rather small at any given time. Maybe changing this system for something that provides a more jerky movement function, such as something closer to a triangle wave (i.e. with very strong changes in the derivative of the position vs time), could provide more of a push to the screws! Maybe mounting the drawers with some spring force that drives them to a position A, and a cam that upon falling smashes the axis and makes them move quickly to a position B?
Honestly a solenoid might do the trick here. But I like the idea of using long slots instead of holes the other comment mentioned.
As a fellow maker of custom screw sorting machines I can say that:
1.) A series of fixed width sieves is a practical way of sorting by diameter without getting too crazy on assembly complexity.
2.) Slots work better than holes but the longer the slot the easier it is to flex and not have a fixed width. Beam deflections scales with L^3 so don't get too long. Stiffness of the beam scales with W^3 as well though so increasing the gap (W) between adjacent slots can help a lot as well.
3.) Massage vibrators are cheap and you can get them at walmart although might not have enough power for the combined mass of the hardware you're trying to get shaking.
4.) If you're going to be at ERRF/3d Printopia, I have an optical + electronic based method to demo this year that accounts for sorting by length and head type.
OMG that is insane. You have a lot of patience doing this. I salute you!
We gave our students the similar problemen. What we saw worked pretty good was using vibration (as otters said) . The teams created that with a motor with an excentric ring, running somewhere around 1000 rpm. This Will create a lot of sound though. With plastic the noise should be less.
I think that a large part of what makes the smaller version work well is the motion on the Z axis, causing the bolts inside to be flung up and down.
Another change I'd make it to fit more holes in each layer, by placing them in a hexagonal grid instead of a rectangular one. Although I think the idea to change them into long slots that @airman2468 mentioned is better
Edit: another idea that came to mind for how to create the Z motion: solenoids! Place 3+ of them below the device, and let them shake the whole thing violently
A similar principle to a semi manual pharmaceutical capsule filler jig might work. It has a self orientation sliding layer to drop the capsules into the holes below in the correct orientation.
The shaking doesn't get the bolts to separate. You need vibration to get them to jump up.
Plus you overloaded it. The feed rate needs to allow for initial sorting, otherwise it will just bridge.
Your inspiration was a vibratory sieve. They also have ones that drain out automatically too, but to the point.
Think you need to vibrate it up and down and maybe add an opening funnel to your slots.
Great job/project. Eagerly anticipating the next rendition.
Yes, I have such container.
The container has even a name:
The workshop...
Why have I never seen your channel before? It's awesome.
Need a vibrator motion to get the bolts through the holes. Maybe a tumbler with different sieve sizes. Change the sieve in larger increments to achieve the sorting
The bolt bucket is an altar to elder gods. When you find a screw, sacrifice it to the bucket. Then, in your hour of need, the bucket gods will provide. All hail the bucket gods.
Very cool build! Could the whole thing be set onto a vibrating table? Like the ones used for casting. Attaching it somehow to something like a vibratory parts tumbler could work. I would think anything that vibrates quickly and violently enough would work.
You could use a weight on a motor to create vibrations, like it's done in game controllers.
Fix the motor to the uppermost tray and fix the lowermost tray on a rubbery mount.
I've seen lots of different types of sorting, and all of the industrial ones use high frequency low amplitude.
Low frequency requires high amplitude, and no matter what you do, they just shake themselves apart. A 3d printed unit is likely to last even shorter.
So, my suggestion is much higher frequency, low amplitude, with some sort of forms/shapes in the bins that will help to turn the screws the right way. But that needs lots of experiments...
Keep the sizing boxes fixed to each other then place the assembly on a separate vibration/shaker unit. The vibration/shaker unit could have a fixing plate which the sorting assembly would attach and detach to. Your sorting can remain modular, but the rigidity of the assembled modules ensures the vibrations and movement are transmitted from top to bottom. If you can modify the vibrations, you can adjust to suit M2 and M3 sorting, vs M8 or larger sizes.
Either that or lift the bolts so with potential energy they can fall down through the sorter, (sorter plates in a series of steps, with the treads angled downwards), and the unsorted bolts be collected and lifted back to the top. May still need some vibration/movement on the steps to sort.
i honestly think a handheld shaker at a large size (think the micronics solution for cleaning parts) is still a better option than the motorized version that, seemingly, has allot of friction in the bearings aswell
since the plastic will form around the balls due to the pressure, is will create a much larger contact patch, causing allot of friction. this is also why the baskets rotate side to side when moving
I think you need more a shaking up and down motion so you pop screws out of holes that the thread fits but head doesnt and so you realign so they can rotate up and down. I would do a circle that grows like a spiral then has a sharp drop and it gets pucked up by springs.
Have the screen tower site on springs then use a higher speed motor with an off balance count weight. Similar to how a shaker deck works.
Use a cheap massage gun for vibration. Pretty helpful for bubble release in concrete forming.
I feel like it's salvageable if you can add some vertical into your agitation. Where can you introduce a cam that could gently lift then drop the trays? Maybe between the trays themselves so the horizontal action forces them up a little then drops them? Not sure if that little motor would cope with all of that though
Cool video! Kudos for posting even though it didn’t work out perfect!
Maybe instead of a back and forth motion, how about a rocking motion? Put a pivot pin in the middle of each drawer and push the forward (or back) up and down. This would require less energy, or force, from the motor to get the screws to move. Replace the holes with slots would also help encourage the screw to rotate easier.
Similar to the rotating drum idea but trying to keep as low profile as possible you could consider moving to as close as a ‘folding’ or ‘tipping’ motion as practicable, where rather than a single screen have two screens of the same sift category ‘tipping’ its load into the other.
I’d imagine the degree you would need to tip would come down to how much you’ve got in there. Unfortunately the more bolts there are the more you have to try and displace each to get them to fall through.
Vibrating sorters are the way to go. You should check them out.
I looks like you should dry your filament. This is my go to when things dont turn out correctly.
drying is soooo important!
Looks like your filament might have a bit of moisture in it (the little voids within a single layer line)
All you really need to dry filament is a heated bed and a box the filament fits in. A small fan in the box will speed up the processes but its not required.
50-55c bed temperature is what I use for pla. 65-70c is what I use for petg.
To me it looks as if there was a far too high layer height set, maybe to speed things up. Better make them in good quality, it may be a source of problems afterwards if low grade prints are being used.
@@MAButh You can print up to 0.3mm layers on a 0.4mm nozzle with out too many issues. Thick layers do tend to highlight imperfections however.
That said, drying the filament won't magically make the part look better but it will fix the voids they have in the layers.
You could experiment with funnel shaped sieves. That could help to bring them all upright. Then the motor may also be strong enough.
Maybe try using a different shaking mechanism. For example, shake a rigid stack, just like in the manual version, but using an up and down motion with a powerful motor. This motion should be large and with a frequency similar to what you would pull of by hand (1-2hz) with an amplitude of a couple centimeters (yes it's a massive amplitude, but similar to what you're doing by hand). This should shake the bolts through the seive. However the seive would need to be tough enough to sustain the dynamic load and not have the stack break apart during shaking.
Tldr: make a mechanical movement similar to what you're doing wuth your hands. (Maybe have a lever that with a small input movement has the stack make a big swing)
This earned you a subscriber :)
What if you make a slider/funnel/feeder that feeds the bolts slowly (similar to the coin sorter where it Slides down a rail). This way it wont be clogged.
If you angle the siv the small motion could be enough to move the bolts downwards(might clog thought)
Thanks for sharing you unfinished project, dont see that often.👊😝
Your peristaltic pump likely generates enough oscillations to power this device. Dual-purpose technology. :D
@ElizabethGreene Love it!...and you're not wrong, that thing has no problems generating vibrations :D
What about an up and down motion instead of the front to back? A camshaft could move up then drop everything down
Reduce the offset for the pin in the slider. This will reduce the torque loading, enabling the motor to run at a higher RPM. This should transmit a higher frequency oscillation to the bolts with a smaller amplitude.
I have about 10.000 screws, bolts, nuts, washers, different lengths, heads, diameters... but obviously I NEVER have the type I need
🤣
Excellent!
Maybe build an oscillating "vibrator/agitator" for the bottom and try and hit the contraption upwards.
Think of a phone vibration motor
What about something that's along the lines of cylinders that are made of the mesh holes, they would cascade and have funnels under each that pour into one another from above, each tipped ~45 degrees so parts drop down them and stay in but also engage with the mesh sidewalls as rotated, then you'd really only need to provide a slow continuous rotation instead of harsh impulses for agitation. One idea for these 'mesh cylinders' being easy to print, Make them flat sheets then wrapping them around a disc base and hoop that lock them to being a container, maybe a cone shape would even be more suitable somehow? Just some ideas, else a flat sorter needs an agitator that is more on-par to a paint-shaker.. longevity would be pretty questionable for printed pieces.
Also I'm totally going to print those bolt sorters and see if they can help tame my bin of screw woes, very clever sussing out the min and max dimensions of certain common screw types and using the heads for the sorting mechanism, I'd have thought that much too uncorrelatable prior.
look into using solenoids for vibrating the modules maybe
Great video!
It looks like you would need to pull all the trays out starting at the bottom to get the top bolts if they have any length to them at all. Not a big deal but would making the tray a little taller, with the sieve towards the middle, or more spacing between trays allow for longer bolts and easier access?
What if you replace the holes in the sieves with long slots? This would allow the bolts to still fall through if they were to small but catch the desired size. That way the bolt does not have to stand up to make it through the sieve.
Do you have any thoughts on how you would then take the sorted bolts and sort each size by length?
I would add a new unit to lift and fall the whole unit by motor and cam at bottom, that way no need to reprint the rest.
Keep the modular part but instead of those forks use pillars of springs around a fork with a shoulder joints. This way you could have a motor on the base that shakes the assembly similar to those earthquake simulators for models. Presumably there is a frequency that creates resonance.
You should get some tupperware and repurpose it as a multi stack pcb etch shaker.
Impressive!
Hack a vibrating cleaner. HF vibrating parts cleaner.
Use a woofer for vibration.
Would rotating drums with long slots work better here ?
Bro telling the first design is too big so he nade it even bigger 😂
Why not go with a tumbler style system. Cylinders instead of drawers that rotate and drop the bolt down a shoot to the next one.
just use a vibratory motor and vibrate the whole thing. fluidize the mass of screws.
How about using a sloping slit that gradually increases the span. You might have to slope the slip pretty aggressively to make the bolts slide by themselves or introduce a shaker to move the bolts along. Same principle as this grading system for oranges: ruclips.net/user/shortsfR3jD5df56A
That print quality is quite sad to look at, anyways quite a cool design, would've never thought of doing it that way.
I saw issues with it instantly.
Why not make the screen a bunch of long ovals instead of teeny tiny holes that do absolutely nothing?
Also yes. You need a strong and SHARP shaking motion.
Pivots don't provide that unless you go too fast.
Concrete vibration motor
Paint mixer?
just make it bigger. tie it to a suspended plank with rope. hit repeatedly from above with your fists. the returning force is done by te plank. this way the number of parts is increased only by 1 with the rope (unless you count the plank). // You've built quite the contraption which is quite mesmerising to watch. Aesthetics aside for practical applications smaller part numbers are usually better. succes
also my dad would pay me some change for sorting his bolts. don't know if thats an option for you tho
It looks like you are printing too fast for your hotend if I were to guess.
use vibration not oscillation
nice design so far, but i guess you have not enough motion in the contruction overall.
It could be possible to overcome that with a solution similar to this: ruclips.net/video/4td_U09cSbQ/видео.html
They use it for sorting caps for bottles. Only if the solid side of the cap is near the metal backside of the "escalator", the cap has its center of mass oriented in a way that it does not tip over and would have to try again. You could try to use the design and put your sieve cutouts in the back. But i have to admit, i don't know how you could ensure that nothing get's jammed - eventually by using a big vibration motor like this one ruclips.net/video/zOsQ2BirLf8/видео.html
What's your layer height? Like 2mm