Thanks everyone for your input! I'm going to pin this comment to highlight some of the helpful input I've gotten, and I'll edit it as more input comes until I inevitably redo this video with all my new knowledge :) - The bearings lock up when under an axial load because they're not designed for that. A bearing that *is* designed to spin under axial load would be a thrust bearing. - A good way to space out bearings on the same shaft is by tightening two nuts against each other. Just make sure you tighten them such that the rod they're on is not under load. - WD-40 is acidic, and may ruin some bearings that have specialty coatings. Your 608/RS/2Z's will be fine though. - The "rings" I referred to are technically races - I misspoke at 0:56. The "correct" notation is SS, because it stands for steel seal. SS, ZZ, 2Z, 22, 2S, they're all the same. SS is technically correct, but it doesn't really matter
Great video! The reason for locking up is the type of bearing. They are designed for vertical loads (therefore it's a vertical bearing). If you use a thrust tapered bearing then it will still spin freely with it tight (depending on how tight of course). It all depends on where the load is coming from. (There's also horizontal bearings as well).
If I recall correctly, The 2S, 2Z suffixes signify the type of seal. 2S means 2 seals (like the ones you prefer in the video. 2S, or maybe written SS) you can get 1S uncommonly and they only have a seal on one side. 2Z are called shields and are used to prevent contact of the seals on both races to reduce seal drag. The don't have as good dust/water resistance though. Can be written 1Z as well with only one shield on one side.
I've used thrust bearings on quite a few things that I've made (knife sharpeners, spinning tool/pen holders, etc.), both ball-types and pin-types, and you can really clamp down on these with a fair bit of axial load. I love using any type or size of bearings in my stuff; I've been fascinated with them since my dad brought home from work a giant (to me at the time) roller bearing, which was around 150mmØ and a few smaller ball bearings that they'd replaced off an induction furnace. Okay, I admit it: I _may_ be quite weird in that respect. 😄
@@troyd-motorsport9933 ⚠️ God has said in the Quran: 🔵 { O mankind, worship your Lord, who created you and those before you, that you may become righteous - ( 2:21 ) 🔴 [He] who made for you the earth a bed [spread out] and the sky a ceiling and sent down from the sky, rain and brought forth thereby fruits as provision for you. So do not attribute to Allah equals while you know [that there is nothing similar to Him]. ( 2:22 ) 🔵 And if you are in doubt about what We have sent down upon Our Servant [Muhammad], then produce a surah the like thereof and call upon your witnesses other than Allah, if you should be truthful. ( 2:23 ) 🔴 But if you do not - and you will never be able to - then fear the Fire, whose fuel is men and stones, prepared for the disbelievers.( 2:24 ) 🔵 And give good tidings to those who believe and do righteous deeds that they will have gardens [in Paradise] beneath which rivers flow. Whenever they are provided with a provision of fruit therefrom, they will say, "This is what we were provided with before." And it is given to them in likeness. And they will have therein purified spouses, and they will abide therein eternally. ( 2:25 ) ⚠️ Quran
You want spacers between your bearings. If you load them perpendicular to their rotation axis, the inner races are squeezed together and the balls inside get pinched between the edges of the inner and outer races.
4:12 that is a quick and easy way to ruin the bearing with the nut on it. Ball bearings are not meant for axial force (like you are putting on it by using a bolt and a nut to pull in another bearing), and will quickly cause indentation damage to the balls and/or the races. Use a press or a clamp for this. If you want to pull in the bearing with a nut and a bolt, use washers on the other end so the force is transmitted either onto the outer-race of the bearing, or directly into the surrounding part. Just a little tip :) I mentor a HS robotics team and I have seen so many bearings ruined by apply force onto the wrong race.
Good point! I suppose I figured if it's taking enough force to insert the bearing that it's ruining it, you probably need a slightly bigger hole 🤣 but thank you for pointing that out, as you are most definitely right.
I wasn't expecting the level of professionalism and the quality and quantity of information provided even within the first few minutes of this video. I normally just watch a few minutes of the beginning of a video, skim the rest in order to get the gist of the information provided, and leave for another one (ADHD is strong with me). However, I'm VERY impressed, and am now subscribing to your channel. I look forward to consuming more of your content.
2:48 This is incorrect. The slicer compensates for the extrusion width. Any discrepancy between the given dimension and the actual is caused by shrinkage of the part as it cools and the tessellated approximation of a solid as a mesh.
The discrepancy is caused by both! I can only speak to Cura -- it definitely doesn't account for extrusion width. I believe on some slicers you can specify if you want your features to be at least their specified dimensions, or at most, but on Cura, it traces the path of the identified diameter. You are right though in that the shrinkage of thermoplastics plays a big part in it, I should have mentioned that!
@@christophersfactory I don't really understand what you mean, I agree with the commenter. The dimension is given by the stl, and if it's a "hole" features the slicer knows it should be 22mm and the path will be slightly outsetted. if it's a "pillar" feature the slicer places the path slightly inwards. Feel free to test this out by setting your extrusion width in the slicer to a comically large number eg. 2mm and you will see the effect - the hole should still be near 22mm. The effect you're seeing is likely coming from poor extrusion tolerances and the fact that the layers aren't perfectly straight Have a good day
@@christophersfactory Cura absolutely compensates for the extrusion width. I've tested it by exporting the toolpaths as a mesh file. This generates an OBJ with all toolpaths modeled with their respective heights and widths. A 20 mm hole measured 20 mm across "corners." As id said, the slicer knows whether features are internal or external based on the normals of the surfaces the mesh is composed of.
Seems like it would be an oversight for a slicer not to compensate for extrusion width, especially when it knows what your extrusion width is to begin with. I think it has more to do with hardware inaccuracy/thermal expansion/contraction than the slicer not compensating… but I will say when working with tight tolerances, my go to tolerance is 0.2-0.3.
I will add a little bit of nuance to this discussion and say that slicers using the slic3r engine (prusaslicer, super slicer, bambu studio, orca slicer, etc) actually do give slightly inaccurate walls, but for a different reason. Slic3r overlaps perimeters for better strength, but the results is dimensions that are a tiny bit off. In Orca Slicer fixing this is as simple as checking the "use precise walls" checkbox (at the risk of reduced part strength, and in extreme cases delamination of the outermost perimeter). Cura, however, does not overlap perimeters (as far as I know), so this is particular cause does not apply there
Thank you, Christopher (and commenters)! I'm a novice just starting out on a prototype project and this will undoubtedly save me soooo much time, trouble, and expense! Very much appreciated.
The main reason that shoulder bolts are much more expensive than the other types you talked about is because of their precision. Their diameter is much more accurate than common hardware store bolts, but that comes at an increased cost. As you noted, they're pretty unnecessary for 3d printed parts. They're designed and used a lot in accurate positioning and fixturing in machinery and manufacturing respectively.
Interesting. It seems even more strange then that an 8mm precision shoulder bolt still isn't 8mm... Are there any bolts that you know of that have an exactly 8mm shaft, with M8 threads at the end? P.S. Thanks for the info!!
@@christophersfactory it's very unusual to see a 8mm shoulder with m8 thread - m6 is the standard. I'm also not sure how you see as much slop on shoulder bolts as you do on normal ones - standard tolerances are on the order of
Friction fit is another term for "Press Fit" or compression fit in metal machined parts. The acceptor metal opening is expected to expand and be under compression. You may try to create an opening with a serrated inner diameter which will deform more easily. Roughening the outer diameter of the bearing may help.
Good thought on the serrated inner surface; triangular teeth on the ID would crush progressively, and should keep the bearing pretty well centered as it’s pressed in, without offering as much resistance as a solid wall. 👍
I chased my tale for ages (about 3 months) with friction fits into plastic (with pins rather than bearings) for a project - trying to optimise my hole sizes, making jigs etc - till I realised that friction fits need precision holes - just not repeatable in fdm plastic - as when you come back and try the same In 3 months it doesn’t seem to work - but most importantly - plastics are prone to creep under static load, so if deformed elastically to hold a part by friction in time it will relax. Nylon - which is less prone to creep changes its dimension with humidity - so that will lose grip because the wa rather changes ! Best stick with glue was my conclusion - as even at screws are still relying on elastic deformation of your 3d part …
He does not claim that it is. He claimed that for his purpose, an unlubricated part would be enough. The steel balls probably won't wear down in the lifetime of his 3d printed machine.
The slop in the fit between the bolt and the inner race is because most bearings are made to go on precision shafts, not bolts. You can order shaft material pretty cheap, especially in the 8mm size. Also, the best way to space bearings on shafts is with external retention rings, but since those require at least a lathe and grooving tool, your next best bet is with shaft collars. They tighten on the shaft wherever you want by way of clamping bolts and/or setscrews.
Glad I could help you fix your nut problems! Sounds like lubricant was your issue all along. In a future video, maybe I can help you calibrate your extruder flow! 😉
For the nut and bolt thing at 6:45 , you can fix this by putting on the first nut up to where you want it, but not too tight, then put on a second nut, and then tighten them into each other.
@@christophersfactory This is how quick-release bike wheels are assembled but instead of a pre-assembled bearing cartridge, the hub has cup-and-cone races with loose ball bearings. Two nuts are tightened against each other on the hollow axle/bolt to hold everything together without squeezing the bearings to death (seizing up). The quick-release skewer (the part most people manipulate to change a flat tire, etc) goes through the axle and is just there to install the wheel on the bike, it doesn't really apply any force to the bearings or hub. Applying the right force with the lockl nuts to hold things together is a bit of an art. Too little or too much can damage the bearing races and/or make the wheel harder to spin.
@@ItsWami I don’t think a locknut would do the same thing: You want to load the bearings with a specific compressive force, then lock the nuts at that position with check-nuts. Locknuts either rely on the force spooled against their flange to keep them from turning, or have nylon inserts to do the same. I don’t think Ny-loks would offer enough resistance to turning to be relied upon in something like the bike-hub application.
Make to hole just smaller than the bearing so that the bearing will not go in but almost will. Put the bearing in the freezer for an hour or so. The steel will contract a small amount but should be enough to push the bearing in the hole. When the temperature rises back up it will expand and tighten the fit to a friction fit. A lil glue wouldn’t hurt but not really needed
that didn't work at all, it made no difference. this might work with metal or if you have very large bearings but mine is only 22 mm OD and the PLA is too soft for this to work. the tiny change in OD form the temperature change makes no difference at all.
@@dronefootage2778 This trick only works if you can very precisely control tolerances between fitted parts and those parts have rather similar thermal expansion coefficients. You can't do that with "standard" 3D printing.
2:54 this is incorrect, slicers do take into account the nozzle diameter. The issue comes down to thermal expansion of the filament, filament getting dragged after its extruded, retraction settings and issues with the machine like E steps, squareness of the frame and axis etc.
Never screw your regular ball bearings the way he does it at the 4:12 mark. This will ruin the bearings. Instead, take two appropriately sized washers so you can clamp on the outer race and screw together that way instead. An alternative to this is using either thrust bearings, tapered roller bearings or angular contact bearings, all of which take MUCH higher axial loads, these kinds of bearings will handle this kind of treatment just fine whereas regular ball bearings will not!
In the robotics pieces I've crafted, I have used copper sheeting (easy to get rolls of it cheap) to create a sleeve that wraps around the 5/16" threaded rod. That takes up the additional space much like your teflon tape method you mentioned, but the teflon tape can continue to deform and a thin copper sheet does that 'less'. Also, it won't degrade as much if I end up needing to lubricate the bearings later. Some kinds of plumbers tape (or similar) can melt in the presence of lubricant/solvent mixes (like WD40). I suppose aluminum foil would also work. Great content! Please keep up the good work. :)
If you want them to spin you need thrust bearings not radial bearings. Radial bearings are designed to take a load on the outside race but not a side load. Thrust bearings are designed with preloaded compression in mind like on a lathe.
2RS = 2 rubber sheild RS = One side rubbe sheild ZZ is not mistyped, many mgf use ZZ. One Z is one side sheild ZZ is sheild both sides. But if the shild is stamped ZZ, this is incorrect. One shield shouild only have one Z on it. Pedantic, I know :-)
No kidding?! This is the first explanation I've seen that made sense, and I've seen a lot. I was told SS just stands for steel seal. Can you tell me this? -- why would one want seals on only one side of the bearing?
@@christophersfactory A one sided sheild is often used in motorbike gearbox for example. Oil flows to the bearing but not right through so oil flows down a oliway.
@@christophersfactory The SS or S prefix is usually stainless steel. it should also be noted that the "R" seals are contact seals. Some seals (eg, designated as V or VV or LLB depending on the manufacturer) are non-contact seals and are used in conditions that are relatively clean and require a high-speed, low friction bearing. There are MANY different types of seals and configurations and designations!
If you use two bearings and crank them together with a bolt the balls on each bearing get compressed into the rings. At a microscopic level the metal has to deform to spin. This is why it becomes harder to spin. There’s a sweet spot of tightness where the shaft runs true and freely. It’s good to use a Belleville washer and two nuts with this configuration to get the best results.
6:41 In short, your problem is due to overconstraint. Underconstraint is the reason why you had the leverage with a single bearing (5:52). I highly recommend you to read "Exact constraint machine design" by Douglass L. Blanding. Its an amazingly good book, and a true eye opener for anyone that likes to design mechanisms and machines. Dont worry, it doesnt have any math.
@@alexwang982 For an object to be fixed, we need to constrain 6 degrees of freedom. Restricting ourselves for now to parallel mechanisms, we can imagine any constraint as a infinitesimally thin beam connected between the fixed world and the object. The question is, how many constraints should we place, and where should we place them. For a problem in 3 dimensional space, we need exactly 6 constraints that independently restrict a single degree of freedom each. More details on how to place them (or a more detailed/better explanation of the concept) can be found in the book from Blanding, or various places online. Regardless, you can consider a bearing a constraint of 2. It constrains translational motion in the plane of the bearing. It releases rotation in all degrees of freedom (rotation around the 'main axis', and also around the other 2 axis as shown in 5:52). Adding an additional bearing as in 6:16 fixates the two other rotational degrees of freedom. Counting our DOFs, this leaves just 2: Rotation around the center axis (which is the one we desire to be left open), and translation along this axis. Now the problem discussed at 6:27 is the constraint of this final degree of freedom. In this case, you try to constrain it from both sides with a rigid element/connection, the bolt itself and the nut. These two constraints essentially do the same thing of fixing the location of the bolt in space, and are acting along the same translational line to constrain this degree of freedom. This means that either you will end up with play (i.e. the nut is not very tight, shaft can rock back and forth) or excessive stress, wear and other problems (as in the video). In short, in the video the overconstraint of the axial translation by the nut and bolt causes excessive forces onto the bearing. In the video the problem is exacerbated by the fact that radial ball bearings cannot deal with these axial loads very well at all.
Exciting! For me, it was one of the best purchases I've made in years. The enjoyment I've gotten out of it is completely disproportionate to the cost. Cheers!
Nice video! To prevent that nut from spinning loose, you can use another nut to lock it in place. Set your first nut right where you want it along the threaded shaft (just enough to let the bearings spin freely). Using two wrenches, torque the second nut tightly to the first. That pressure locks them both in place. the 'locking' nut can be removed easily when you want to disassemble it.
The other tip I would add is to add a small chamfer around the edge of the bearing recess to compensate for elephant's foot if it will be on the bottom of the print.
HOLY SHEEEET! The bearings in your link are now down to $19.59; the same bearings over the pond, here in the UK, cost £71.27! 😲 I can even get them with 7-day delivery from the US for £34. Yay for Brexsh|t and an incompetent government. 🤦♂ Anyway, thanks for the video; I've just subbed. 👍
Holy cow! That’s quite the discrepancy. I’m not familiar enough with the politics of Brexit but that’s certainly unfortunate. Cheers, friend, thanks for the comment 🙏🏻 glad you’re here
I had to chuckle at the use of MS Paint for a crude "hand drawn" explanation of tool compensation, immediately followed by some footage of precision drawing in Fusion. As with everyone of your videos, I learned some things. Keep it up!
I forgot to say this in reply to the last comment of yours I saw, but I have to thank you for your positivity and kind words. It's very inspiring to me to keep doing this channel and make more videos. It goes a long way. Thanks again, from the bottom of my heart. Christopher
@@christophersfactory You are very welcome. There's too much negativity on the internet. A little kindness can go a long way. You mentioned a few days ago that you finally received the electronics you needed to start collecting electricity from your windmill. Will there be a video showing the results coming soon? 🍿
@@schmidtyyt You're awesome. Yes, the parts finally came, but I realized I had a fatal flaw that rendered the PCB's I designed nearly useless. I saved a few by cutting the traces out with a utility knife and re-routing some connections with hook-up wire, but being the perfectionist I am, I didn't want to show that in the video, haha. On top of that, I redesigned the turbine, but I'm now in the process of redesigning it again. I've used pre-cut plywood up to this point, but I think to have it work really well, I need to get better at learning how to cut and work with thicker and more durable woods. The plywood I've been using warps horribly. I just started working with this stuff called hardboard, I think it's made from compressed sawdust. I can already tell it's much less prone to warping, but it's slightly harder to work with. I'll probably have to redesign the turbine hub to accommodate the heavier and thicker boards.... this is all to say that there's a lot I'm re-working; I told myself I'd wait to get the "build video" out until I was done, but now I'm thinking I should just make regular, almost vlog-style updates on what I'm doing. Viewers would probably get more benefit out of that anyway, because learning what doesn't work is arguably just as valuable as learning what does. I have a few other non-windmill-related videos already shot and edited scheduled for this week, but windmill content should start back up next week, especially after these words of encouragement :)
@@christophersfactory That sucks about the PCBs not working as hoped. I know nobody likes to admit failure much less broadcast it to the world, but showing what went wrong may help others avoid the same mistake. As a wise man recently said, "learning what doesn't work is arguably just as valuable as learning what does." 😉 I'd expect the hardboard to work better for you, as well. And I totally get the design re-considerations that brings, too. However, I have to ask: why not 3D print them? As for working with the hardboard, check out the Dremel Ultra Saw (great Black Friday deal at Amazon today). I found it to be a great little saw for cutting thin wood and metal.
@@schmidtyyt I completely agree about the failure bit. That was one of my goals when I first made this channel, was to candidly show people my efforts in these projects, not just showcase a finished project. As for why I'm not particularly inclined (although not opposed) to printed blades, there are a few reasons. PLA warps really badly in the sun, especially with large and thin features. If the blades are not thin, it will use a lot of PLA, which is much more expensive than wood. I could try printing in nylon or ABS, but I've never worked with either before. My short experience with wind turbines has taught me that the name of the game is wind-collecting surface area. This is faster to design with panels of wood, I think. It requires less design consideration. Additionally, I could probably cut 250 wooden blades in any afternoon, but 3D-printing them is going to require a significant amount of time. So, to answer the question more directly, I guess there isn't really anything that I've found that strictly deters me from 3D printing the blades, but rather several small things that incline me toward wooden blades instead. People often ask me what the goal of this windmill project is, as I've been working on it for well over two years and through twenty-one designs. Really, there is no goal. I do this because it makes me feel happy and fulfilled. I'm not necessarily trying to create a to-market product, nor am I trying to design something that anyone could assemble with only filament and a printer. So, as the Cheshire Cat said, “If you don't know where you want to go, then it doesn't matter which path you take." I'm just taking whatever path I feel like and feeling things out. Maybe I'll take the path of 3D printed blades at some point. That Dremel looks pretty slick. I have a knock-off rotary tool, I wonder if it could use Dremel saw blades? Something to look into for sure. How about you? I think you're probably my most loyal subscriber already, haha. What gravitates you toward this content? Any plans to build a windmill of your own?
This just what I am looking for. A few bearings in the mix open up so many possibilities. Thanks for all your diligence. I'd like to show you a few of my designs.
A vice is something that everyone should have and is very good for press fitting parts, like those bearings and is probably a bit quicker to set up than the nut and bolt solution.
@@christophersfactory my local supermarket (grocery store) sells miscellaneous crap on a weekly rotation so I picked up a couple of cheap vises of various sizes. Mine can grab parts up to about 150mm in the dimension you're squeezing, and as Conor said they're great for 3d printed and maker stuff. That bigger one cost like €15 max
@@christophersfactory they really aren’t that expensive and you don’t need large ones that bolt down to workbenches, even a cheap one that just clamps onto a desk is enough for most people.
After nuts and bolts, bearings are one of those things I buy in excess, in sizes I don’t need for ideas I haven’t thought of just because they are so satisfying. Sometimes when I really need some relieve I’ll slide a shaft in and give it a spin. Maybe slide it back and forth if it’s a snug fit. I started collecting used copy machines to take apart because people just give them away and I’ve pulled literally buckets of bearings out. I took apart a plotter and it had a 1m 13mm slide mounted to a sturdy boxed beam. No clue what to do with it but I ordered a set of 13mm sliding blocks just in case. It would make a good X | Y axis on a CNC router.
Great video, and helpful info in the comments. I'm thinking about making some 3D printed projects with wheels, but I had trouble finding beginners advice. This video was the first useful thing that came up.
Great video and I'm more intrigued at the wind generator! Fantastic point about logging all the research. I'm probably going to create a form to print out for the things I work on as I normally wing it. Then end up printing the same mistaken part twice, or not really try to push myself beyond the normal limits. I can see documenting my stages and steps is critical. Usually I'm the "oh that needs X" and adjust it a tad and then start again, but yes I do find myself wondering which version I'm about to use by the third round. One idea would be to make a cap that fits over the bearing as well that would allow both friction fit an C.A. Glue so the bezel of the cap holds it in place.
6:40 the reason is because those are radial bearings. They're design to be able to take force in the radial direction (think of where they go in a skateboard). However, they are not designed to take forces that are directed along their axis of spin such as thrust forces (Think of an office chair that swirls around with you on top of it). For this, you want thrust bearings (hence the name). These are designed to take this kind of compression force along the axis of spin.
Your printer isn't calibrated right. Slicers take line width into account. They are sophisticated pieces of software. You only need to consider actual tolerance required and the impact different filament and environments conditions has. In reality, print mechanical parts with a 0.6mm nozzle and similar line width (those are two different things in a slicer) I find the last step to fit is easily achieved by filing the rough bits off which also ensures that the fit won't get loose over time.
I'm 99.9% sure what explanation about what the slicer doing at 3:05 is wrong. This may have been true when 3d printing first started but modern slicers can certainly account for the thickness of the bead being laid down. That's not the reason your bearing doesn't fit into that hole.
My favourite "take the entire weight of a person" story is bike spokes. How can spokes hold up the entire weight of a person? The spokes *above* the hub hold you up with *tensile* strength. The load path is you > hub > spokes > top arch of wheel > bottom arch of wheel > ground
Instead of brass inserts, have you tried to embed steel nuts in a print? I design in recesses for the nut and screw, including tolerance for nut thickness and in my slicer I put a pause after the recess is printed, before the next layer. I then place the nut in the recess and resume the print. The next layer covers the nut. I do the same with magnets.
Sometimes I'll design a feature that can accommodate a steel nut, although I've never tried embedding them. Doing it during the print process certainly sounds more promising than trying to solder it in. Does that method work pretty well for you? What I'll often do is just make a square hole that's the height of my nut, and slide my nut in sideways through it until the screw can meet it. Perhaps a strength test video is in order 👀
@@christophersfactory Embedding nuts and magnets works very well for me. When I'm in a hurry I'll use the open square hole method, but when I'm fine tuning production prints I'll embed. I haven't tested strength but it seems to me if the wall thickness between the nut and the head of the screw is the same they should be about equally strong, though I think with an embed it might be a bit stronger, because the nut is pretty much locked into place while the opening for a sliding nut has to be pretty large to accommodate tolerances on the square nuts, resulting in some slop. In any case, embedding the nuts or magnets (when embedding magnets I use a tiny drop of CA adhesive in the recess to make sure they won't jump out of place, attracted to the other magnets) results in a cleaner exterior surface that looks less DIY.
@@lovrocatela8727 If you're imbedding a nut, you'll want to wait until the past possible opportunity in the print to do so, otherwise it's likely that the nozzle will bump it. Personally, I tend to use that method less than just making features in the print that nuts can fit into.
@@lovrocatela8727 After initial slicing, I insert a pause at the first level above the cavity for the nut and then reslice. I make sure that the cavity for the nut is deep enough to accommodate the dimensional tolerances of the nuts to make sure there's no chance of the extruder nozzle crashing into the nut. I also block supports in the cavities for the nut and screw. The layer above the nut is printed as a bridge. While that first layer isn't always perfect, it's almost always good enough to retain the nut and then subsequent layers go down just fine. The pause feature is availabe in the Prusa slicer (and Bambu slicer, too, which is based on the Prusa). I don't have to monitor the printer as the gcode will tell it to pause the print. The Prusa printers will beep to notify you that the print has been paused. I then insert the nut(s) and resume the print.
I need to spin something like your mill, but smaller and much much much faster >20k rpm (200Hz) range. Balanced but not by an expert. Do you think your method could work? It's almost identical to what drew up
6:42 It's likely that your bearings are not axially concentric. 3D printing and friction fitting will introduce a lot of dimensional inaccuracy causing the 2 bearings to be misaligned. If the nut is screwed down tight, the center of the off-axis bearing will worm its way up/down the bolt as it spins causing it to seize. A potential fix: Use a long bushing or sleeve to span the gap between the two inner rings. A brass tube or even a 3d printed tube might work. The tube will force the rings to be concentric as the nut is tightened down. Aligning the outer rings will be more challenging though; you might leave the second bearing loose in its housing to help. Alternatively, you could use a whole stack of bearings to support the entire length of the bolt or mount the whole bearing assembly in a pipe.
If you want to use a nut to clamp the two bearing together (without them seizing), try using tapered roller bearings. They can handle axial loading (thrust) and radial loading. Try McMaster-Carr part number: 6677K51
I have seen many videos on dimensional accuracy, and you are the first video ive seen that explained it correctly by using microsoft paint! lol. Very well made video!
A crown nut and cotter pin are meant to keep a nut in place without the nut being very tight. It's a bit more complicated to build since the bolt or shaft needs a hole for the cotter pin in the right place but it looks cleaner and is more compact than two nuts jammed together.
WD40 will eventually turn to goo or varnish. In the clock restoration business we hated it. Flushing out the bearings with solvent like naphtha then adding a light oil will likely give better long term service as the cheap grease used in low cost bearings will be gone rather than thinned in WD40 which will mostly evaporate and leave yet more goo behind.
Possibly mentioned already, bear with me if this is redundant : Assuming you have room in the build - To secure the threaded end of the bolt and keep it from loosening or being too tight *use two nuts *set the first nut against the bearing / socket *thread the second nut against the first *hold the first in the exact position you want and tighten the second nut against it (tighter the better, and thread lock if you like). The torque between the nuts will keep either from moving along the thread. Hope this makes sense and hope it helps :)
Very cool, thanks for the tips. I tried to design some TPU based wheels for my baby stroller, and had the bearing with bolt acting as lever slop issue. This created quite a sag when a load (the baby) was placed into the stroller. I didn't think of the two bearings as far apart acting as a choke idea, that's great!
they are races not rings. the reason they lock up when you tighten the bolt is because they are coaxial bearings not trust bearings. the balls bunch up against the side of the race and causes misalignment in the races and this multiplies the issue and locks up.
I find 0.15mm added to the OD of the bearing is good for press fit. And it's not good to force bearings into place by the inner race, unless you're pressing it onto a shaft. You can damage the bearing. I made a 3d printed tool to press in, and remove, bearings in pulleys. It has a tube that only rests on the outer race and uses a bolt that's the same diameter as the bearing's ID.
Great video. Informative and great pacing. Do a followup about spacers maybe (what you to keep them from locking up). Having disassembled a ton of printers I learned they are used constantly, and so old printers are a good source of pretty decent metallic spacers (although printed spacers do the job too).
Mahalo , just got my 100 pack bearing from Amazon.. arrived in 3 days.. Well also try your measurements for the 3d printers to fit them into my projects.
That's totally a doable method, it can be a little difficult to ensure that the plastic only deforms where you want it to, and I would wager that that bond is still not as strong as just using superglue.
The part about 3D printers just doing "whatever happens happens" is (fortunately) not true ... unless your printer isn't properly calibrated (which unfortunately happens often!). What *is* true is that usually we're talking about printers that are not industrial-grade - they're consumer-grade! But yes, slicers *do* take this kind of problem into account (even if imperfectly). 🙂
Another method instead of glue could be to pause the print at the right layer number, put in the bearing, and then continue the print, printing also covering the other side just enough.
I'd be interested to try this! I haven't done much with inserting parts mid-print. I know many people do, to fair success. I just worry about the nozzle colliding with the part. Do you know of any way to avoid that?
In cura there is a setting I believe it is called slicing tolerance (not really sure off hand) but the choices are inclusive and exclusive. This will let you tell cura if it needs to be at least or at most and it will put all .4 on the side of the line it needs to be.
When u use two bearings one should be fixed in place while other allowed to slide on on at least one of the parts. There was long explaination on my uni why it was, but i remmember analysis of forces on that was nightmare (like literally impossible to do static analysis on that) and just go for one of them sliding
I think the reason people don't like using wd 40 is it's acidic and may cause issue with certain coatings. made from fish oil apparently. it's great to get rid of water.
There's another way to solve the problem of a model that has a hole that is too big or too small for the bearing: change the horizontal expansion setting in the print settings. Usually -0.1mm to +0.1mm does it for me.
Interesting -- where is that located in Cura? I've seen several people talk about it but the only one I've ever seen is the vertical tolerance in the experimental settings. Thanks for the comment and kind words :)
If you learn best with your hands, take a look at how a bicycle wheel bearing stack is set up to better understand how the stress physics of a windmill wheel could be minimized. Bicycle wheels have been on the engineering block for a long time and optimized to gain an edge to win world championships.
Dont recommend using set screws to secure bearings you will deform the outer race and make it oval. It's better to use a clamping style collar or press fit.
Thanks everyone for your input! I'm going to pin this comment to highlight some of the helpful input I've gotten, and I'll edit it as more input comes until I inevitably redo this video with all my new knowledge :)
- The bearings lock up when under an axial load because they're not designed for that. A bearing that *is* designed to spin under axial load would be a thrust bearing.
- A good way to space out bearings on the same shaft is by tightening two nuts against each other. Just make sure you tighten them such that the rod they're on is not under load.
- WD-40 is acidic, and may ruin some bearings that have specialty coatings. Your 608/RS/2Z's will be fine though.
- The "rings" I referred to are technically races
- I misspoke at 0:56. The "correct" notation is SS, because it stands for steel seal. SS, ZZ, 2Z, 22, 2S, they're all the same. SS is technically correct, but it doesn't really matter
Great video! The reason for locking up is the type of bearing. They are designed for vertical loads (therefore it's a vertical bearing). If you use a thrust tapered bearing then it will still spin freely with it tight (depending on how tight of course). It all depends on where the load is coming from. (There's also horizontal bearings as well).
Hey, why didn’t you sponsor it? There’s a lot of views ,I’ll bet a lot of people ordered. You should do it. I ordered them just now.
If I recall correctly, The 2S, 2Z suffixes signify the type of seal. 2S means 2 seals (like the ones you prefer in the video. 2S, or maybe written SS) you can get 1S uncommonly and they only have a seal on one side. 2Z are called shields and are used to prevent contact of the seals on both races to reduce seal drag. The don't have as good dust/water resistance though. Can be written 1Z as well with only one shield on one side.
I've used thrust bearings on quite a few things that I've made (knife sharpeners, spinning tool/pen holders, etc.), both ball-types and pin-types, and you can really clamp down on these with a fair bit of axial load.
I love using any type or size of bearings in my stuff; I've been fascinated with them since my dad brought home from work a giant (to me at the time) roller bearing, which was around 150mmØ and a few smaller ball bearings that they'd replaced off an induction furnace. Okay, I admit it: I _may_ be quite weird in that respect. 😄
@@troyd-motorsport9933 ⚠️ God has said in the Quran:
🔵 { O mankind, worship your Lord, who created you and those before you, that you may become righteous - ( 2:21 )
🔴 [He] who made for you the earth a bed [spread out] and the sky a ceiling and sent down from the sky, rain and brought forth thereby fruits as provision for you. So do not attribute to Allah equals while you know [that there is nothing similar to Him]. ( 2:22 )
🔵 And if you are in doubt about what We have sent down upon Our Servant [Muhammad], then produce a surah the like thereof and call upon your witnesses other than Allah, if you should be truthful. ( 2:23 )
🔴 But if you do not - and you will never be able to - then fear the Fire, whose fuel is men and stones, prepared for the disbelievers.( 2:24 )
🔵 And give good tidings to those who believe and do righteous deeds that they will have gardens [in Paradise] beneath which rivers flow. Whenever they are provided with a provision of fruit therefrom, they will say, "This is what we were provided with before." And it is given to them in likeness. And they will have therein purified spouses, and they will abide therein eternally. ( 2:25 )
⚠️ Quran
You want spacers between your bearings. If you load them perpendicular to their rotation axis, the inner races are squeezed together and the balls inside get pinched between the edges of the inner and outer races.
Interesting. I figured it was something like that, but glad someone smarter can verify! Thank you!
Indeed If you supported the inner race you would find the pressure of the nut has no effect
@@scatdawg1 right on!
yeah this is the issue, put spacers on either side and it wont lock up
Yeah, good point. Nobody wants their balls pinched 🤌🏾
4:12 that is a quick and easy way to ruin the bearing with the nut on it. Ball bearings are not meant for axial force (like you are putting on it by using a bolt and a nut to pull in another bearing), and will quickly cause indentation damage to the balls and/or the races. Use a press or a clamp for this. If you want to pull in the bearing with a nut and a bolt, use washers on the other end so the force is transmitted either onto the outer-race of the bearing, or directly into the surrounding part. Just a little tip :) I mentor a HS robotics team and I have seen so many bearings ruined by apply force onto the wrong race.
Good point! I suppose I figured if it's taking enough force to insert the bearing that it's ruining it, you probably need a slightly bigger hole 🤣 but thank you for pointing that out, as you are most definitely right.
I wasn't expecting the level of professionalism and the quality and quantity of information provided even within the first few minutes of this video. I normally just watch a few minutes of the beginning of a video, skim the rest in order to get the gist of the information provided, and leave for another one (ADHD is strong with me). However, I'm VERY impressed, and am now subscribing to your channel. I look forward to consuming more of your content.
2:48 This is incorrect. The slicer compensates for the extrusion width. Any discrepancy between the given dimension and the actual is caused by shrinkage of the part as it cools and the tessellated approximation of a solid as a mesh.
The discrepancy is caused by both! I can only speak to Cura -- it definitely doesn't account for extrusion width. I believe on some slicers you can specify if you want your features to be at least their specified dimensions, or at most, but on Cura, it traces the path of the identified diameter.
You are right though in that the shrinkage of thermoplastics plays a big part in it, I should have mentioned that!
@@christophersfactory I don't really understand what you mean, I agree with the commenter. The dimension is given by the stl, and if it's a "hole" features the slicer knows it should be 22mm and the path will be slightly outsetted. if it's a "pillar" feature the slicer places the path slightly inwards. Feel free to test this out by setting your extrusion width in the slicer to a comically large number eg. 2mm and you will see the effect - the hole should still be near 22mm.
The effect you're seeing is likely coming from poor extrusion tolerances and the fact that the layers aren't perfectly straight
Have a good day
@@christophersfactory Cura absolutely compensates for the extrusion width. I've tested it by exporting the toolpaths as a mesh file. This generates an OBJ with all toolpaths modeled with their respective heights and widths. A 20 mm hole measured 20 mm across "corners." As id said, the slicer knows whether features are internal or external based on the normals of the surfaces the mesh is composed of.
Seems like it would be an oversight for a slicer not to compensate for extrusion width, especially when it knows what your extrusion width is to begin with. I think it has more to do with hardware inaccuracy/thermal expansion/contraction than the slicer not compensating… but I will say when working with tight tolerances, my go to tolerance is 0.2-0.3.
I will add a little bit of nuance to this discussion and say that slicers using the slic3r engine (prusaslicer, super slicer, bambu studio, orca slicer, etc) actually do give slightly inaccurate walls, but for a different reason. Slic3r overlaps perimeters for better strength, but the results is dimensions that are a tiny bit off. In Orca Slicer fixing this is as simple as checking the "use precise walls" checkbox (at the risk of reduced part strength, and in extreme cases delamination of the outermost perimeter). Cura, however, does not overlap perimeters (as far as I know), so this is particular cause does not apply there
Thank you, Christopher (and commenters)! I'm a novice just starting out on a prototype project and this will undoubtedly save me soooo much time, trouble, and expense! Very much appreciated.
The main reason that shoulder bolts are much more expensive than the other types you talked about is because of their precision. Their diameter is much more accurate than common hardware store bolts, but that comes at an increased cost. As you noted, they're pretty unnecessary for 3d printed parts. They're designed and used a lot in accurate positioning and fixturing in machinery and manufacturing respectively.
Interesting. It seems even more strange then that an 8mm precision shoulder bolt still isn't 8mm... Are there any bolts that you know of that have an exactly 8mm shaft, with M8 threads at the end?
P.S. Thanks for the info!!
@@christophersfactory it's very unusual to see a 8mm shoulder with m8 thread - m6 is the standard.
I'm also not sure how you see as much slop on shoulder bolts as you do on normal ones - standard tolerances are on the order of
@@christophersfactory shoulder bolts need to be "precision ground" for bearing applications, normal ones are undersized by up to 0.1mm
Friction fit is another term for "Press Fit" or compression fit in metal machined parts. The acceptor metal opening is expected to expand and be under compression. You may try to create an opening with a serrated inner diameter which will deform more easily. Roughening the outer diameter of the bearing may help.
Aka interference fit.
Good thought on the serrated inner surface; triangular teeth on the ID would crush progressively, and should keep the bearing pretty well centered as it’s pressed in, without offering as much resistance as a solid wall. 👍
I chased my tale for ages (about 3 months) with friction fits into plastic (with pins rather than bearings) for a project - trying to optimise my hole sizes, making jigs etc - till I realised that friction fits need precision holes - just not repeatable in fdm plastic - as when you come back and try the same In 3 months it doesn’t seem to work - but most importantly - plastics are prone to creep under static load, so if deformed elastically to hold a part by friction in time it will relax.
Nylon - which is less prone to creep changes its dimension with humidity - so that will lose grip because the wa rather changes ! Best stick with glue was my conclusion - as even at screws are still relying on elastic deformation of your 3d part …
WD-40 is not a lubricant, as it was never designed as such. Rather use a bearing grease as they are cheap and used in the automotive industry.
He does not claim that it is. He claimed that for his purpose, an unlubricated part would be enough. The steel balls probably won't wear down in the lifetime of his 3d printed machine.
he is trying to delub monkey brain
The slop in the fit between the bolt and the inner race is because most bearings are made to go on precision shafts, not bolts. You can order shaft material pretty cheap, especially in the 8mm size.
Also, the best way to space bearings on shafts is with external retention rings, but since those require at least a lathe and grooving tool, your next best bet is with shaft collars. They tighten on the shaft wherever you want by way of clamping bolts and/or setscrews.
Love this channel for all my nut problems❤❤❤ also other problems I’ve had with my shaft as well! Thanks for the tips Cristian(:
Glad I could help you fix your nut problems! Sounds like lubricant was your issue all along. In a future video, maybe I can help you calibrate your extruder flow! 😉
After hours of skimming through youtube I finally found a video that actually makes sense
For the nut and bolt thing at 6:45 , you can fix this by putting on the first nut up to where you want it, but not too tight, then put on a second nut, and then tighten them into each other.
This is such a good remedy for that! Good thinking, thanks!
@@christophersfactory This is how quick-release bike wheels are assembled but instead of a pre-assembled bearing cartridge, the hub has cup-and-cone races with loose ball bearings. Two nuts are tightened against each other on the hollow axle/bolt to hold everything together without squeezing the bearings to death (seizing up). The quick-release skewer (the part most people manipulate to change a flat tire, etc) goes through the axle and is just there to install the wheel on the bike, it doesn't really apply any force to the bearings or hub.
Applying the right force with the lockl nuts to hold things together is a bit of an art. Too little or too much can damage the bearing races and/or make the wheel harder to spin.
Or use a locknut :)
@@ItsWami I don’t think a locknut would do the same thing: You want to load the bearings with a specific compressive force, then lock the nuts at that position with check-nuts. Locknuts either rely on the force spooled against their flange to keep them from turning, or have nylon inserts to do the same. I don’t think Ny-loks would offer enough resistance to turning to be relied upon in something like the bike-hub application.
Make to hole just smaller than the bearing so that the bearing will not go in but almost will. Put the bearing in the freezer for an hour or so. The steel will contract a small amount but should be enough to push the bearing in the hole. When the temperature rises back up it will expand and tighten the fit to a friction fit. A lil glue wouldn’t hurt but not really needed
Interesting! I wouldn't have thought it contracts enough to work. I'll have to try this! Thanks!
wow thank you! that might be the answer i was looking for. i will let you know how it went.
that didn't work at all, it made no difference. this might work with metal or if you have very large bearings but mine is only 22 mm OD and the PLA is too soft for this to work. the tiny change in OD form the temperature change makes no difference at all.
@@dronefootage2778 This trick only works if you can very precisely control tolerances between fitted parts and those parts have rather similar thermal expansion coefficients. You can't do that with "standard" 3D printing.
@@zkasprzyk that makes sense. the solution i came up with is to print a cap that gets screwed on.
2:54 this is incorrect, slicers do take into account the nozzle diameter. The issue comes down to thermal expansion of the filament, filament getting dragged after its extruded, retraction settings and issues with the machine like E steps, squareness of the frame and axis etc.
I worked at Boeing we used WD-40 to wash stuff out. We then applied Sterrit M-1 to do the job.
With WD -40 the statement was reapply after 40 days.
Never screw your regular ball bearings the way he does it at the 4:12 mark. This will ruin the bearings. Instead, take two appropriately sized washers so you can clamp on the outer race and screw together that way instead. An alternative to this is using either thrust bearings, tapered roller bearings or angular contact bearings, all of which take MUCH higher axial loads, these kinds of bearings will handle this kind of treatment just fine whereas regular ball bearings will not!
Good points! Thanks for the heads-up :)
Good point, but it will only ruin the bearing if the force on the inner race is high. For a light press fit in PLA, maybe it won’t hurt the bearing.
Bearings make the world go round, great video!
Indeed! Thank you :)
In the robotics pieces I've crafted, I have used copper sheeting (easy to get rolls of it cheap) to create a sleeve that wraps around the 5/16" threaded rod.
That takes up the additional space much like your teflon tape method you mentioned, but the teflon tape can continue to deform and a thin copper sheet does that 'less'.
Also, it won't degrade as much if I end up needing to lubricate the bearings later. Some kinds of plumbers tape (or similar) can melt in the presence of lubricant/solvent mixes (like WD40).
I suppose aluminum foil would also work. Great content! Please keep up the good work. :)
If you want them to spin you need thrust bearings not radial bearings. Radial bearings are designed to take a load on the outside race but not a side load. Thrust bearings are designed with preloaded compression in mind like on a lathe.
Not sure how this video was made without this point included.
6:40 I believe this is because you start loading the bearings in the axial, which they aren't made to do, which causes binding.
You are absolutely right! Glad to have so many smart people that follow me 🤙🏻 thanks!
This is the highest density, getting started guide i've seen in awhile. About to place my order for some bearings, thanks for the tips!
2RS = 2 rubber sheild
RS = One side rubbe sheild
ZZ is not mistyped, many mgf use ZZ. One Z is one side sheild ZZ is sheild both sides. But if the shild is stamped ZZ, this is incorrect. One shield shouild only have one Z on it.
Pedantic, I know :-)
No kidding?! This is the first explanation I've seen that made sense, and I've seen a lot. I was told SS just stands for steel seal. Can you tell me this? -- why would one want seals on only one side of the bearing?
@@christophersfactory A one sided sheild is often used in motorbike gearbox for example. Oil flows to the bearing but not right through so oil flows down a oliway.
@@christophersfactory The SS or S prefix is usually stainless steel. it should also be noted that the "R" seals are contact seals. Some seals (eg, designated as V or VV or LLB depending on the manufacturer) are non-contact seals and are used in conditions that are relatively clean and require a high-speed, low friction bearing. There are MANY different types of seals and configurations and designations!
If you use two bearings and crank them together with a bolt the balls on each bearing get compressed into the rings. At a microscopic level the metal has to deform to spin. This is why it becomes harder to spin. There’s a sweet spot of tightness where the shaft runs true and freely. It’s good to use a Belleville washer and two nuts with this configuration to get the best results.
Thanks so much! I'm always amazed at the collective knowledge of a community. Thanks for being a part of that :)
6:41 In short, your problem is due to overconstraint. Underconstraint is the reason why you had the leverage with a single bearing (5:52).
I highly recommend you to read "Exact constraint machine design" by Douglass L. Blanding. Its an amazingly good book, and a true eye opener for anyone that likes to design mechanisms and machines. Dont worry, it doesnt have any math.
Explain?
@@alexwang982 For an object to be fixed, we need to constrain 6 degrees of freedom. Restricting ourselves for now to parallel mechanisms, we can imagine any constraint as a infinitesimally thin beam connected between the fixed world and the object. The question is, how many constraints should we place, and where should we place them.
For a problem in 3 dimensional space, we need exactly 6 constraints that independently restrict a single degree of freedom each. More details on how to place them (or a more detailed/better explanation of the concept) can be found in the book from Blanding, or various places online.
Regardless, you can consider a bearing a constraint of 2. It constrains translational motion in the plane of the bearing. It releases rotation in all degrees of freedom (rotation around the 'main axis', and also around the other 2 axis as shown in 5:52). Adding an additional bearing as in 6:16 fixates the two other rotational degrees of freedom.
Counting our DOFs, this leaves just 2: Rotation around the center axis (which is the one we desire to be left open), and translation along this axis.
Now the problem discussed at 6:27 is the constraint of this final degree of freedom. In this case, you try to constrain it from both sides with a rigid element/connection, the bolt itself and the nut. These two constraints essentially do the same thing of fixing the location of the bolt in space, and are acting along the same translational line to constrain this degree of freedom. This means that either you will end up with play (i.e. the nut is not very tight, shaft can rock back and forth) or excessive stress, wear and other problems (as in the video). In short, in the video the overconstraint of the axial translation by the nut and bolt causes excessive forces onto the bearing. In the video the problem is exacerbated by the fact that radial ball bearings cannot deal with these axial loads very well at all.
I can't believe you only have 120 subscribers, I'm starting my 3d printing journey and your channel is amazing.
This comment just made my day. Thank you, li'l bacon :')
Thanks for that. I've got the skate board bearings, (had them sitting around waiting for inspiration). Now I just need a 3D Printer.
Exciting! For me, it was one of the best purchases I've made in years. The enjoyment I've gotten out of it is completely disproportionate to the cost. Cheers!
Nice video! To prevent that nut from spinning loose, you can use another nut to lock it in place. Set your first nut right where you want it along the threaded shaft (just enough to let the bearings spin freely). Using two wrenches, torque the second nut tightly to the first. That pressure locks them both in place. the 'locking' nut can be removed easily when you want to disassemble it.
The other tip I would add is to add a small chamfer around the edge of the bearing recess to compensate for elephant's foot if it will be on the bottom of the print.
YES! I should have said this in the video. That's a huge one that helps with fitting these bearings in.
HOLY SHEEEET! The bearings in your link are now down to $19.59; the same bearings over the pond, here in the UK, cost £71.27! 😲
I can even get them with 7-day delivery from the US for £34. Yay for Brexsh|t and an incompetent government. 🤦♂
Anyway, thanks for the video; I've just subbed. 👍
Holy cow! That’s quite the discrepancy. I’m not familiar enough with the politics of Brexit but that’s certainly unfortunate. Cheers, friend, thanks for the comment 🙏🏻 glad you’re here
I had to chuckle at the use of MS Paint for a crude "hand drawn" explanation of tool compensation, immediately followed by some footage of precision drawing in Fusion.
As with everyone of your videos, I learned some things. Keep it up!
I forgot to say this in reply to the last comment of yours I saw, but I have to thank you for your positivity and kind words. It's very inspiring to me to keep doing this channel and make more videos. It goes a long way. Thanks again, from the bottom of my heart. Christopher
@@christophersfactory You are very welcome. There's too much negativity on the internet. A little kindness can go a long way.
You mentioned a few days ago that you finally received the electronics you needed to start collecting electricity from your windmill. Will there be a video showing the results coming soon? 🍿
@@schmidtyyt You're awesome. Yes, the parts finally came, but I realized I had a fatal flaw that rendered the PCB's I designed nearly useless. I saved a few by cutting the traces out with a utility knife and re-routing some connections with hook-up wire, but being the perfectionist I am, I didn't want to show that in the video, haha.
On top of that, I redesigned the turbine, but I'm now in the process of redesigning it again. I've used pre-cut plywood up to this point, but I think to have it work really well, I need to get better at learning how to cut and work with thicker and more durable woods. The plywood I've been using warps horribly. I just started working with this stuff called hardboard, I think it's made from compressed sawdust. I can already tell it's much less prone to warping, but it's slightly harder to work with.
I'll probably have to redesign the turbine hub to accommodate the heavier and thicker boards.... this is all to say that there's a lot I'm re-working; I told myself I'd wait to get the "build video" out until I was done, but now I'm thinking I should just make regular, almost vlog-style updates on what I'm doing. Viewers would probably get more benefit out of that anyway, because learning what doesn't work is arguably just as valuable as learning what does.
I have a few other non-windmill-related videos already shot and edited scheduled for this week, but windmill content should start back up next week, especially after these words of encouragement :)
@@christophersfactory That sucks about the PCBs not working as hoped. I know nobody likes to admit failure much less broadcast it to the world, but showing what went wrong may help others avoid the same mistake. As a wise man recently said, "learning what doesn't work is arguably just as valuable as learning what does." 😉
I'd expect the hardboard to work better for you, as well. And I totally get the design re-considerations that brings, too. However, I have to ask: why not 3D print them?
As for working with the hardboard, check out the Dremel Ultra Saw (great Black Friday deal at Amazon today). I found it to be a great little saw for cutting thin wood and metal.
@@schmidtyyt I completely agree about the failure bit. That was one of my goals when I first made this channel, was to candidly show people my efforts in these projects, not just showcase a finished project.
As for why I'm not particularly inclined (although not opposed) to printed blades, there are a few reasons. PLA warps really badly in the sun, especially with large and thin features. If the blades are not thin, it will use a lot of PLA, which is much more expensive than wood. I could try printing in nylon or ABS, but I've never worked with either before.
My short experience with wind turbines has taught me that the name of the game is wind-collecting surface area. This is faster to design with panels of wood, I think. It requires less design consideration. Additionally, I could probably cut 250 wooden blades in any afternoon, but 3D-printing them is going to require a significant amount of time.
So, to answer the question more directly, I guess there isn't really anything that I've found that strictly deters me from 3D printing the blades, but rather several small things that incline me toward wooden blades instead.
People often ask me what the goal of this windmill project is, as I've been working on it for well over two years and through twenty-one designs. Really, there is no goal. I do this because it makes me feel happy and fulfilled. I'm not necessarily trying to create a to-market product, nor am I trying to design something that anyone could assemble with only filament and a printer. So, as the Cheshire Cat said, “If you don't know where you want to go, then it doesn't matter which path you take." I'm just taking whatever path I feel like and feeling things out. Maybe I'll take the path of 3D printed blades at some point.
That Dremel looks pretty slick. I have a knock-off rotary tool, I wonder if it could use Dremel saw blades? Something to look into for sure.
How about you? I think you're probably my most loyal subscriber already, haha. What gravitates you toward this content? Any plans to build a windmill of your own?
A very needed topic as there are thousands of vids on bearing and hardly any on how to mount them.
This just what I am looking for. A few bearings in the mix open up so many possibilities. Thanks for all your diligence. I'd like to show you a few of my designs.
Why not use 8mm stainless steel rods?
The RC Car industry stocks these 8mm rods. Have an RC Hobby shop near you?
A vice is something that everyone should have and is very good for press fitting parts, like those bearings and is probably a bit quicker to set up than the nut and bolt solution.
You're totally right. They're super expensive though 😅
@@christophersfactory my local supermarket (grocery store) sells miscellaneous crap on a weekly rotation so I picked up a couple of cheap vises of various sizes. Mine can grab parts up to about 150mm in the dimension you're squeezing, and as Conor said they're great for 3d printed and maker stuff. That bigger one cost like €15 max
@@cian.horgan No kidding? That's dope. I'll have to check around my area to see if I can find something like that. Otherwise, maybe a garage sale.
@@christophersfactory they really aren’t that expensive and you don’t need large ones that bolt down to workbenches, even a cheap one that just clamps onto a desk is enough for most people.
After nuts and bolts, bearings are one of those things I buy in excess, in sizes I don’t need for ideas I haven’t thought of just because they are so satisfying. Sometimes when I really need some relieve I’ll slide a shaft in and give it a spin. Maybe slide it back and forth if it’s a snug fit. I started collecting used copy machines to take apart because people just give them away and I’ve pulled literally buckets of bearings out. I took apart a plotter and it had a 1m 13mm slide mounted to a sturdy boxed beam. No clue what to do with it but I ordered a set of 13mm sliding blocks just in case. It would make a good X | Y axis on a CNC router.
Great video, and helpful info in the comments. I'm thinking about making some 3D printed projects with wheels, but I had trouble finding beginners advice. This video was the first useful thing that came up.
That was incredibly HELPFUL.
I'm glad I found your channel.
Great video and I'm more intrigued at the wind generator! Fantastic point about logging all the research. I'm probably going to create a form to print out for the things I work on as I normally wing it. Then end up printing the same mistaken part twice, or not really try to push myself beyond the normal limits. I can see documenting my stages and steps is critical.
Usually I'm the "oh that needs X" and adjust it a tad and then start again, but yes I do find myself wondering which version I'm about to use by the third round.
One idea would be to make a cap that fits over the bearing as well that would allow both friction fit an C.A. Glue so the bezel of the cap holds it in place.
6:40 the reason is because those are radial bearings. They're design to be able to take force in the radial direction (think of where they go in a skateboard). However, they are not designed to take forces that are directed along their axis of spin such as thrust forces (Think of an office chair that swirls around with you on top of it). For this, you want thrust bearings (hence the name). These are designed to take this kind of compression force along the axis of spin.
Your printer isn't calibrated right. Slicers take line width into account. They are sophisticated pieces of software. You only need to consider actual tolerance required and the impact different filament and environments conditions has.
In reality, print mechanical parts with a 0.6mm nozzle and similar line width (those are two different things in a slicer)
I find the last step to fit is easily achieved by filing the rough bits off which also ensures that the fit won't get loose over time.
I'm 99.9% sure what explanation about what the slicer doing at 3:05 is wrong. This may have been true when 3d printing first started but modern slicers can certainly account for the thickness of the bead being laid down. That's not the reason your bearing doesn't fit into that hole.
My favourite "take the entire weight of a person" story is bike spokes. How can spokes hold up the entire weight of a person?
The spokes *above* the hub hold you up with *tensile* strength. The load path is you > hub > spokes > top arch of wheel > bottom arch of wheel > ground
Love it. Can't wait to deep dive the rest of your content when I have time. Thank you.
Of course! Glad I could help. Much more exciting content to come :)
I like the flanged 608s as well. They make great timing belt pulleys when stacked inverted.
Some excellent info in here, thanks for posting! I'm about to start my new drybox project and needed some decent bearings.
it was very useful for my final project of my major, you gave me very good ideas and really love yor explanations and the edition
Hi there! Really enjoy your videos from the UK, Could I ask what size washer you use to secure the inner ring of the 608 2rs? Thanks
Instead of brass inserts, have you tried to embed steel nuts in a print? I design in recesses for the nut and screw, including tolerance for nut thickness and in my slicer I put a pause after the recess is printed, before the next layer. I then place the nut in the recess and resume the print. The next layer covers the nut. I do the same with magnets.
Sometimes I'll design a feature that can accommodate a steel nut, although I've never tried embedding them. Doing it during the print process certainly sounds more promising than trying to solder it in. Does that method work pretty well for you?
What I'll often do is just make a square hole that's the height of my nut, and slide my nut in sideways through it until the screw can meet it. Perhaps a strength test video is in order 👀
@@christophersfactory Embedding nuts and magnets works very well for me. When I'm in a hurry I'll use the open square hole method, but when I'm fine tuning production prints I'll embed. I haven't tested strength but it seems to me if the wall thickness between the nut and the head of the screw is the same they should be about equally strong, though I think with an embed it might be a bit stronger, because the nut is pretty much locked into place while the opening for a sliding nut has to be pretty large to accommodate tolerances on the square nuts, resulting in some slop. In any case, embedding the nuts or magnets (when embedding magnets I use a tiny drop of CA adhesive in the recess to make sure they won't jump out of place, attracted to the other magnets) results in a cleaner exterior surface that looks less DIY.
How do you know when is the time to place a nut during print? Do you wait the whole time until it's ready?
@@lovrocatela8727 If you're imbedding a nut, you'll want to wait until the past possible opportunity in the print to do so, otherwise it's likely that the nozzle will bump it. Personally, I tend to use that method less than just making features in the print that nuts can fit into.
@@lovrocatela8727 After initial slicing, I insert a pause at the first level above the cavity for the nut and then reslice. I make sure that the cavity for the nut is deep enough to accommodate the dimensional tolerances of the nuts to make sure there's no chance of the extruder nozzle crashing into the nut.
I also block supports in the cavities for the nut and screw. The layer above the nut is printed as a bridge. While that first layer isn't always perfect, it's almost always good enough to retain the nut and then subsequent layers go down just fine. The pause feature is availabe in the Prusa slicer (and Bambu slicer, too, which is based on the Prusa). I don't have to monitor the printer as the gcode will tell it to pause the print. The Prusa printers will beep to notify you that the print has been paused. I then insert the nut(s) and resume the print.
Chris this video was wildly useful and informative
Glad you think so! Thanks
I need to spin something like your mill, but smaller and much much much faster >20k rpm (200Hz) range. Balanced but not by an expert.
Do you think your method could work? It's almost identical to what drew up
Hats off mate this is an awesome video !! You are underrated and deserve a huge following!
Thank you so much! That means a lot to hear :)
6:42 It's likely that your bearings are not axially concentric. 3D printing and friction fitting will introduce a lot of dimensional inaccuracy causing the 2 bearings to be misaligned. If the nut is screwed down tight, the center of the off-axis bearing will worm its way up/down the bolt as it spins causing it to seize.
A potential fix: Use a long bushing or sleeve to span the gap between the two inner rings. A brass tube or even a 3d printed tube might work. The tube will force the rings to be concentric as the nut is tightened down. Aligning the outer rings will be more challenging though; you might leave the second bearing loose in its housing to help. Alternatively, you could use a whole stack of bearings to support the entire length of the bolt or mount the whole bearing assembly in a pipe.
If you want to use a nut to clamp the two bearing together (without them seizing), try using tapered roller bearings. They can handle axial loading (thrust) and radial loading. Try McMaster-Carr part number: 6677K51
Oh yeah, and basically point the tapers away from each other, then you could clamp down with up to 2,000 lbs!
Woah, those are some badass bearings! Thanks for the heads-up. At $35 a pop, I'll have to save my quarters though lol!
Thanks for sharing this. Your math was very close, and it didn' take much for me to make perfect prints.
I have seen many videos on dimensional accuracy, and you are the first video ive seen that explained it correctly by using microsoft paint! lol. Very well made video!
I doubt this is how slicers actually work. The slicer usually knows which nozzle diameter you use and will acount for that
Thanks! I wondered what to do with my heavily greased bearings.
A crown nut and cotter pin are meant to keep a nut in place without the nut being very tight. It's a bit more complicated to build since the bolt or shaft needs a hole for the cotter pin in the right place but it looks cleaner and is more compact than two nuts jammed together.
WD40 will eventually turn to goo or varnish. In the clock restoration business we hated it. Flushing out the bearings with solvent like naphtha then adding a light oil will likely give better long term service as the cheap grease used in low cost bearings will be gone rather than thinned in WD40 which will mostly evaporate and leave yet more goo behind.
for slightly loose tolerance I wrap some painters tape around the outside of the bearing for a tight fit. For low load applications of course.
That's a good idea! Thanks for sharing :)
Possibly mentioned already, bear with me if this is redundant : Assuming you have room in the build - To secure the threaded end of the bolt and keep it from loosening or being too tight *use two nuts *set the first nut against the bearing / socket *thread the second nut against the first *hold the first in the exact position you want and tighten the second nut against it (tighter the better, and thread lock if you like). The torque between the nuts will keep either from moving along the thread. Hope this makes sense and hope it helps :)
Very cool, thanks for the tips. I tried to design some TPU based wheels for my baby stroller, and had the bearing with bolt acting as lever slop issue. This created quite a sag when a load (the baby) was placed into the stroller. I didn't think of the two bearings as far apart acting as a choke idea, that's great!
Thanks! I'm glad it helped you! If you ever get the stroller design set up, I'd love to see a picture of it, that sounds cool. Cheers!
EXACTLY the video I needed. Thank you!
Knurled nut is a great idea! Nice breakdown thank you
wd40 is not for greasing, it is for cleaning previous junk before lubricating/greasing
Appreciate the time you take to share your wisdom!!
I have trouble finding bearings on Ali, which are not factory rejects being passed off as the good stuff.
First video I’ve seen of yours very interesting thanks
they are races not rings. the reason they lock up when you tighten the bolt is because they are coaxial bearings not trust bearings. the balls bunch up against the side of the race and causes misalignment in the races and this multiplies the issue and locks up.
Wow, I feel like I learn much more people than I have to share. I'll definitely need to make a Part II. Thanks for the info!!
I find 0.15mm added to the OD of the bearing is good for press fit.
And it's not good to force bearings into place by the inner race, unless you're pressing it onto a shaft. You can damage the bearing. I made a 3d printed tool to press in, and remove, bearings in pulleys. It has a tube that only rests on the outer race and uses a bolt that's the same diameter as the bearing's ID.
Thank you! You saved me a lot of wasted money because I thought I needed to buy expensive bearings since the cheap ones didn't spin easily.
CK3D has a free 3D printable needle bearing that uses filament as the needles. It's free on their patron page.
Great video. Informative and great pacing. Do a followup about spacers maybe (what you to keep them from locking up). Having disassembled a ton of printers I learned they are used constantly, and so old printers are a good source of pretty decent metallic spacers (although printed spacers do the job too).
Thank you so much! I very much agree, and I probably will make a part 2.
Mahalo , just got my 100 pack bearing from Amazon.. arrived in 3 days.. Well also try your measurements for the 3d printers to fit them into my projects.
Why don't you heat plastic if hole is too small and then insert bearing while plastic is soft?
That's totally a doable method, it can be a little difficult to ensure that the plastic only deforms where you want it to, and I would wager that that bond is still not as strong as just using superglue.
No freeze the bearing so that it constricts place the bearing and as it expands it’ll fit very tight. It’s how we change wheel bearing on cars.
Have you thought of using thruster bearings
GREAT VIDEO ! Very informative, thank you
underrated, great video
This was so helpful. Now subbed and liked! Thank you!
thanks. best video on the subject on youtube.
Subscribed because of that last statement. LOL, who knew??? IKR?
"Assumedly"... Like!
Using regular metal gear and bearing for base 3d printing is a good idea.
ผมเห็นด้วยครับ ขอบคุณที่มาคอมเม้นท์ครับ :)
I want the files for your windmill. My niece would love that windmill!
The part about 3D printers just doing "whatever happens happens" is (fortunately) not true ... unless your printer isn't properly calibrated (which unfortunately happens often!). What *is* true is that usually we're talking about printers that are not industrial-grade - they're consumer-grade! But yes, slicers *do* take this kind of problem into account (even if imperfectly). 🙂
Another method instead of glue could be to pause the print at the right layer number, put in the bearing, and then continue the print, printing also covering the other side just enough.
I'd be interested to try this! I haven't done much with inserting parts mid-print. I know many people do, to fair success. I just worry about the nozzle colliding with the part. Do you know of any way to avoid that?
@@christophersfactory You'll have to stop the print at the right height. Check in your slicer. You can do that with nuts too.
In cura there is a setting I believe it is called slicing tolerance (not really sure off hand) but the choices are inclusive and exclusive. This will let you tell cura if it needs to be at least or at most and it will put all .4 on the side of the line it needs to be.
That setting is for the vertical axis, not where on what side of a trace it should extrude.
hello, great video on a useful topic, thanks for sharing your skills :) 👍
When u use two bearings one should be fixed in place while other allowed to slide on on at least one of the parts. There was long explaination on my uni why it was, but i remmember analysis of forces on that was nightmare (like literally impossible to do static analysis on that) and just go for one of them sliding
Thanks for the tips!
You're more than welcome. Glad they proved useful :)
WD 40 is not a good oil, you want a light machine oil like 3-in-1
You're right. I only suggest WD 40 because it's something that most people have around the house. :)
I think the reason people don't like using wd 40 is it's acidic and may cause issue with certain coatings. made from fish oil apparently. it's great to get rid of water.
Which is what it's designed to do. It's full name is Water Displacement formula #40. It's not a lubricant.
There's another way to solve the problem of a model that has a hole that is too big or too small for the bearing: change the horizontal expansion setting in the print settings. Usually -0.1mm to +0.1mm does it for me.
Good video, thanks for the info.
You're welcome! Glad you enjoyed it.
Great channel, new subscriber. There is a setting in Cura slicer for horizontal tolerance
Interesting -- where is that located in Cura? I've seen several people talk about it but the only one I've ever seen is the vertical tolerance in the experimental settings.
Thanks for the comment and kind words :)
@@christophersfactorySlicing Tolerance in experimental settings, lost in tech has video about printing threads in Cura covers it
Ball bearings aren't designed to be constantly side loaded. Needle bearings are designed for that. They are pretensioned to eliminate all freeplay.
If you learn best with your hands, take a look at how a bicycle wheel bearing stack is set up to better understand how the stress physics of a windmill wheel could be minimized. Bicycle wheels have been on the engineering block for a long time and optimized to gain an edge to win world championships.
great video, thanks!
Great video!
Thanks! :)
Dont recommend using set screws to secure bearings you will deform the outer race and make it oval. It's better to use a clamping style collar or press fit.