[Metrology] How Thick Is A Sliver Of Light

According to "Handbook of Dimensional Meaurement", 30 millionths, with blue light.

https://books.google.com/books?id=qFiUBgAAQBAJ&pg=PA279&lpg=PA279&dq=straightedge+light+gap&source=bl&ots=WRcrfRuw0p&sig=AnDYuQje1_qlsUqd8deUKtb8hsY&hl=en&sa=X&ved=0ahUKEwjsu9Xy97DJAhWJNiYKHUiJB8IQ6AEIJzAC#v=onepage&q=straightedge light gap&f=false

I would have to suppose as far as reliability and trust, it would depend the individual and several other variables. I believe they author is referring to ideal conditions. Defraction of white light would cause you to see a shift from red @ 70 millionths down to the blue at 30. I would assume this would require a full visible spectrum light source.
 
Whatever, it's small enough to tram my mill spindle against the cylinder square!
 
you should have little problem seeing .002", unless you have poor eyesight
.0002" sounds right to me as being very possible.
if you are used to looking at very small gaps, you can quickly tell the differences.

i know it's not the same thing, but when i was a little younger, i changed so many ignition points that i'd set the gap by eye and check with a go-no feeler gauge- the feeler gauge was merely for show...

my point is, once you put a straightedge to something that's not flat- the imperfection will stick out like a sore thumb.
if you have a well lit shop or direct sunlight, you can see different colors of the spectrum.

here's a neat trick if you have bad eyes,
take a piece of scrap opaque material and drill a very small hole (1/16" or smaller) and hold the scrap so that you can see through the hole.
then try to read something you normally could not read without your glasses and tell me the result....
 
Last edited:
Ahahahaha. And here I was thinking someone had stuck a feeler gauge through the slit and come up with a measurement !

Oh wait, I'll run down to the auto part parts store and pick up a new feeler gauge set. Let's see, I need one that measures between 30 and 70 tenths of a tenth. Those are on row 3 I think. That should do it.

Seriously, it looks like 30 millionths is a good candidate from what several people have posted above, accounting for such variables as bifocals, not enuf coffee, crummy batteries in my flashlight, etc.

What brought this on, is my desire to find an expedient method to check the depth of the wear on the ways of used lathes with a straight edge. Many lathes actually, as I am starting to look for an affordable, tight, replacement for my existing big iron - a good portion of which is laying on the floor at the moment - both as parts in need of replacement and as a trail of cosmic dust - having departed the ways in copious quantities over the years it was used to keep the Canadian Armed Forces in the field.
 
See page 279, Tony Wells Google Books reference above (post #11):

Slit of white light : visible down to .0001
Visible spectrum changes to red light (refraction): .00007 -.00005
Red light changes to blue light (refraction): .00003

Way cool to know!
 
I, too, have experienced the blue light phenomena. When checking two surfaces for square, it only slight divergence between the "solid square" and to body being checked, you can see at one end no light, and as you approach the other end, first blue then white light. I don't remember red light, I never gave a thought to measuring it. My Dial Cal doesn't go down that far.
 
It would depend upon the intensity and wavelength of the light. The human eye is most sensitive to the green portion of the spectrum.

I set up an experiment with a parallel (the best that I have for a straight edge) and my surface plate. I used a 300 lumen LED flashlight for the light source and turned off the room lighting. With the parallel edge resting on the surface plate, there was a very small amount of light getting through. I then put a .0015 shim under one end. This gave me a variable width slit theoretically going from .0015 to zero. The illumination of the target was very strong next to the shim, only fading in the last 1/4" of the parallel. The inference being that light is visible through a slit as narrow as .00008" (2 microns).

Light will pass through a much narrower slit but due to the scattering caused by diffraction, there won't be enough gathered by the eye to be detectable.

Bob
 
I've been using a LED flashlight that gives a blue tint and slowly move my eye around looking for light between the part and cylindrical straight edge.
When no light shows and the blue pattern is full or solid, it's far truer than I can measure or can machine.
In the eighty's it was said that a 1 mill gap was about the smallest the human eye could see so several of us tested this and found that with proper lighting we could see gaps of .0002" .
That was without trying hard. Not scientific but I'm not going to spend a lot of money to find how many decimal places I can see to.
It is a good question, an interesting subject, and worth some googling.
I like bob's experiment.
 
Wikipedia recons its about half a micron,

I do know that the size of integrated circuits is limmited in getting much smaller due to the photo lithography aspect of making the waffers. :)

Stuart
 
I've been restoring an old Dalton lathe, made in 1919. This afternoon got the headstock bolted back on the bed and decided to lay a straightedge along the ways to estimate how much wear there might be. I knew before hand it is in pretty good shape- lucked into it in this condition earlier this year... Sure enuf, I got a glimpse of blue light up against the headstock, and a little more pronounced slit of white light a few inches away - in the typical position - just abaft the chuck position. Mostly it was a very narrow slit - maybe a couple of Thou at the worst, then smooth and dark sailing down to the end- where the tailstock will reside. Even better the tailstock ways showed intermittent grains of light most of the way along the bed. Caused I think by the rough surface of my crude aluminum hardware straightedge- not a precision measuring device by any means.

Nevertheless, very happy the blue light test showed what I think is minimal wear on my 97 year old lathe! Hooaah.
 
You must log in or register to reply here.
Back
Top