Optical flats.

[SC_RUFCTR]

Hi All

Educate me about Optical Flats. I've seen them around in labs and tool making areas but I have no real idea what they're used for.
The reason I ask is because they are such a precision item with impressive tight tolerances and characteristics.

[simonl]

Hi All

Educate me about Optical Flats. I've seen them around in labs and tool making areas but I have no real idea what they're used for.
The reason I ask is because they are such a precision item with impressive tight tolerances and characteristics.

Hello,

I'm certainly no expert but in my Uni days we used optically flat glass (optical flats) when using laser for very accurate and sensitive distance measurement. There are many different types of apparatus that are also designed to be optically flat. The glass is typically made from thermally stable material e.g. Pyrex so that diminesional changes due to temperature differences are minimised, similar to the manufacture of telescope lenses and mirrors. Example of use: There are certain circumstances when you may wish to pass a beam of light through a medium such as glass and not have it scattered by surface imperfections. Another reason why the glass needs to be accurate relative to the wavelength of light is because small variations in path travelled by the same beam of light (due to surface imperfections) will cause optical interference which may be undesirable in certain circumstances. The term "optically flat" infers that the surface imperfections are configured to within the order of wavelengths of light. For instance, an optically flat piece of glass may be configured to within a half wavelength. The optical spectrum of light is between 400 - 700 nano metres (1 nanometre = 1000 microns) so take 555nm (middle of the spectrum, colour green) half a wavelength would be approx 280nm or 0.28 microns or just over 1/100 of a thou.

Just a small example for the uses of optically flat glass and what is means.

Cheers,

Simon

[GZBMW]

Optical flats are deemed to be (in the world of metrology) as the flattest an object can be.

They offer flatness within 25 nm (0.000001").

The are use to check out of flatness condition of gauges and instruments such as gauge block or the anvils of micrometers.

So long as the surface in question is relatively flat, flat ground and lapped and reflective, the light passing through the optical flat placed on top of the item being checked will reflect back. A fringed pattern will appear as shown attached. This is a quick way to see the flatness of the entire surface. Under green or pink monochromatic light, the dark bands are roughly 250 nm apart.

[Stustoys]

Hi Guys,
Great stuff but I am struggling with some of your numbers
"(1 nanometre = 1000 microns)"? Backwards or did you mean something else?

"roughly 250 nm apart" If the lines are only 0.000250mm apart doesnt that mean that block in the picture is only about 0.000750mm wide? What am I missing?(or is this how it looks under normal light?)

Should "roughly" and "nm" be in the same sentence? ;)


To many 0's

Stuart

[Greg Q]

I don't think the dimensions of the fringes are significant, just the population. I have a nifty little Mitutoyo mic calibration set here that has an odd assortment of blocks, and an optical flat. It was meant for cal labs to use in certifying micrometers. The flat is used to asses the anvils, and the gauge blocks are used to check the mics at random positions on the screw*

* Most user cal rods that you see measure mics at multiples of the screw pitch. Thus, they are only checking the same radial position on the screw. A set of cal blocks, on the other hand, checks the thimble screw at non-repeating intervals and hence is a much more trustworthy staistical approach to metrology calibration.

If you were measuring for sheep stations you'd of course want the same deviation map as your gauge blocks. My set varies all over the place +/- a millionth or two, but they were last checked 40 years ago. Now they could be anywhere until they are again sent in for cal. which ain't going to happen any time soon. If the optical flats say a particular block is suspect then it'll get red sharpie all over it and I 'll make do without it.

Greg

Aiming to have 1.137 beers before dinner

[simonl]

Optical flats are deemed to be (in the world of metrology) as the flattest an object can be.

They offer flatness within 25 nm (0.000001").

The are use to check out of flatness condition of gauges and instruments such as gauge block or the anvils of micrometers.

So long as the surface in question is relatively flat, flat ground and lapped and reflective, the light passing through the optical flat placed on top of the item being checked will reflect back. A fringed pattern will appear as shown attached. This is a quick way to see the flatness of the entire surface. Under green or pink monochromatic light, the dark bands are roughly 250 nm apart.

Hi there,

I'm interested in the fringes in the picture. Being used under monochromatic light (light of a single wavelength) I was under the impression they would be interference lines or an interference pattern. If this is the case then wouldn't they be spaced by integers of whole wavelengths?

Or are these lines created by some other optical phenonema?


Hi Stuart,

It is customary to group numbers into groups of thousands, or 3 orders of magnitude. For instance,

SI unit for distance is Metre unit m

so going up, 1 Km = 1000m (nothing new here)

and, going the other way in smaller dimensions....

1m = 1000 millimetres (milli means 1/1000)

1mm = 1000 micrometres (micro means 1/1000000)

1 micrometre = 1000 nanometres (nano means 1/1000000000)

1 nanometres = 1000 picometres (pico means 1/1000000000000)

insert another measurement here in addition to the standard numbering system:

The angstrom. 1 angsrom = 10 nanometres. Angstroms are generally used in crystalography and other sciences that deal with dimensions of approx the size of atomic spacings since the average atomic spacing is approx 1 Angstrom or multiples of.

1 picometres = 1000 femtometres

1 attometre = 1000 femtometres

So on and so on.....

Hope this helps!

Cheers,

Simon

[Stustoys]

Hi Simon,
I was with you till I got to here
"1 picometres = 1000 femtometres

1 attometre = 1000 femtometres"
A quick google got me to "1 femtometer = 1000 attometer"
Not likely to be a problem for me for awhile yet :)

Thanks

Stuart

[Ueee]

Hi Simon,
I was with you till I got to here
"1 picometres = 1000 femtometres

1 attometre = 1000 femtometres"
A quick google got me to "1 femtometer = 1000 attometer"
Not likely to be a problem for me for awhile yet :)

Thanks

Stuart


Don't you hate it when your shed warms up by 2 degrees and everything is many femtometres to big. Geez even spell check doesn't know what that is. Good work Simon, you have opened my eyes to the truly small......:2tsup:

Ewan

[simonl]

Hi Simon,
I was with you till I got to here
"1 picometres = 1000 femtometres

1 attometre = 1000 femtometres"
A quick google got me to "1 femtometer = 1000 attometer"
Not likely to be a problem for me for awhile yet :)

Thanks

Stuart

Oops. Sorry. I did it from memory but i must admit i may be a tad rusty!

[fubar]

Did a precision grinding diploma back in the eighties The first assignment was to grind an optical flat by hand Cannot say how long that took other than wore out a pair of runners stepping in a circle going round a grinding plate going down through the different grades of emerys to the oxides .Could probably still do the stroke in my sleep today.

[Stustoys]

Hi Simon,
I do it all the time.
Just making sure I wasn't missing something(which I also do all the time:doh:)

Sounds interesting fubar, I've only read about making lenses like that. I guess making a flat is pretty much the same?

Stuart

[Ueee]

Oops. Sorry. I did it from memory but i must admit i may be a tad rusty!

I'll have forgotten by tomorrow morning, but i guess i have never come across it before. At least next time i do i'll go "hey where have i heard that before?"

Thanks again for the fascinating incite

Ewan

[Pete F]

Sounds interesting fubar, I've only read about making lenses like that. I guess making a flat is pretty much the same?

Stuart

I too thought it was a similar process and would be very keen to hear more about the process. I've heard it's actually not all that difficult (though I guess that's a relative term), but does take a lot of patience.

Pete

[simonl]

I too thought it was a similar process and would be very keen to hear more about the process. I've heard it's actually not all that difficult (though I guess that's a relative term), but does take a lot of patience.

Pete

I'd be interested to hear about grinding a flat too. I know that grinding a parabolic mirror is a relatively simple (but long) process of random movements of one flat over another in a W movement using varying grades of grinding paste. Turning the pieces bit by bit as you go, and the randomness of the process tends to turn the top into a concave and the bottom convex. The longer and more random the process, the more accurate you get. Accuracies in the order of a 1/4 wavelength of light (150nm) are easily achievable by the backyard punter who wants to grind his/her own mirror for a telescope. In theory, a good telescope mirror should be at least a 1/4 of the diameter in thickness. So, an 8 inch mirror should be at least 2 inches thick.

More useless crap from ME!

Simon

Simon

[RossA]

An Attometer is only 10-18 while a


Yoctometer (ym) is 10-24 (0.000000000000000000000001) m


At the other end of the scale we have a Yottameter (Ym) and it is 1024 (1,000,000,000,000,000,000,000,000) m


The estimated size of the universe is 930 Ym (93,000,000,000 light years)


Can’t claim any real knowledge, very interesting website which shows it all graphically is HTwins.net - The Scale of the Universe (http://htwins.net/scale/)


Cheers,
RossA

[simonl]

An Attometer is only 10-18 while a


Yoctometer (ym) is 10-24 (0.000000000000000000000001) m


At the other end of the scale we have a Yottameter (Ym) and it is 1024 (1,000,000,000,000,000,000,000,000) m


The estimated size of the universe is 930 Ym (93,000,000,000 light years)


Can’t claim any real knowledge, very interesting website which shows it all graphically is HTwins.net - The Scale of the Universe (http://htwins.net/scale/)


Cheers,
RossA

Cheers Ross! I never new that. I'm familiar with light years and the speed of light in a vacuum (2.997x 10 8 M/s) or 3.0 rounded up. but not big big numbers like that!

Does anyone know how to show "to the power off" on the keyboard?

Assume not many people would understand 3.0E8?

Simon

[RossA]

For some strange reason a number which sticks in my head is 39.37007874

Most of you will recognise it as the number of inches in a metre, thou's to a mm, etc. but when I quote it to my non engineering friends they look at me strange.

Don't understand why. :wacko:

Cheers,
RossA

[simonl]

For some strange reason a number which sticks in my head is 39.37007874

Most of you will recognise it as the number of inches in a metre, thou's to a mm, etc. but when I quote it to my non engineering friends they look at me strange.

Don't understand why. :wacko:

Cheers,
RossA

I'm going to keep that one! Ive never seen it quoted to so many sig figs before!
I'm ashamed to say I work off 40! And, it's still more accurate than any of my work so far! :B
Cheers

Simon

[Stustoys]

Hi Simon,
Isn't ^ "to the power of"?
I'd understand 3.0E8 faster than 3.0^8

Stuart

[Ueee]

The one number i never forget is 16.387064. The no of CC in a CI. I used to fly model aircraft and we were always converting engine sizes. Some reason the number stuck!

A Yoctometer reminds me of the Y page in my youngens A-Z book, kind of a cross between a yacht and a yak.....

One thing that baffles me is a Googleplex(a one folowed by a google 0's). I have heard even if you write a 0 on every surface on earth, you would still run out of space- it could be crap though.
Ewan

[simonl]

The one number i never forget is 16.387064. The no of CC in a CI. I used to fly model aircraft and we were always converting engine sizes. Some reason the number stuck!

A Yoctometer reminds me of the Y page in my youngens A-Z book, kind of a cross between a yacht and a yak.....

One thing that baffles me is a Googleplex(a one folowed by a google 0's). I have heard even if you write a 0 on every surface on earth, you would still run out of space- it could be crap though.
Ewan

Love model airplanes. Stupid imperial system though. 40 size engines, 60 size etc etc. So tell me, are you an electric motor fan or do you still prefer the sound of the glow plug methanol motor?

I would love to get back into it. So many cool hobbies, so little time.....

Simon

[Ueee]

Love model airplanes. Stupid imperial system though. 40 size engines, 60 size etc etc. So tell me, are you an electric motor fan or do you still prefer the sound of the glow plug methanol motor?

I would love to get back into it. So many cool hobbies, so little time.....

Simon

Glow plug only! I haven't flown in about 8 years, would probably just crash now. Those ABC motors are pretty amazing. 50 size (8cc ish) will do 15,000 rpm(dependent on prop size), but they are sooooo simple! Just straight ported 2 stroke, single needle carb, I reckon 20% of the fuel doesn't even get burnt, it goes in and out the exhaust in the one stroke. For those who have never seen one, they also run on a pressurized fuel tank, so they wont starve themselves if you fly upsidedown etc. All they have is a pipe running from the muffler to the tank. As always, simple is best!

[simonl]

Glow plug only! I haven't flown in about 8 years, would probably just crash now. Those ABC motors are pretty amazing. 50 size (8cc ish) will do 15,000 rpm(dependent on prop size), but they are sooooo simple! Just straight ported 2 stroke, single needle carb, I reckon 20% of the fuel doesn't even get burnt, it goes in and out the exhaust in the one stroke. For those who have never seen one, they also run on a pressurized fuel tank, so they wont starve themselves if you fly upsidedown etc. All they have is a pipe running from the muffler to the tank. As always, simple is best!

A bit OT but hey I'm hearing you. I like the way you tune them too. Max noise and then increase the mixture by a couple of clicks!

The only thing I wouldn't miss is all the caster oil over the wings, fuse and my hands!

Still I love the way they scream. Only one thing better than a glow plug motor would be a turbine. Now we're talking! :U:U

[Ueee]

A bit OT but hey I'm hearing you. I like the way you tune them too. Max noise and then increase the mixture by a couple of clicks!

The only thing I wouldn't miss is all the caster oil over the wings, fuse and my hands!

Still I love the way they scream. Only one thing better than a glow plug motor would be a turbine. Now we're talking! :U:U

I went to a model air show in Nowra once, and other than the usual historical stuff and heli's cutting the grass with their rotors the main event was a speed test. There were 3 turbine engined "wings" all testing. I can't remember their top speed, but just the way they climbed vertical was amazing. I also recall the starting technique, and air gun to wind them up and a blowtorch to ignite. I think they went through about 2L of avgas in an average 10 minute flight.

Back to the subject of flats (its funny how threads can get a life of there own)

I have seen enough shows on Hubble etc to know how you lap an optical lens, ie one that is convex or concave, but how do you make a dead flat, er, optical flat?
I saw a thing about the Kek? observatory and how the mirrors were finished using an electron microscope, but there must be another way?
Surely if you lap your soon to be optical flat on a known flat, the known flat will not be flat after the first stroke?

Ewan

[simonl]

I went to a model air show in Nowra once, and other than the usual historical stuff and heli's cutting the grass with their rotors the main event was a speed test. There were 3 turbine engined "wings" all testing. I can't remember their top speed, but just the way they climbed vertical was amazing. I also recall the starting technique, and air gun to wind them up and a blowtorch to ignite. I think they went through about 2L of avgas in an average 10 minute flight.

Back to the subject of flats (its funny how threads can get a life of there own)

I have seen enough shows on Hubble etc to know how you lap an optical lens, ie one that is convex or concave, but how do you make a dead flat, er, optical flat?
I saw a thing about the Kek? observatory and how the mirrors were finished using an electron microscope, but there must be another way?
Surely if you lap your soon to be optical flat on a known flat, the known flat will not be flat after the first stroke?

Ewan

Dunno how they make optical flats. I never knew they made them to the accuracy of 25nm. You can't make something that accurate using light since the dimensional accuracy is much smaller than the wavelength of light that you use. It's a bit like trying to look at a single atom using an optical mircoscope. It's never going to happen since an atom is about 5000 smaller than the actual wavelength of the light you are using to "see it" That's why they use electron mircoscopes. When electrons are accellerated to required speed, they exhibit "wavelike properties" the higher their speed, the smaller their wavelength known as the Debroglie wavelength (spelling not right) gets. Wavelengths of the order of atomic spacings can be achieved, that's how we are able to "see" pictures of atoms using electron microscopes.

If you use a wavelength that is too large, for an object that is too small, you will end up with a diffraction pattern, where the light bends and you end up with a distorted image or no image at all.

Cheers,

Simon

[RayG]

Ewan, Simon,

I'm another ex-model aircraft nut, mostly gliders in the later years, but started flying FAI combat with oliver tigers and super tigre G15's back in the 1960's... still got a 1/4 scale kestrel up in the roof storage somewhere... must get back into it one day, but I think I'd be going electric nowadays.

Back to optical flats,
I think maybe you can see smaller than a wavelength, by measuring the spacing between the interference fringes between two plates, wouldn't that be measuring less than a wavelength?

Regards
Ray

[simonl]

Ewan, Simon,

I'm another ex-model aircraft nut, mostly gliders in the later years, but started flying FAI combat with oliver tigers and super tigre G15's back in the 1960's... still got a 1/4 scale kestrel up in the roof storage somewhere... must get back into it one day, but I think I'd be going electric nowadays.

Back to optical flats,
I think maybe you can see smaller than a wavelength, by measuring the spacing between the interference fringes between two plates, wouldn't that be measuring less than a wavelength?

Regards
Ray

I assume you mean use light to take measurements smaller than the wavelength of light? Well, yes maybe but not 20 times smaller, as in the case of an optical flat being accurate to 25nm.

Like I said in a previous post, if those lines are indeed interference fringes then their spacing is exactly one wavelength apart.

This is the way it works (in a nutshell) if you have two beams of light of the same wavelength (monochromatic, also each light source must be of a single phase, otherwise you will have interference within the beam. That's why they use laser) and you bring them together, they will undergo interference. Where a crest of one from a wave of one source meets a crest from the other source, you get constructive interference. Constructive interference are the bright bands on the interference fringes. Conversely if you have a trough meet a crest you get destructive interference and they are the dark bands.

LASER: Light Amplification by Stimulated Emission of Radiation light sources are highly monochromatic and all the waves are of the same phase angle, so they produce a constructive light pattern. Thats why they appear so bright even at low powers

Simon

[RayG]

Hi Simon,

Yes, we are on the same wavelength... (pardon the pun)

If you put two optical flats together, you will see interference fringes across the flats, they are one wavelength apart, but if we look at the spacing between the fringes, then the wider the spacing the flatter the surface. So we can measure the spacing between adjacent fringes and interpolate down to fractions of a wavelength.

Let's say we can measure to 0.1mm, and the spacing between adjacent fringes is 10mm, then we are in effect measuring flatness to within 1/100 of a wavelength.. That is the slope of the difference between the two surfaces is 500 nm over 10 mm ...............of course I could be completely wrong... :)

Of course that doesn't tell you if either surface is in fact flat, all it tells you is the difference between the two, if one is a master that we know for sure is flat, then the other surface must be flat as well.

Regards
Ray

[Graziano]

Making optical telescope glass mirrors and optical flats both start with the same basic materials in terms of glass blanks, grinding and polishing media. When you make an optical flat you have to make a set of three flats at a time and just like in scraping, if you check all possible combinations of contact and if they mate then they must be flat.

When you lap two discs together, the disc on the bottom (facing upwards) will go convex or dome outwards and the disc on the top (facing down) will go concave and dish inwards. The longer you go, the more pronounced the curve gets. you want this to happen for a glass telescope mirror so you always have the sacrificial glass disc "Tool" on the bottom and the wanted mirror on the top when lapping them. In terms of time taken, making an optical paraboloidal mirror takes a small fraction of time to grind vs a set of optical flats at only a ~100 hours or so for a 6-8" inch mirror. For an optical flat you need to label the three flats A,B & C and ensure all three get equal lapping time on top and bottom so they will tend to go dished when lapped on top (lapped surface facing down) and tend to go domed when lapped on the bottom (surface facing up). You end up with a lapping sequence A-B, A-C, B-A, B-C, C-A, C-B over and over.

A local guy has made them for fun and subsequently sold sold them for use as reference standards Astro-Tel - Flat Tester (http://www.turbofast.com.au/astrotel/flattester.html), they can be tested by a couple of different methods and once made can be checked against each other to determine how flat they are so in a way they are self checking. The photo below shows two of the local Cairns guy's 8 inch pyrex flats with a small shim between the two flats and illuminated with a single colour of orange sodium vapour discharge to make the fringes much more defined.

http://www.turbofast.com.au/astrotel/img/flattest.jpg

The fringes you see are the combined errors of the two flats, the lines curve to the left at the bottom of the photo which indicates the flats have a turned down edge which in this case were fixed by grinding off and chamfering the edge rather than spending a few hundred more hours lapping off. The dark line in the middle is the wire straight edge to eliminate parallax and provide a reference straight line. Just placing your thumb on the glass will expand it and give a small rise of several fringes. The necessary homebrew cider bottles are to the left of the pic while you wait the three hours or so for the glass flats to thermally cool before optically testing around 3AM when there are no trucks going past on the road to upset the measurements.

If you want to see some fringes for yourself you can get two small sheets of clean float glass and place them one on top of each other with a white sheet of paper underneath. Under fluoro illumination you should see a few hundred~1000 faint greeny-pink fringes per inch.

[simonl]

Hi Simon,

Yes, we are on the same wavelength... (pardon the pun)

If you put two optical flats together, you will see interference fringes across the flats, they are one wavelength apart, but if we look at the spacing between the fringes, then the wider the spacing the flatter the surface. So we can measure the spacing between adjacent fringes and interpolate down to fractions of a wavelength.

Let's say we can measure to 0.1mm, and the spacing between adjacent fringes is 10mm, then we are in effect measuring flatness to within 1/100 of a wavelength.. That is the slope of the difference between the two surfaces is 500 nm over 10 mm ...............of course I could be completely wrong... :)

Of course that doesn't tell you if either surface is in fact flat, all it tells you is the difference between the two, if one is a master that we know for sure is flat, then the other surface must be flat as well.

Regards
Ray

Thanks Ray,

I understand what you are saying. I'm very rusty with my optics and physics but it does make sense to me and sounds like you know what your talking about. I still find that stuff very interesting. Thanks for your info on the subject and thanks.

Hi Graziano, thankyou for taking the time to explain the methods and science behind the system. I very much enjoyed reading your post. Thanks!

Cheers,

Simon

[Ueee]

Hi Graziano and Ray,
Thats exactly what i wanted to know, thanks for the incite!

Ewan

[Pete F]

If you put two optical flats together, you will see interference fringes across the flats, they are one wavelength apart, but if we look at the spacing between the fringes, then the wider the spacing the flatter the surface. So we can measure the spacing between adjacent fringes and interpolate down to fractions of a wavelength.

Let's say we can measure to 0.1mm, and the spacing between adjacent fringes is 10mm, then we are in effect measuring flatness to within 1/100 of a wavelength.. That is the slope of the difference between the two surfaces is 500 nm over 10 mm ...............of course I could be completely wrong... :)

Of course that doesn't tell you if either surface is in fact flat, all it tells you is the difference between the two, if one is a master that we know for sure is flat, then the other surface must be flat as well.

Regards
Ray

I don't know about that Ray. I don't know the first thing about optical flats, but just logically working through your explanation, if I understand correctly it seems to me you're assuming the surface changes linearly and thus you're able to interpolate between the fringes to presume the dimensions at the intermediate points. In practice it's probably pretty close to the truth too, but what's to stop the surface being dead flat, and then suddenly stepping down a wavelength? Whether the surface slopes linearly or goes in a series of steps, if the steps are in the correct locations you could theoretically see the fringes in precisely the same locations.

Hopefully that makes sense. My point being that I can see how the flat can measure down to wavelengths of light, and as you said, that is given by the spacing (and shape?) of the fringes. However I feel what happens in between the fringes (ie at less than multiples of the wavelength) is anyone's guess. I happen to think your suggestion is probably how the surface would behave in real life, but I don't think it could be said definitively.

All a bit hypothetical, but I like to try to understand how things work, sometimes by nutting things out publicly, so hopefully I'm understanding these. What is it about curiosity and cats? :D

Pete

[RayG]

Hi Pete,

What you do is, when you put the two flats together, is to slide them, or rotate one against the other and watch how the fringes move, if there is a step-wise discontinuity, then the you will see it in the way the fringes move. ( or don't move )

Regards
Ray

[Pete F]

Ok that would make sense. From what you said then I'm assuming optical flats are never so flat that they won't have any fringe errors? If one flat was "perfect" your suggestion won't work and the fringes wouldn't move.

Pete

[simonl]

Hi Pete,

What you do is, when you put the two flats together, is to slide them, or rotate one against the other and watch how the fringes move, if there is a step-wise discontinuity, then the you will see it in the way the fringes move. ( or don't move )

Regards
Ray

By any chance, are these interference rings also known as Newtons Rings? I remember setting up a similar situation but one surface was not flat and there was a fluid medium between the two surfaces. It created a series of rings, not lines and they were called Newtons Rings. Assume Issac Newton did some work on a paper on these some time back....

It's a bit fuzzy (my memory that is) as it was a long time ago at RMIT Applied Physics Department.

Simon

[RayG]

Ok that would make sense. From what you said then I'm assuming optical flats are never so flat that they won't have any fringe errors? If one flat was "perfect" your suggestion won't work and the fringes wouldn't move.

Pete

Hi Pete,

No, they would both have to be perfectly flat, and you would see the number of fringes decreasing as it got closer to flat. Eventually one fringe, then zilch... Then you can always shim one corner to see the fringes.

Hi Simon,

Newton was the pioneer of optics, I know the term, but can't remember what it refers to....

Ok, good old Wikipedia.... try here Newton's rings - Wikipedia, the free encyclopedia (http://en.wikipedia.org/wiki/Newton%27s_rings)

Regards
Ray

[Pete F]

Hi Pete,

No, they would both have to be perfectly flat, and you would see the number of fringes decreasing as it got closer to flat. Eventually one fringe, then zilch... Then you can always shim one corner to see the fringes.

Regards
Ray

Ray you said if the flats were rotated the fringes would change. I can see how that would be the case, as the distance between each individual point will change according to the irregularities of the surfaces with respect to each other. However if one flat is "perfect" then I can't see how the fringes can change as the flats are rotated; the relative distance between each point on the surface won't change. Maybe I'm not explaining myself very well. What I mean is that assuming you put two "imperfect" flats together, rotate them and the fringes will change. Put two "perfect" flats together and you won't (without shimming) see any fringes at all, so rotating them won't indicate anything. Now take one of those "perfect" flats and place it against the imperfect flat and rotate it. I can't see any reason the fringes should change in the latter case.

Pete

[RayG]

Ray you said if the flats were rotated the fringes would change. I can see how that would be the case, as the distance between each individual point will change according to the irregularities of the surfaces with respect to each other. However if one flat is "perfect" then I can't see how the fringes can change as the flats are rotated; the relative distance between each point on the surface won't change.

Pete

Hi Pete,

If one is perfectly flat and the other isn't, then the fringes will move as the surfaces move in relation to one another, Imagine a sheet of corrugated iron on a flat surface, with the corrugations representing the fringes, now rotate the sheet of iron, did the fringes rotate as well.... :)

Regards
Ray

[Pete F]

No sorry Ray, that analogy definitely isn't valid. The fringes are a product of an imperfect surface, they are not the imperfect surface. I do like your analogy however, so imagine a sheet of corrugated iron and a sheet of flat frosted glass. Look through the frosted glass and where the two surfaces touch you will see the corrugated iron. They are the fringes. Now rotate the frosted glass, the pattern doesn't change. If on the other hand the frosted glass was also corrugated the pattern would change.

Thinking about it some more, if this is the addition and subtraction of wavelengths of light wouldn't the fringes surely be a half a wavelength and not a full wavelength?

Pete

[RayG]

No sorry Ray, that analogy definitely isn't valid. The fringes are a product of an imperfect surface, they are not the imperfect surface. I do like your analogy however, so imagine a sheet of corrugated iron and a sheet of flat frosted glass. Look through the frosted glass and where the two surfaces touch you will see the corrugated iron. They are the fringes. Now rotate the frosted glass, the pattern doesn't change. If on the other hand the frosted glass was also corrugated the pattern would change.

Thinking about it some more, if this is the addition and subtraction of wavelengths of light wouldn't the fringes surely be a half a wavelength and not a full wavelength?

Pete

Hi Pete,
Well to be pedantic, odd multiples of 1/2 wavelength will be dark, and even multiples of 1/2 wavelength will be bright..

I see your point about rotating a perfectly flat surface, that might be an innovative test for perfect flatness in itself.... I'd be patenting that.... :)

Regards
Ray

Here is an image from Wikipedia
http://upload.wikimedia.org/wikipedia/commons/thumb/4/42/Optical_flat_interference.svg/310px-Optical_flat_interference.svg.png

[Michael G]

Pete, the fringes are a result of the air gap between the flats and the refraction of the light. Two perfectly flat glass flats will still produce fringes if one end is jacked up producing an air gap. Rotating one or both of those flats will not change where the fringe pattern is (provided the item producing the air gap (hair, sheet of paper etc) is still in the original position).
However, if one of the flats has a defect such that it is not perfectly flat that deviation from flatness will be seen as a distortion of the fringe pattern, which should be uniform and even if things are perfect.

Michael

[Pete F]

Yes bear in mind Ray I'm only speaking hypothetically in order to understand better how these would work. Many MANY years ago (groan too many now come to think of it) a very wise lecturer said to me "Pete if you want to test something, test it at F^&*-all and infinity. If it proves at both those extremes the chances are it will test everywhere in between!" As a young man fresh out of high-school, hearing a lecturer swear was a memorable event :D

Michael, yes that was precisely my point. Not having used them I could only go on logical deduction from what others said. I've tried to find them on ebay over the years (thought quite why I want optical flats is a complete mystery), but have never found any, or at least none at prices that I would be prepared to pay for!

In the process of all this I found this article, which seems to be terrific in explaining how they're used. Ray, they show how to measure sub-wavelengths, so you're indeed quite right it is possible, albeit I think a little different to your approach. http://www.edmundoptics.com/learning-and-support/technical/learning-center/marketing-literature/files/eo-optical-flat-manual.pdf

I'm not sure why I find metrology so fascinating, it should be as boring as bat-$%^& ... well, maybe that IS the reason :D

Pete

Edit: Ray I didn't see that image you posted when I replied. Yes that is why the fringes will be 1/2 a wavelength apart.

[RayG]

Hi Pete,

I did a few years of pure maths (too many, as I recall), and derive some degree of rigor in logical thought processes from that background, and if you could prove a hypothesis failed at some extreme boundary condition then the hypothesis failed.

But the real world isn't like that at all, there are things that are true within the bounds of normal experience.

Best summed up with this quote...

In theory, theory and practice are the same, in practice they are not.

Still, I agree metrology is fascinating stuff, can't see them making it into a reality TV show however...:D


Regards
Ray

[Pete F]

Still, I agree metrology is fascinating stuff, can't see them making it into a reality TV show however...:D

I don't know mate, I honestly don't really watch TV*, but from what I've seen it would surely have to be as exciting as much of the crap they put to air :p

Pete

*Excepting Grand Designs, of which I have the entire DVD set anyway