Micro CNC

[Mike Everman]

I've been waiting for something interesting to show on this, and tomorrow will be the first parts made with our Bell-Everman Mini Polar CNC.

The basic architecture is A, B, Z axes. "A" rotates the material, which has been bonded to a 6mm stud. The stud is placed in a 180 degree indexer, so that one side can be finished and then flipped to do the backside.

The "B" axis rotates the spindles, with twin spindles being an initial take on eliminating the future need for a tool changer that I will ultimately want.
We are using a Delta Tau Clipper control, so that all part data coming in as XYZ is automatically converted to ABZ polar coordinates. The video shows a test pattern, and I will have a part making video up soon.

I plan several material carrier styles, and am thinking toward watch making applications, where precious metal chips need careful collection.

The current design has two spindles with coolant sprayers. The spindles are 200,000 rpm air driven types, with 1/16" shank cutters, but they have recently been discontinued, so we are in search of a replacement. I may in fact be making our own, with an ER-8 collet extension as a beginning point. The approach I'm taking for spinning it up (again with air) has been breadboarded and reached 165,000 rpm, so almost there!

Video and more is at http://www.youtube.com/watch?v=LziQ4uTXhBg
And more in the way of pics and write up on my blog: mikeeverman.com

Hope you guys like it! Almost ready for prime-time!

Any CNC control scheme would work if you don't mind doing the polar conversions at the CAM stage.

BTW, I had to take down the previous build log site, as almost all photos had a proprietary bit of my customer's. This and all future info will be more generic.

[whiskeybelly]

Sounds great Mike! I'm curious about the spindle RPM though.

I know when doing the calculations for very small diameter cutters and drills to obtain the ideal surface speed very high RPM is required, but I've also observed many times that the feed rate and thus the chipload per cutting tooth is overlooked resulting in overly light loading which leads to early cutter failure.

In the non-CNC world I have seen a number of little Hauser jig borers converted from their pokey 3K RPM spindles to high speed spindles of 20-30K. The result, with manual feeding, is usually disappointing as the required feed rate is frighteningly fast at those spindle speeds. With a human 'computer' at the helm the 3K spindle is actually plenty fast and tool life and quality of finish is quite good.

With a CNC of course the feedrate is entirely programmable and you can push the cutter to it's max potential very easily. But 200K is in the realm of grinding speeds; I'm sure this is already in your calculations, I just bring it up because I've seen some very clever mechanics scratch their heads over tooling failures when overlooking chipload/feedrates.

[Mike Everman]

This high speed is for cutting ceramics with diamond bits. You make a good point, I'll have to look very closely at the spindle speed options for cutting metals. It occurs to me that i've never seen feeds and speeds for gold and the other rare metals.
I've been so absorbed in the application of the moment, I've not been looking so hard at variable lesser rpm's.
I am not very limited by feedrate, which is nice.

[whiskeybelly]

Mike-

For a rough rule of thumb you can figure .01mm per cutting lip per rev. for feed rate. This would be for 1mm cutters/drills down to around .5mm. From .5mm down cut the feed in half- .005mm. Above 1mm you can add a little more. This is rough, and roughly applicable to anything from steel in its annealed state to brass, german silver, and gold.

With that in mind, with a 3K spindle speed and a two lip .5mm drill as an example, you will be turning 50 revs per second, times the 2 lips, so the advance is 1mm (50x2x.01mm) per second- which is about perfect for a human .

Common spindle speeds on decent CNC machines run from 20K-50K RPM. Imagining a 30K spindle the feed is 10X the above example so 10mm per second. A little crazy for a human (though I have seen plenty of manual drill presses designed for similar speeds.. and hate them!), but fine for a computer.

At 100K the feed would be 33mm per second, for a two lip tool. A four flute endmill would double that. And 200K doubles that . I don't doubt that a well spec-ed CNC can handle that, especially a small well designed one considering some of the insane moves I've seen on multi-ton 5 axis monsters. I would think though that there may be some problems simply dealing with inertia with sudden toolpath direction changes on complex parts. I only have the most rudimentary grasp of G code and programming so this may well be something you experts have down without even thinking about it. And I'm certainly not trying to be critical of your design- if anything I suppose I'm suggesting that a lower speed spindle would not be a hindrance to performance when using cutting rather than grinding tools.

[Mike Everman]

Thanks much! I will consider this carefully when moving to metal cutting and spindle choice.
I just posted some vids on my blog of it cutting Zirconia with diamond tools. It was pretty wild, and we broke half a dozen cutters trying to go too fast.

Speed will not be an issue with this setup, and I expect I'd be using 20-30k spindle speed for metal micromachining in the future.

[Ophiuchus]

Mike, that video of the zirconia machining is impressive. I've never heard of anyone machining that material before.

After looking more into Gene Clark's work, I noticed he used iridium for some parts, and I was looking into machining iridium myself. I've never come across any machining data for milling or turning pure iridium, or pure beryllium either- I was thinking of using both.

I'd be very interested to see what other extremely exotic stuff your amazing new mill can handle.

[Mike Everman]

I don't know anything about machining iridium, but I do know that beryllium is extremely hazardous to your health!

[Paul Loatman]

Mike,

This project is coming along very well. I've never seen a machine quite like this before, and definitely not at this scale. Do you plan on selling a generic version of these to the public? If you do, and if i can scrounge up the cash, i'd buy one.

Also, in my opinion, the health concerns regarding beryllium far outweigh any real benefits you'd see from using it in watchmaking, it's not really that great. I would just stay away from it, too much of a hassle if you ask me, and that's coming from a guy that spent, literally, 12 hours cutting a 2inch square block of steel in half with a piercing saw.

-Paul

[Mike Everman]

Thanks, Paul.
I definitely will be selling this for other than dental applications, and I'm presenting it here mainly to get input from guys that do the strictest type of micro machining. And of course, I'm a timepiece enthusiast, so one of my dreams is for this to become a "go to" tool for watchmakers.

I'm trying to keep it under $20k list with single spindle and controls, but the spindle is a wild card, and I may design it myself so I can have a tool changer. It currently is a Delta Tau clipper control.

I do think that I'd like to sell some to intrepid souls that want to work out the kinematics and controls for themselves, to see what they come up with.

There is going to be a direct feedback option for stupid high resolution, too.