The smallest (measurable) dimensions you've worked with?

Joined
Jun 7, 2002
Messages
3,411
This question is open not just to knife makers, polishers and sharpeners. Gemologists, craftsmen, and other industrial workers are welcome to add. Perhaps describing your work would be beneficial for these fora.

Mine is 30 microns (0.03 mm) most rocks and minerals are actually translucent and a very thin section can be analyzed in a polarizing microscope. Preparing mineral samples for microscopy is what i did for a short time with our geologic survey:

1. Cut a slab of that rock/mineral your want to examine using a tungsten saw. Make sure one side is less than 1 inch wide and 2 inches long.
2. Polish the above-mentioned side of the slab on a lapidary (horizontal steel wheel.) First put a small mound of medium-grit (~.5mm-.2mm) carbonrundum powder in the middle and wet with water. Turn on the wheel and polish your work to remove the saw marks and rough features. Move the work side-to-side in a radial direction in relation to the wheel. This will prevent the steel wheel from concaving (and your work from convexing.) It has to be perfectly flat.
3. Shift to fine (< 0.1 mm) carborundum to (visibly) smoothen the work.
4. Final polish of the work is done manually on a glass plate using aluminum oxide power (micron-size).
5. Glue that polished side of your work on a glass slide using melted canada balsam. Care should be taken to prevent bubbles of air in the balsam between the glass and the mineral surface.
6. Back to the tungsten saw. Slice off the main portion of the mineral slab, leaving only a chip glued to the glass slize (less than 3mm thickness.)
7. Back to the lapidary. Stick a small suction cup on the opposite side of the glass slide to use as a handle. Use medium grit to reduce the chip sample to less than 1 mm thickness (careful this time.) Use the fine to reduce below 0.1 mm (100 microns.)
8. Below 100 microns, you stand to work anywhere between 2 hours and 2 days on the lapidary and glass plate, depending on the mineral hardness. In the case of massive quartz, I spent 3 days on my first time. Experienced polishers can actually "see" 0.03mm or 30 microns. I couldn't. For quartz, you determine correct thickness by going to the polarizing microscope and viewing your work. Under crossed polars, 0.03mm of quartz should appear pale yellow whereas mine looked dark violet (> 60 microns!)
9. So back to the glass plate. One polisher already helped me out by clipping the slide down and grinding it to 50 microns using a diamond wheel. anything smaller than 50, the wheel can't help me. so back to the glass and hand-grind using corundum powder. finally, the foreman looked at my sample and said "it think you've finally done it." i went back to the microscpe and sure enough, it was pale yellow. one of my best experiences as a young geologist.
***
 
interesting.....

.03mm approx 1/1000 inch right?? So I used to help out at my families machine shop, made small parts for pratt and whitney, so pieces of F-17 jet engines and others. In general most tolerances were +/- 3/1000 if I remember correctly, I was in high school and they had me doing odd stuff, but I did run some CNC occasionally. Pretty fascinating, machined quite a bit of titanium.
 
I've worked with +/- 1 half thou before (thats .5/1000)....ya...thats small and it sucks working with those tolerances, I work in a machine shop in the summers and when you work with tolerences like that, if you so much as bump the grinder (these grinders weigh probably 800 lbs) you will have to throw the piece out.

when we make these parts, we make double before we start grinding on the small tolerance..... and usually screw up almost too many (ie: if we needed 100 we made 200 and ended up having like 115 good ones or so....ya not fun)
 
wait, 1/1000 of an inch , right? so .5 thou is around 15 microns. i don't think i can see that small a difference. how do you measure, linear or just by fit?

i don't know if my friend is ribbing me but in their microchip plant, machinists can "feel" 1 micron. if they can't they're fired outright. now how true is this?
 
wait, 1/1000 of an inch , right? so .5 thou is around 15 microns. i don't think i can see that small a difference. how do you measure, linear or just by fit?

i don't know if my friend is ribbing me but in their microchip plant, machinists can "feel" 1 micron. if they can't they're fired outright. now how true is this?

No, you can't see that small of a difference, except for visual polish (ie almost mirror v the rest of TW piece). We measured it with micrometers with accuracy of one tenth of a thousandth or as a machinist knows it one tenth or one thou

Ya, I can't tell you about being able to feel one micron, we can only get so much accuracy on a mill and then it is moved to the grinders.....I usually worked the grind shop
 
In a class project, I had to make a dice. Using a micrometer, the results were within 0.00 . So basically 8.00mm , 8.00mm , 8.00mm....maybe if I had another decimal to work to :p. Today I lathed 2 pieces of nylon at 12. 03 to 12.05 mm for a 12.00 tube
 
I can change the angle of an edge 1/100 part of 1 degree &#8211; and go back to the exact starting point again.

Now, there is no need to change an edge as small as that - in my mind, it was a spin off function of a construction I made. I use a screw to do this on a sharpening tool. 1 turn of this screw changes the sharpening angle 7/100 parts of one degree. What I did was to make this screw larger (wider) and I divide it in to 7 equal sections. I can use this screw to visible change the sharpening angle 1 seventh of 1 turn = 1/100 part of 1 degree.

When I do this, I can see the new surface on an edge when I use a really fine sharpener &#8211; but I cannot even feel it with my fingertip. Nice, but unnecessary, spin off function.

Thomas
 
The smallest tolerances I regularly worked in was 0.1um or a tenth of a micron. we were building mechanical filters that could alter wave energy, say x-ray. I worked in the MEMS industry and we built parts using LIGA technology. The biggest part we made was 1x1x0.35cm and it contained 7 parts in a mechanical safe and arming micro scale firetrain. The most dense array I ever designed of mecanical parts was over 72,000 on a 4 inch wafer! All these parts were mechanical parts, no micro electronics. I used a Nikon VMR3020 vmm and laser profilometry. Programming the machine was a pain too.


-Xander
 
.0001"

The parts I design for punching plates for the printing industry have a clearance of .0001".

We have the parts wire EDMed.

.03mm would be .00118"
 
.001", which is the finest any of my measuring tools will do. I once trued a bandsaw wheel to within 1 thou, which made a remarkable improvement on how smoothly the saw ran compared to the 12-14 thou tolerance it had before.

I also made a bandsaw tension gauge, which used a lever action to magnify the blade stretch under tension (which would be something like .0025" at 20,000 psi) so that it could be read accurately by a dial indicator graduated in thousandths. It can measure in increments of 1000 psi (blade stress), which correlates to a stretch of about 0.000125". So technically that is the smallest measurement I have worked with.
 
8 micrometers. I've spliced plenty of fiber, both by hand and using a fusion splicer.
 
Routinely bore holes to plus/minus .0001
 
Last edited:
I really have no idea, but this is a fun topic.

Personally, I do everything by hand, so I can tell you what grits I used to sharpen/polish, but I cannot tell you how effective my human hands are at producing results on par with the grits I am using (I often wonder how many people contemplate this concept when I see maximum grit sizes posted in this and other forums...)

Anyway, I have been doing this for some time and 2 nights ago I achieved what I think to be my finest results yet, and did so on a VERY large blade. Details are not important, but suffice to say, this thing would be "extremely sharp" by any standard I have seen posted here;) I finished on 1.0 micron diamonds which is not all that fine (I used to go down to 0.25) but I have found that there are many limits that exist in materials and methods...and these recent results demonstrate to me that I was previously wasting time with grits smaller than 1.0 microns as evidenced by the superior results I recently achieved.

FWIW this is what I used to think was "really sharp";)
[video=youtube_share;D7wdYJUU9vo]http://youtu.be/D7wdYJUU9vo[/video]
 
Routinely bore holes to plus/minus .0001

I wouldn't say I ROUTINELY bore holes to +/- a tenth.

A more common bearing fit where I work is +.0005/-0.0, or plus 5 tenths minus zero.

So that would be +/- .00025.

Boring to within a tenth is a pain, especially if you land within a tenth or two of tolerance. You adjust the boring bar up a tenth at a time and it doesn't take anything off due to cutter deflection or rubbing/burnishing, then suddenly it will take 3 tenths off.

A tenth (.0001") is a tenth of a thousandth (.001").

A piece of paper or a hair is two or three thousandths thick.
 
Last edited:
I used to work building high speed drilling machines for manufacturing printed circuit boards (this was about 30 yrs ago). They were guaranteed to be accurate within a ten-thousandth in any axis (assuming it was properly set up in a climate controlled facility). To get that, we had to align the rails that the table moved on to within a one-hundred thousandth (.00001) of straight. We used a colimated beam of light to check it and would have to check it (and usually reset it) until it passed 3 days in a row.

It was amazing what would affect the straightness of an @1-1/2" steel rail. They used to cost about $50,000 before adding the CNC or programs. The drills themselves ran on air bearings- remeber, this was 30 years ago- very high tech!
 
For comparison:

The wavelengths of visible light range from about 0.4 microns to 0.7 microns.
A modern razor has an edge of around 0.4 microns (see Verhoeven's electron microscope pictures of knife edges in _Experiments in Knife Sharpening_).
Diamond microtome knives are sharp to about 0.005 microns.

Personally, I've done some very basic machining to around 0.003" (around 76 microns) on a metal lathe.

Sincerely,
--Lagrangian
 
I work for a company that works to the 22 nanometer (22 nm) node for the CMOS process
 
hmmm... so far:

1 mineralogist (me)
6 machinists
3 in micro-electronics
1 in printing
1 in nanotech
2 knife sharpeners
***
still no eye doctors and microsurgeons, no surveyors, none in biotech, no artillery/ordnance men, no astronomers.
 
If you let me jump off the deep end, and include biotech and astonomers, then you get the following:

1. For drug design, biochemists like to know molecular structures to sub-angstrom levels (1 Angstrom = 10^-10 meters). A carbon-hydrogen bond is approximately 1 angstrom. Often, they can use x-ray crystallography to find the shape of molecules to this level of accuracy. Alternatively, they can use nuclear-magnetic resonance (NMR). My graduate degree is in using computers to interpret and analyze NMR data.

2. Astronomers use telescope mirrors which are accurate to 1/10th or 1/20th wavelength. For optical telescopes, this would mean 0.040 microns (40 nanometers) or smaller.

And if you include material science people, then you get to the atomic level (just like biochemists, so around 1 angstrom). In electron microscopes, scanning-tunnelling microscopes, and atomic-force microscopes, they get down to looking at single atoms in crystals and on surfaces. Scanning-tunnelling and atomic-force microscopes use the sharpest needles in the world, which often have a pointy tip of just one atom.

And if allow theoretical physicsts, they talk about the Planck length. Distances smaller than the Planck length do not exist in any meaningful way in quantum field theory. The Planck length is 1.6×10^&#8722;35 meters and represents the intrinsic "fuzziness/blurriness" of space-time (according to quantum field theory).

I'm not sure including the theoretcial physicts would be "fair" because it's just theory; so far, there are no observations or experiments at the Planck Scale. But the chemistry and material science people study, design, and make stuff at the atomic and molecular level. If you want, you can include nuclear physicists, who work with atomic nuclei, which are about 10,000x to 100,000x smaller than an atom. So somewhere around 10^-15 meters. Since some nuclear physicsts create new elements, I suppose you could argue they also create things this small. :)

Okay, enough of jumping off the deep end!
Mostly I'm interested in stuff in the 0.5 micron range for knife sharpening and abrasive grits.

Sincerely,
--Lagrangian

P.S. http://scaleofuniverse.com/
 
Back
Top