Air Bearing Spindle Diamond Turning CNC Lathe: Complete Design Review and Introduction
ฝัง
- เผยแพร่เมื่อ 28 ก.ย. 2024
- For some reason the framerate on the in person video got axed and I can't seem to fix it, sorry. Anyway this is what the air bearings are for and what I'll be working on for the next while. Ultra precision to a micron is the goal
Beautiful concept! I know you changed this quite a bit since making the video, and this comment is no longer valid for this design, but maybe you can use it for a future design: Incase you were not aware, McMaster sells ultra fine thread screws, and press fit brass threaded inserts. They don't go to 1/2" diameter, but do to 1/4-80 and 1/4-100. PN is 98625A200 for the screw and 98625A586 for the insert. (multiple lengths of both are offered). The screws also have a ball at the end which could potentially interface with the pivoting jaw directly. I was going to use these to build a fine adjust into an indicator holder. (never got around to it).
Did you have a look at Dan Gelbert's air bearing lathe with granite surface plates? I like the way he uses them to achieve his micron precision cross slide action.
At the scale of this lathe, air bearing ways wouldn't be quite stiff enough for me. Dan is a huge inspiration for this project if it wasn't already obvious, his lathe is awesome.
You can use differential threads for fine adjustment. Basically a hollow bolt with a different thread pitch on the inside bore, then screw in that bore. Say if you have a 1/2-20 main screw. That screw is drilled down the middle and tapped #10-24. One full turn of the outer bolt advances the screw 0.050" like normal. But if you hold the inner screw still, then that screw will retreat 0.042" and give you a nice fine 0.008" per turn advance at the end of that inner screw. Locking that inner screw without disturbing it might be difficult though. Maybe if you made your main bolt from scratch with a large thumbwheel you could have a perpendicular setscrew, and have a lead or copper slug to bear against the thread of the inner screw to keep it backing out. Or if you can find a small screw that's long enough, a locknut against the head of the outer screw will work fine.
(Well crap. I wrote this while watching and didn't know others had pointed it out.)
I get more more comments telling me to use differential threads on my channel than any other kind of comment lol. I was familiar with them going into this design and chose to not use them for reasons I’ve explained in other comments
For your Radial Adjustments, I would highly recommend making several Differential Threads...
that way you can achieve almost micron adjustment very accurately...additionally, you can
then use conventional thread sizes and Not have to make a Tap...I have made several that
achieve nearly the equivalent of 200 TPI
On machining the Graphine bearing final shape a Jig Grinder with a diamond tool would work
nicely...
Following your build, as I am also designing an Air Bearing lathe...having just purchased a
Brown & Sharpe #1 Universal Cylinder Grinder and next a Moore #3 Jig Grinder...Great work!!!
As I mentioned in another comment, differential screws would be a little difficult to implement into this particular design, and id like to keep it as rigid as possible. You are certainly right in that a Moore jig bore would make easy work of the bearing radii, but the shop I work in doesn’t have one and I’m not made of money so I’ll have to stick to standard boring heads for now.
thanks for the vid and the time you took to explain things. My real question is this. What is it that you plan on making that will actually take advantage of this precision? Do you have a list of interesting applications that you can highlight to illustrate the need for such precision? thanks again for the vids, great work
That is a good question I never addressed. I should mention, the design is obviously completely different now than it was in this video. I’ll make a new video once the design is done in November. Diamond turning lathes are the precision behind a lot of stuff, namely cell phone camera lenses and many many other optics, as well as applications in automotive, medical, physics, and a bunch of industries. Personally, I am interested in using the technology to produce telescope Mirrors that require 0 post processing and are sufficiently accurate and reflective directly off the machine.
@@cylosgarage that is what I want to do as well. Astrophotography is a hobby of mine. I have a Software Bisque paramount mount on a Takahashi (other way around actually)
Very cool , we should keep in touch
very nice work so far. Air bearings have also been on my todo list, but still far away.
Where do you get the graphite material from?
Most generic graphite off of amazon seems to work great. Just search graphite block or graphite rod or something and pick one that says its for EDM electrodes. This ensures you don't get any impure clay bonded graphite, as only the pure stuff works. Thanks!
One thing to enhance the cutting actions of the "teeth" cut in the bearing, Considering the shape of it and the fixed rotation axis, you cut them at the wrong angle, you need to turn them let's say 45 degrees. In your current setup, one set of lines do not cut at all, the others cut at the same time the whole length of the "tooth". That's not good. It cuts way too abruptly and it gives you worse looking scratches in the graphite. Look at end mills for metal - look at the helix and the angle they have. Basic routing end-mills for wood are just straight pieces of carbide brazed without any helix to a steel base - reasons are reduced price, and they do not tear out the fibers of wood up or down.
bruh what lmao
@@cylosgarage what what? 26:45 - you're showing the cylindrical bearing you say you used for lapping the cylindrical shape into graphite and the lines you cut are perpendicular and parallel to any kind of motion you would use during "lapping" the graphite to shape. I just offered you a suggestions of of how such lines (teeth of cutting action) should be angled for better performance. Also look at ben Krasnow's video, he has the teeth at not 90 degree also. Look at this vid th-cam.com/video/Wojlc3VsuHQ/w-d-xo.html . The lines are 45 degree to the motion of the durin the lapping of prismatic ways. Same was done on the prismatic ways on Moore Jig Grinders as per book "Foundations of Mechanical accuracy" by Wayne.R.Moore
@@chronokoks oh i thought you were talking about something completely different. I see now. regardless if the lapping had worked well or not, the fact remains that even shafts from the same batch have some variance in diameter; the one i bought for lapping was 6 tenths smaller than the full shaft for the spindle, for example. This difference in diameters is fine for demos like ben's, but for best performance they need to be matched much better. Therefore I've abandoned lapping for boring. Thanks for the advice though, you're totally right.
why are you using graphite? yeah I have seen the videos of everyone using graphite for air bearings but that's good for youtube videos. For production machine you would like to use some strong plastic material, or metal. Graphite could deteriorate over time and the spindle wouldn't be durable. If your spindle is going to be held vertically, then putting traditional bearings (made with metal balls) at the bottom would prevent the spindle from being damaged on pressure loss during movement. And even if the spindle would have to be held horizontally I would avoid graphite and look for other solutions
“One of the great challenges in life is knowing enough to think you're right but not enough to know you're wrong”
-Neil deGrasse Tyson
@@cylosgarage you forgot to add the statement about nonexistence of an absolute truth
You should look into differential screws
I'm actually using differential screws in an indicator base I'm building right now. I would have liked to have used them in this but it would have been complicated to implement and sacrificed a little rigidity.
@@cylosgarage Have you shared any information about that online? If so can you link me to it?
@@evren.builds unfortunately no I haven’t, perhaps I should. If you look at Tom Lipton’s video on precision mechanics though, it’s basically the same thing as what he did in that video to demonstrate differential screws
@@cylosgarage Thanks!, will do
Epic. This design is really clever. Have you thought about thermal expansion, could this be a problem? I mean you have an constant stream of air "cooling" the spindle. Maybe a state of thermal equilibrium will be reached, without active contol. On the other hand even a one degree change could get you in trouble reaching sub micron precision. Maybe you could activly control the power of the compressor and some thermal sensors to build an feed back loop that compensate for any thermal expansion.
The "cheap" granit plates I have don't reach sub micron levels. They have a measure protocol showing 2µm deviation on a 300x500mm plate. Grade 00 Plates on the other hand are really expensive. But with a air bearing spindle, maybe you can grind your own to your specifications. PS I am your 2^9 subscriber. Keep up that good work, get a workshop at home :D
One of the great things about air bearings is their thermal stability due to lack of friction. They’ll only generate heat at very high speeds, faster than I would run them. The only potential source of heat is the motor itself at the rear of the spindle, and the self cooling effects of all the airflow through the assembly acts to counteract this. The surface plate flatness tolerances don’t necessarily need to be AA to achieve good precision either. It’s more an alignment tool to make getting the axis and spindle nice and inline easier. An out of flat surface plate that was, let’s say, convex 2 tenths would effect things like diametrical tolerance through a pseudo cosine error type of relationship and wouldn’t effect the tool position relative to the work by much at all. Thanks for the support!
One more question, and I may have missed it (but I did watch the whole video). Are there 2 or 4 radial bearings? Since you are showing a section its difficult to tell. I'm working on a lathe chuck concept with pivoting jaws in order to hold tapered objects. What you had with the ball and socket was very similar to what I have brainstormed thus far. I was using a cylinder for a pivot, rather than a ball, as I didn't think I needed more degrees of freedom than one. You commented that your design is self aligning. (meaning the shaft runs coaxial to the intended headstock Z axis). Is that true if it had 2 radial bearings, or 4 radial bearings, or both? I was actually surprised to hear it was self aligning. With my chuck design I'm looking at 4 pivoting jaws with a screw behind them, just like a conventional lathe 4 jaw independent chuck. And having 2 complete chucks with pivoting jaws separated by a several inches. I thought I needed 2 chucks to do the alignment. But you comment that it self aligns might negate the need for a second chuck in my case. Thanks.
Have you thought about outsourcing the manufacturing of the shaft and sleeve to get the precision you want instead of compromising the design? Or if you lap the critical parts, the limit of what you can make is what you can measure?
Outsourcing something of this precision would become very expensive very fast. I am thoroughly confidant I can lap the thrust faces flat and perpendicular to under 50 millionths (measurement limit is 20 millionths for what I have), which is more than accurate enough for good function.
@@cylosgarage I am just thinking that the spindle would deflect easily for precision application, it should have 2 sets of the pad bearings if you go with that type of design at least.
@@MF175mp the two sets of bearing pads would technically be over-constraint and would make aligning the entire assembly extremely difficult. The moment loads are actually taken up by the thrust bearings counterintuitively, i'm thinking about making a video soon on the spindle statics and forces to better explain how it's rigid despite those floppy radial bearings.
Great project! have you considered using two Hemispheres back to back, |)--(| , I know some other youtubers were discussing it. It will make lapping a lot easier, and likely present some great results.
not entirely sure what you mean by that, please elaborate
@@cylosgarage th-cam.com/video/Jq4a25W_emI/w-d-xo.html
@@DUIofPhysics Ahhh yes i remember that. It so happens that it is much easier for me to manufacture flat planes and simple radii rather than manufacture a massive ultra precision metal spheroid and matching bearing. If I had better manufacturing capabilities I'd consider it but I'm pretty happy with the system I've got worked out now. Thanks for the support!
@@cylosgarage I look forwards to seeing how it works out for you. However, just to note, I've found online sources, you can buy those ball bearings up to 5" diameter, just saw one in half then machine it from there. Here is a 5" one! www.bcprecision.com/collections/chrome-steel-ball-bearings/products/5-inch-chrome-steel-ball-bearings-g25
@@DUIofPhysics wow! that's impressively cheap! Hopefully Adam keeps his word and attempts one of such spindles in the future