Nice work Manuel! By the way, there is a relatively easy way to make glass or even zerodur air bearings. Instead of using diamond milling the surface channels, you can use sandblasting with SiC. Using 3d printing resin, which is relatively blast resistant, you can create the pattern and then blast ths glass surface to depths ranging from a few to a few hundred microns deep.
@@DiffractionLimited Fuck youtube comments everything gets filtered :( If you want you can adjust the filtering in your setting to be not so strict. But IDK if you have bot problems. (UA-cam Studio -> Settings -> Community -> Defaults) Here you can set the strictness (for new videos, old one have to be adjusted manual)
@huygensOptics - You wouldn't have very precise control of the depth using that method, correct? Can I infer that the depth consistency of the channels is not super critical?
@@willusher3297 It depends, since the presented method doesnt provide control either, the goal is to make them sufficiently deep so that variations dont make much of a difference and rely on the restrictor. In industry more advanced designs do use precise channel depth to control the properties of the bearing.
Thank you for the video. Its super nice to watch and intresting 😊 well done. A little safty tip: dont use gloves while working with rotating tools like a drill press or lathe. It might not be that dangerous with small diameter drill bits because they break easily. But worst case the glove can get catched and will rip the glove, including your finger, off. It works like the finger trap kids toy. A coworker lost his finger this way.
Nice. I saw a complimentary diy to make rotational air bearings. It involved using an extra metal axel part that was covered in epoxy and abrasive which was then used to grind a depression in a graphite block. Then a hole was drilled in the graphite for the air supply. Thinking of making one for a turntable, using an old CPAP as the pressure source. I plan to use a linear induction motor to turn it, but I have not figured out how to suspend the tone arm yet, though I am leaning towards using cord rather than bearings...
Very nice. Maybe one could avoid super gluing metal disc blank to molding 3 point jig by adding to that jig a magnet? Disks are not that heavy, glass reference surface is diamagnetic and demolding would be cleaner and bit easier
I learned a lot from this video. I was planning to use same epoxy layer application, but was not so sure about if it will work or not. Thanks to you, now I know that it works ! Nice ending music btw :)
You just made UA-cam a better place. Really nice job and very well executed and explained. Just one question tough: how much is the pressure loss and how powerfull has to be the compressor you use for the air to push into this air bearing? Thank you in advance.
Thank you. I use two tiny airpumps in series, which produce about 2 bar. The amount of air consumed by the pads is really low (The air within the small 1m hose is enough to keep the bearings flying for about 20s after turning off the compressor.) You can even use canned air for testing.
This video feels like getting smarter. The plaster is a nice trick and a lot easier than drilling a tiny hole. I have my doubts about a plastic drill guide. If there was a steel bushing, it would be more n00b-proof. But that's a nit compared to everything else you've shared. Can you recommend a book?
I'd love to see you do an air spindle like this. Would have to be small to find a good reference surface as good as float glass. They make ER collet to shaft converters which is maybe what I'd try
Yes its a homemade drill press and its unfortunately not finished yet. But I can do a short video about it in future. It uses a JMC Servomotor as spindle allowing for very silent operation with position, torque and velocity control (e.g. for tapping). It has a small esp32 microcontroller and an Oled display for showing RPM/Torque. Its really quite neat!
Very nice work and explanation of your process. It look like is can support a good amount of force. Do you happen to have any recommendations for literature or books treating the engineering design of air bearings?
Wow this is really really really old tech a lot of height stands and measuring equipment in the manufacturing industry have been using this for decades
It was amazing! As an amateur, I learned a lot from this project. But there are some doubts for me, please clarify them! What is the pressure of the air compressor? And why do you use plaster in the center hole? I would have guessed that you might drill the hole with a precision drill, but you didn't. I guess it prevents pressure drop.
A precision hole works too but might be more difficult to manufacture, thats why I choose plaster as a porous media restrictor. The restrictor is very important and leads to a sudden pressure drop under the pad when the gap hight increases. This results in a stiff bearing force. The pressure for the demo was about 2 bar.
@@DiffractionLimited It's amazing how even a slight pressure difference over a big area can produce an ungodly amount of force. E.g. Vacuum work holding with a simple MDF sheet and a household vacuum cleaner is often more than enough for light milling operations. The MDF acts as restrictor and due to its fibrous nature can in some cases even grip into the surface of the work piece. Resulting in a rocksolid clamp.
@@DiffractionLimitedWhat restrictor would you use for a hydrostatic system? Oil instead of air, you get so much more stiffness. Personally I was thinking about using needles as the restrictors.
@@LeoNordin-c8b Maybe try the Ferrules from optical fibers (i have a video on how to extract them) ? They have a precision hole of about 125µm in a very hard ceramic body that could easily be glued into a drilled hole.
Why do you use the epoxy casting vs polishing the metal disc directly (e.g. using a high-grit diamond stone and/or polishing compound)? Is it ease of manufacture, or is there an advantage the resin provides over the metal? Excellent work, thank you!
Its ease of manufacture. The overall flatness is much more impoartant thatn the actual surface finish. You can use any mothe you like, as long as you keep the surface very flat and free of burs.
Hi, did you test other mold release agents as well? I have been sourcing for precision options and I added this one to my list. But have not choosen yet. Would adding pva as recommended make the demoulding easier?
I'm curious what you'd estimate is the weight bearing capability of these. I've been rearranging my garage lately and this has me dreaming of making some of these to use as feet for the heavier items. If they can handle the load and work on the concrete floor, then I could imagine these would make moving things around so much easier.
Nice project, well executed. Is there any special reason for not just going with glass and laser engraving, which I believe would be even simpler and just as effective? Or acrylic glass, the simplest to work with and offering all the machining possibilities, precise engraving depths and geometry etc.? I also noticed that graphite wasn't mentioned, although at least in my opinion, it might work well if used as an insert, for example with acrylic shell, eliminating the need for engraving because of its porousness and good air distribution. I'm not criticizing anything, just trying to understand the reasoning behind the material and method choice.
>>Is there any special reason for not just going with glass and laser engraving: I tried it once with a diode laser and painted glass and I ran into troubles because tiny melted glass 'blobs' protruded the surface. So without additional polyishing/grinding it seemed unfeasable. I dont know if it works better with a CO2 or Fiber laser since I dont have access to those, but it would be very interesting to try since from a manufactioring perspective it would be great. >> Acrylic glass Tried that as well but it was way to flexible to hold its planarity under load. In my experience it just didnt work very well. >>I also noticed that graphite wasn't mentioned I never tried graphite bearings but I couldnt think of a way to make them as small as 'solid' pads. Also graphite isnt a very strong material so it needs to be supported very well by the 'host structure' making it more bulky. Also stiffness of the air bearing film tends to be lower than with solid pads. >> I'm not criticizing anything, just trying to understand the reasoning behind the material and method choice. I think those are very good and valid questions and I am happy to answer this kind of requests. Also feel free to critice ;-)
That's a pretty cool project. I like it. I followed and understood everything you did but the one thing I don't understand and please help me understand how does it actually work. You plugged the nozzle at the end with plaster so how does the air get out? Is the plaster porous? Is that how the air is getting out? Because in my mind you just plugged up the air hole. I don't understand how the air is getting out
You are right, the plaster is porous and you can control the porousity in a certain range by changing the ratio of plaster and water (here I used 1:1 by mass - quite a lot of water). The plaster has a very important task! It restricts the air flow. If it was unrestricted the pad would be hoovering way to high above the surface and you loose stiffness (it would bee like a soft air cusshion). Also the pressure in the feedline would drop for all other bearings if one of them was lifted up. Maybe take a look at the 'How air bearing work' section again and try to imagine what would happen if an infinite amount of air could flow to the pad.
@DiffractionLimited oh ok I get that. Okay so the plaster is porous and it restricts the air flow but it restricts all three of them evenly to a measured amount and that measuring amount is adjustable based on how much water you use the ratio and that can change how much air flow you have and I suppose that can affect how much weight can be lifted and that's something you can probably experiment with
Hi, thanks for the great video - really useful. I want (at some time!) to make a cylindrical air bearing, or air bushing. It's topologically identical to your bearing, but harder I fear to cut the channels in the cylinder interior. Before I start thinking too hard about this, I wondered whether you'd ever tried using the plaster as the bearing rather than the epoxy? As it's porous it should act like a commercial bearing, which are all made that way?
I tried but it doesn work well, I deforms when curing and is to soft. Also whenever there is a particle under the bearing it tends to create more plaster particles thus quickly destroying the bearing surface.
@@DiffractionLimited Hi - thanks so much for getting back to me. I suspected that might be the case, although hoped that perhaps the air pressure might keep it all together. In the meanwhile I've (re)discovered Applied Science's video - ua-cam.com/video/K_N_h_mKf-4/v-deo.html - where he uses graphite, and I might try that instead. I'm really keen on using a porous bearing if I can...
As you have seen the air gap is very small. Plastics are usually not strong/stiff enough to hold their shape under load very well. Remember the force is inserted at the center where the ball bearing is. With larger bearing s this becomes more and more of a problem. You also have to make the print air-tigh, which can be done but is also some work. Plastics also deform alot under temperature variations. Overall I believe metal ist just the better choice here.
@@DiffractionLimitedthat would depend entirely on the desired end result. They sold a lot of air hockey tables with less precision than this. Can the air be turned on and off, so I can set the load down after moving it?
21:25 "Take care to create even channels that do not intersect with the circular channel" Did I hear that correctly? The straight channels should not connect to the circular channel?
Looking up the word "intersect" gives "divide by passing or laying across it." So, I think it is correct. You connect the circle, but not pass across it
Wow! Can you use gauge indicator and measure Can you use a gauge indicator to measure how much the indicator arrow moves if you press it with your finger from above?
@@timbehrens9678 The plaster filling acts as a restrictor and is extremely important. The bearings will not work without them. You only want a tiny amount of air entering the gap so that the pressure collapses as soon as the gap gets slightly larger. This produces a 'stiff' bearing force.
I used about 2 bar (slightly on the low side for the restrictors used) and yes the pressure definitely influences the load capacity and the optimal operation point. More pressure, higher loads - but you alos need more preload force to get the most out of them.
Do you know of any resources that a person could use to be able to calculate the specs of a bearing for a given purpose? I'd love to be able to use them for my own project, but I'm not sure how to figure out the specs that I'd need.
I think for now you would have to experiment on your own to get the specs. There will also be an open source simulator program in future, but it will take a while until it is ready for release.
Damm many youtubers I watch since before seem to comment here. So much I have been missing out on. I wonder if the demoulding is easier on a flatter surface, if it is even possible over a large diameter like lets say 30cm.
I can only make an educated guess, since I havent measured. But at 2 bar input pressure I would put them between 2 and 6 N/µm for a single bearing pad, when used with optimal preloading.
The platform is like in ASML photolithographic machines! Now it remains to make the same displacement drives and firmware on raspberry PI and can be sold to countries that cannot buy these machines due to US sanctions! Subscribe!
Nice work Manuel! By the way, there is a relatively easy way to make glass or even zerodur air bearings. Instead of using diamond milling the surface channels, you can use sandblasting with SiC. Using 3d printing resin, which is relatively blast resistant, you can create the pattern and then blast ths glass surface to depths ranging from a few to a few hundred microns deep.
Thank you. Also great tip! If I ever get around building a sandblasting chamber I will try that out.
@@DiffractionLimited Fuck youtube comments everything gets filtered :(
If you want you can adjust the filtering in your setting to be not so strict. But IDK if you have bot problems.
(UA-cam Studio -> Settings -> Community -> Defaults) Here you can set the strictness (for new videos, old one have to be adjusted manual)
Why not just use etching for the glass?
@huygensOptics - You wouldn't have very precise control of the depth using that method, correct? Can I infer that the depth consistency of the channels is not super critical?
@@willusher3297 It depends, since the presented method doesnt provide control either, the goal is to make them sufficiently deep so that variations dont make much of a difference and rely on the restrictor. In industry more advanced designs do use precise channel depth to control the properties of the bearing.
Thank you for actually narrating your video!! It is greatly appreciated, as opposed to just using captions.
Superb work my friend. Thank you for making the high quality educational videos that I never did
Thank you :)
Thank you for the video. Its super nice to watch and intresting 😊 well done.
A little safty tip: dont use gloves while working with rotating tools like a drill press or lathe. It might not be that dangerous with small diameter drill bits because they break easily. But worst case the glove can get catched and will rip the glove, including your finger, off. It works like the finger trap kids toy. A coworker lost his finger this way.
Mechanical poetry with a top class video showcasing the design and build process. WoW
fantastically accessible: i especially appreciate the 3d printed tooling. will definitely have to try this out
Thanks, let me know how it worked for you. Also note that the CAD files are now online.
The do’s and do not’s are incradabily pragmatic,well done esquire. Thanks for sharing
Translucent material allows UV quick glue. I love the no smell/mixing/hassle of it.
Telling people to skip the section they don’t like is such a power move on youtube 😂
Nice. I saw a complimentary diy to make rotational air bearings. It involved using an extra metal axel part that was covered in epoxy and abrasive which was then used to grind a depression in a graphite block. Then a hole was drilled in the graphite for the air supply. Thinking of making one for a turntable, using an old CPAP as the pressure source. I plan to use a linear induction motor to turn it, but I have not figured out how to suspend the tone arm yet, though I am leaning towards using cord rather than bearings...
What a lovely set of bearings.
Very nice work and excellent result!
Very nice. Maybe one could avoid super gluing metal disc blank to molding 3 point jig by adding to that jig a magnet? Disks are not that heavy, glass reference surface is diamagnetic and demolding would be cleaner and bit easier
I learned a lot from this video. I was planning to use same epoxy layer application, but was not so sure about if it will work or not. Thanks to you, now I know that it works ! Nice ending music btw :)
You just made UA-cam a better place. Really nice job and very well executed and explained. Just one question tough: how much is the pressure loss and how powerfull has to be the compressor you use for the air to push into this air bearing? Thank you in advance.
Thank you. I use two tiny airpumps in series, which produce about 2 bar. The amount of air consumed by the pads is really low (The air within the small 1m hose is enough to keep the bearings flying for about 20s after turning off the compressor.) You can even use canned air for testing.
First introduction clear and informative. Thank you. Noe let see what you make.
nice job
This video feels like getting smarter. The plaster is a nice trick and a lot easier than drilling a tiny hole. I have my doubts about a plastic drill guide. If there was a steel bushing, it would be more n00b-proof. But that's a nit compared to everything else you've shared. Can you recommend a book?
The idea ro use bushing is very good. 3D print worked for me since I used a dill press as well. Thanks for the idea
Nice. Now you need the right line/scratch pattern on the pads, so your able to steer the setup by rotating the disks/pads.
That's a cute project! Nice
This is great! Thank you for sharing.
Excellent work
I'd love to see you do an air spindle like this. Would have to be small to find a good reference surface as good as float glass. They make ER collet to shaft converters which is maybe what I'd try
How would you make the air channels accurately spaced?
Can you do a video on that what looks to be homemade drill press? Please.
Yes its a homemade drill press and its unfortunately not finished yet. But I can do a short video about it in future. It uses a JMC Servomotor as spindle allowing for very silent operation with position, torque and velocity control (e.g. for tapping). It has a small esp32 microcontroller and an Oled display for showing RPM/Torque. Its really quite neat!
@@DiffractionLimited omg video please. I’d like to build something similar.
Thanks for sharing!
I think i missed how the air escapes 😅 but damn impressive build!
Very nice work and explanation of your process. It look like is can support a good amount of force. Do you happen to have any recommendations for literature or books treating the engineering design of air bearings?
Good job! what pressure is needed to operate and what is the air consumption.
Wow this is really really really old tech a lot of height stands and measuring equipment in the manufacturing industry have been using this for decades
It was amazing! As an amateur, I learned a lot from this project.
But there are some doubts for me, please clarify them!
What is the pressure of the air compressor?
And why do you use plaster in the center hole? I would have guessed that you might drill the hole with a precision drill, but you didn't.
I guess it prevents pressure drop.
A precision hole works too but might be more difficult to manufacture, thats why I choose plaster as a porous media restrictor. The restrictor is very important and leads to a sudden pressure drop under the pad when the gap hight increases. This results in a stiff bearing force. The pressure for the demo was about 2 bar.
@@DiffractionLimited It's amazing how even a slight pressure difference over a big area can produce an ungodly amount of force.
E.g. Vacuum work holding with a simple MDF sheet and a household vacuum cleaner is often more than enough for light milling operations.
The MDF acts as restrictor and due to its fibrous nature can in some cases even grip into the surface of the work piece.
Resulting in a rocksolid clamp.
@@DiffractionLimitedWhat restrictor would you use for a hydrostatic system? Oil instead of air, you get so much more stiffness.
Personally I was thinking about using needles as the restrictors.
@@LeoNordin-c8b Maybe try the Ferrules from optical fibers (i have a video on how to extract them) ? They have a precision hole of about 125µm in a very hard ceramic body that could easily be glued into a drilled hole.
Nice. Next up JB weld air bearing ball joint using a ball bearing? Or making a preloaded air bearing with two sones for vacuum and air?
Wow nice
Why do you use the epoxy casting vs polishing the metal disc directly (e.g. using a high-grit diamond stone and/or polishing compound)? Is it ease of manufacture, or is there an advantage the resin provides over the metal?
Excellent work, thank you!
Its ease of manufacture. The overall flatness is much more impoartant thatn the actual surface finish. You can use any mothe you like, as long as you keep the surface very flat and free of burs.
Hi, did you test other mold release agents as well? I have been sourcing for precision options and I added this one to my list. But have not choosen yet.
Would adding pva as recommended make the demoulding easier?
I'm curious what you'd estimate is the weight bearing capability of these.
I've been rearranging my garage lately and this has me dreaming of making some of these to use as feet for the heavier items.
If they can handle the load and work on the concrete floor, then I could imagine these would make moving things around so much easier.
Nice project, well executed. Is there any special reason for not just going with glass and laser engraving, which I believe would be even simpler and just as effective? Or acrylic glass, the simplest to work with and offering all the machining possibilities, precise engraving depths and geometry etc.? I also noticed that graphite wasn't mentioned, although at least in my opinion, it might work well if used as an insert, for example with acrylic shell, eliminating the need for engraving because of its porousness and good air distribution. I'm not criticizing anything, just trying to understand the reasoning behind the material and method choice.
>>Is there any special reason for not just going with glass and laser engraving: I tried it once with a diode laser and painted glass and I ran into troubles because tiny melted glass 'blobs' protruded the surface. So without additional polyishing/grinding it seemed unfeasable. I dont know if it works better with a CO2 or Fiber laser since I dont have access to those, but it would be very interesting to try since from a manufactioring perspective it would be great.
>> Acrylic glass
Tried that as well but it was way to flexible to hold its planarity under load. In my experience it just didnt work very well.
>>I also noticed that graphite wasn't mentioned
I never tried graphite bearings but I couldnt think of a way to make them as small as 'solid' pads. Also graphite isnt a very strong material so it needs to be supported very well by the 'host structure' making it more bulky. Also stiffness of the air bearing film tends to be lower than with solid pads.
>> I'm not criticizing anything, just trying to understand the reasoning behind the material and method choice.
I think those are very good and valid questions and I am happy to answer this kind of requests. Also feel free to critice ;-)
That's a pretty cool project. I like it. I followed and understood everything you did but the one thing I don't understand and please help me understand how does it actually work. You plugged the nozzle at the end with plaster so how does the air get out? Is the plaster porous? Is that how the air is getting out? Because in my mind you just plugged up the air hole. I don't understand how the air is getting out
You are right, the plaster is porous and you can control the porousity in a certain range by changing the ratio of plaster and water (here I used 1:1 by mass - quite a lot of water). The plaster has a very important task! It restricts the air flow. If it was unrestricted the pad would be hoovering way to high above the surface and you loose stiffness (it would bee like a soft air cusshion). Also the pressure in the feedline would drop for all other bearings if one of them was lifted up. Maybe take a look at the 'How air bearing work' section again and try to imagine what would happen if an infinite amount of air could flow to the pad.
@DiffractionLimited oh ok I get that. Okay so the plaster is porous and it restricts the air flow but it restricts all three of them evenly to a measured amount and that measuring amount is adjustable based on how much water you use the ratio and that can change how much air flow you have and I suppose that can affect how much weight can be lifted and that's something you can probably experiment with
Hi, thanks for the great video - really useful. I want (at some time!) to make a cylindrical air bearing, or air bushing. It's topologically identical to your bearing, but harder I fear to cut the channels in the cylinder interior. Before I start thinking too hard about this, I wondered whether you'd ever tried using the plaster as the bearing rather than the epoxy? As it's porous it should act like a commercial bearing, which are all made that way?
I tried but it doesn work well, I deforms when curing and is to soft. Also whenever there is a particle under the bearing it tends to create more plaster particles thus quickly destroying the bearing surface.
@@DiffractionLimited Hi - thanks so much for getting back to me. I suspected that might be the case, although hoped that perhaps the air pressure might keep it all together. In the meanwhile I've (re)discovered Applied Science's video - ua-cam.com/video/K_N_h_mKf-4/v-deo.html - where he uses graphite, and I might try that instead. I'm really keen on using a porous bearing if I can...
I was thinking that a syringe with about an 18g needle would allow you to fill the plaster from the bottom, eliminating any chance of air bubbles.
Awesome, I want them I just have no idea what to do with them
You will find out for what to use them when you have them ;-)
Is the use of JB Weld purely for ease of manufacture? Or do you think it is better than a scraped or lapped steel/CI surface?
Why instead of the trouble of drilling the metal disc etc dont make the disc and channel with 3d print?
As you have seen the air gap is very small. Plastics are usually not strong/stiff enough to hold their shape under load very well. Remember the force is inserted at the center where the ball bearing is. With larger bearing s this becomes more and more of a problem. You also have to make the print air-tigh, which can be done but is also some work. Plastics also deform alot under temperature variations. Overall I believe metal ist just the better choice here.
@@DiffractionLimitedthat would depend entirely on the desired end result. They sold a lot of air hockey tables with less precision than this.
Can the air be turned on and off, so I can set the load down after moving it?
21:25 "Take care to create even channels that do not intersect with the circular channel" Did I hear that correctly? The straight channels should not connect to the circular channel?
Looking up the word "intersect" gives "divide by passing or laying across it." So, I think it is correct. You connect the circle, but not pass across it
Is the plaster of paris porus enough for the air to reach the scribed channel?
Wow!
Can you use gauge indicator and measure
Can you use a gauge indicator to measure how much the indicator arrow moves if you press it with your finger from above?
Yes, I will do stiffness measurements eventually.
What is the role of the plaster filling? Can those bearings work without it?
@@timbehrens9678 The plaster filling acts as a restrictor and is extremely important. The bearings will not work without them. You only want a tiny amount of air entering the gap so that the pressure collapses as soon as the gap gets slightly larger. This produces a 'stiff' bearing force.
how much pressure did you use on those? and is it correlated with the weight that it hold?
I used about 2 bar (slightly on the low side for the restrictors used) and yes the pressure definitely influences the load capacity and the optimal operation point. More pressure, higher loads - but you alos need more preload force to get the most out of them.
Do you know of any resources that a person could use to be able to calculate the specs of a bearing for a given purpose?
I'd love to be able to use them for my own project, but I'm not sure how to figure out the specs that I'd need.
I think for now you would have to experiment on your own to get the specs. There will also be an open source simulator program in future, but it will take a while until it is ready for release.
Air squeezes though plaster in the center, not crazy bout that it;'s abrasive and variable.....
Easy to drill out plaster but use hot glue etc.
How did you drill the small central hole?
Using a 0.1mm pcb drill. But it was just a test and is not needed for building the pads.
This reminded me of Minecraft Elegance, it is mechanical elegance. Good work
Damm many youtubers I watch since before seem to comment here. So much I have been missing out on.
I wonder if the demoulding is easier on a flatter surface, if it is even possible over a large diameter like lets say 30cm.
How stiff are these bearings?
I can only make an educated guess, since I havent measured. But at 2 bar input pressure I would put them between 2 and 6 N/µm for a single bearing pad, when used with optimal preloading.
Air hockey anyone?
I hate when people are doing something for the first time and explain the process like they are an absolute pro and mastered it to perfection
Why is every surface that you rub together in this video make a horrifically awful noise
The platform is like in ASML photolithographic machines! Now it remains to make the same displacement drives and firmware on raspberry PI and can be sold to countries that cannot buy these machines due to US sanctions!
Subscribe!