What's new
What's new

CNC Machining to Six decimal places in metric measurements How to do it?

Street

Hot Rolled
Joined
Apr 16, 2015
Location
Melbourne Victoria Australia
Just looking at a advert Machining titanium they quote to 0.000001 mm as a working tolerance.

How is this even achieved on insert tip turning tooling?

Just how would you measure it with accurate repeatability?

Controlled coolant would be a given, so would warm up

Any ideas?
 
Maybe they meant .000001 meter, there are some machines out there that can hold a micron (small parts) but I don't think they have them.
 
Yes but machinist commonly don't use six decimal places of a metre....maybe some non machinist?
well can take off 3 decimal places. 0.001mm is better but still not great on tip turning tooling, careful grinding but they don't mention that.
Will take it with a grain of salt on the advert and the speil

Thanks.
 
All machining can be done to six decimal places. The trick is knowing what the last two or three will be before you start...

I think they missed an order of magnitude, or two.

Chip
 
Just looking at a advert Machining titanium they quote to 0.000001 mm as a working tolerance.

That's because the marketing department doesn't work in the shop. They don't have to know complex things like where the decimal goes. Then there's typos. A friend of mine had somebody do their website. They had them in a 24,000 acre facility. It's on the internet, so it must be true.
JR
 
That's one nanometre or about the width of two larger atoms.

1 nanometer = 1 e -9 meter

1 angstrom = 1 e -10 meter

An atom is about one angstrom diameter give or take.

A nanometer is about ten atoms, thereabouts. Current lithography is in the nanometer range, tens or hundreds of 'em.

To get to the angstrom range you need something a bit better:

IBM (atoms) - Wikipedia

Publicity thy name is don eigler...
 
That's one nanometre or about the width of two larger atoms.

SAG and Jim Rozen have the numbers that are meaningful. A Carbon hydrogen bond length is about 109 picometers (10^-12m). This is about 10^-10 meters, or actually about 0.1 nanometer (check my math, though!). The spec of 0.000001mm is 10^-9. So the spec is asking for precision to about 10 molecular bond lengths. I'm guessing that they meant meters (m) not millimeters (mm).

The DRO that could handle 0.000001mm would be interesting! They have 'em for special electronic microscopes. I wonder what ABEC class one's mill bearings would have to be?
 
One who did not get the puck up between meteres and mm should be hit up side of head with a dull shovel many times.
They also should not be machinists as duh thinking is way outside their capability.
Seen this in machinist, a number for a hole that makes no sense yet they dump a pile of chips on the desk.
Absolute morons who stand proud, are all technicality right, but look like a jackass to the real world.
The ability to actually think helps.
Sure it is a oh-shit. Can you figure that out or is it way above your pay grade? The fact you posted already answers this.

Engage your brain. Don't be a robot. We need people who can think in manufacturing.
Try to be smarter than my dog.
Bob.
 
Yes a nanometre is ten small atoms wide and two large atoms wide: Diameter of an Atom - The Physics Factbook


1 nanometer = 1 e -9 meter

1 angstrom = 1 e -10 meter

An atom is about one angstrom diameter give or take.

A nanometer is about ten atoms, thereabouts. Current lithography is in the nanometer range, tens or hundreds of 'em.

To get to the angstrom range you need something a bit better:

IBM (atoms) - Wikipedia

Publicity thy name is don eigler...
 
Work IS done to nanometers. And I have actually done it; actually just a little bit finer than that. No CNC involved. Not bragging as many others have also done the same thing. I did it on my screen porch in Florida. A three legged stand made from wood. Two glass blanks; well, one was Pyrex. Some abrasives. Some rouge. A pitch lap. A lot of work and testing. Total investment, some years ago, was less than $100; I was raising a family then and my money was dear.

It is called making a telescope mirror. For test equipment, I used a knife edge tester. Mine is fancy with an aluminum base that was salvaged from some film splicing equipment, a night light bulb in an aluminum tube, a pinhole in the tube, and a knife edge (I used a razor blade) on a mount with a shop made micrometer screw to move it. All put together with wood working tools. It's cost is included in that $100 I mentioned above. OK, truth be known, I did use a lathe to make that micrometer screw; but it was a small lathe so many here will not count that as real machine work.

It is amazing what can be done with the proper technique and some sweat.

Can this level of work be done with modern CNC machines? I don't know, but I would not bet against it. In the worst case, they can't be more than one decimal place away from it. The big problem would be measuring the result. A block of metal may change that much just by turning it on it's side.
 
Another reference: The most precise stuff I've had interest in has been building telescopes. The wavelength of light is 400-700 nanometers. A very good spec for mirror accuracy is 1/10 of a wavelength, so at the low limit, 40nm. This spec calls for 40-fold better better accuracy than a telescope mirror surface!
 
There is the possibility that someone in advertising is using the internal mathematical precision of the control.

Sort of like lying with statistics but technically correct.
 
There is the possibility that someone in advertising is using the internal mathematical precision of the control.

Sort of like lying with statistics but technically correct.

Why/how would someone in advertising do that? Yes the control is a computer, but the position readout doesn't even go that far and most controls will reject the values if they are strung out that long as a redundancy. And what statistic would they be using to manipulate the idea that it would even be plausible? Not technically correct in any way, shape or form.

Sorry, not bashing you over the head, but no.
 








 
Back
Top