Using Optical Flat to check Flatness of Granite Surface Plate

[electrosteam]

Hi all,
I have just acquired (at the right price) a 3 inch optical flat about 1/2 inch thick.
I remember from university physics laboratory days using a sodium light (I think) to check the flatness of a surface by observing interference fringes.

Is this approach workable for a granite surface plate ?
If it is, what is the work method and suggestions on a source of a suitable monochromatic light source?

So far, all I have tried was a squirt of WD40, the optical flat floats effortlessly across the surface.

Getting into totally unknown territory is fun !

John.

[Techo1]

Normally the optical flat is larger than the part being checked so unless you have a very small surface plate it will only give a general idea of flatness in the area it covers.

Lex.

[Michael G]

I'm not sure it will be - the fringes are down at the wavelength of light level and the granite surface plate won't be that flat.

Michael

[Pete F]

I'm curious as to what others use for light sources too. I just use my regular workshop lighting but the fringes are a bit hard to see. I think a monochromatic source would be better, but I've never really looked in to small light sources and whether the room would then also need to be darkened etc etc. I seem to recall reading somewhere about using a diffuser on a lasers? Lasers are a dime a dozen these days so I'd be interested if there was any truth in that

[electrosteam]

Pete,
You hinted that you have done something like this.
Can you give us any details - granite polishing, size of the flat, any interposing fluid, angle of light incidence etc.
John.

[simonl]

it's very blurry but I have some memories of doing the same tests. The interference fringes are called Newtons Rings (I think) and we used a light, clear oil between the flat and the object. I can't remember out light source but we did do a lot of work with 4mW HeNe lasers at the time. I'm not sure if indeed you need a laser or just a monochromatic light source. In terms of usefulness with granite surface plates, I think it may give you localised flatness but not overall since it's not big enough to cover the plate. Also you would need to work out the calculations so that you can measure the fringe separation (and variation of) and relate that to flatness in terms of uM.

Cheers,

Simon

[Pete F]

Pete,
You hinted that you have done something like this.
Can you give us any details - granite polishing, size of the flat, any interposing fluid, angle of light incidence etc.
John.

Such good questions John, I hope you're therefore prepared to be let down with some really crap answers, as I can't answer most if them :D

I've only ever used an optical flat on lapped surfaces and I haven't ever dropped my optical flat on either of my granite plates, so don't know how they go there. Oh gawd I shouldn't even joke about dropping it anywhere, they're not cheap! My guess is that neither would have a surface texture fine enough to allow the optical flat to work properly, but I can't say for sure. I'm away at present, but I'll give it a go when I can.

I'm trying to picture the size of my flat and I'd say it was around 60 mm(?) I haven't had it out for a while and it's not exactly something I would wear out, so I could be off with that size. I was repairing and servicing about 20 micrometers a year or so ago, and was mainly using it then to check their faces.

No idea of the angle if incidence etc. I have fluorescent lights in my workshop, I can't recall what colour temperature they are, but the fringes appear just fine. As Michael suggested, the distances we're talking here are weeny, so it doesn't take much to upset the apple cart. As I mentioned in my post above, the fringes don't exactly leap up and punch me in the nose, and I think that could be improved with a more determined light source. If it's something easy to do and is small to store with the flat I will however invest the time. My idea was to get a laser pointer and diffuse the light, but NFI if that would work as I hope. Unfortunately I'm very busy when I'm in the workshop, as I have one job in particular that needs to be done by a specific date, however if I get a chance I'll try to pull it out and better answer your questions.

[RayG]

The guys at the National Measurement Institute that we did a group visit last year have a special room for optical flats, and I recall the choice of light source was one of the critical factors in calibrating the flats.. Even to the extent that a blown globe, requires them to re-calibrate the system for the replacement globe.

The accuracy of the measurement between fringes depends on knowing the wavelength.

In answer to the original question, optical flats are pretty much useless for surface plate calibration.

Differential levels or laser interferometer are the preferred methods.

Ray

[jhovel]

I seem to recall from the standards lab visit that it required a Sodium vapor light source to be a meaningful tool - but of course I could be quite wrong...

[Steamwhisperer]

Hi Joe,
They needed the Sodium Vapour light source to be what.?? Lol

Phil

[simonl]

I seem gto recall from the standards lab that required a Sodium vapor light source to be m eaningfuk That's Gold! I dare say that Joe may have been on his phone when posting! It beats my effort of saying lactating instead of loctiting.... Simon

[Oldneweng]

Hi Joe,
They needed the Sodium Vapour light source to be what.?? Lol

Phil

I would say that Joe should sack his proofreader. :D

Dean

[Michael G]

Sodium is the preferred light source because the two frequencies the light comes out at are close together and there are only 2. To accurately count fringes you need to be as close to monochromatic as you can otherwise there is interference in unwanted places.

Michael

[Pete F]

The guys at the National Measurement Institute that we did a group visit last year have a special room for optical flats, and I recall the choice of light source was one of the critical factors in calibrating the flats.. Even to the extent that a blown globe, requires them to re-calibrate the system for the replacement globe.

The accuracy of the measurement between fringes depends on knowing the wavelength.

In answer to the original question, optical flats are pretty much useless for surface plate calibration.

Differential levels or laser interferometer are the preferred methods.

Ray

That would certainly make sense for some of the work done by the lab, as the level of precision available with optics is quite extraordinary. It is however far greater than the required level of precision required by any machine tool, so I dare say we would be free to use any convenient light source without any real concern as to how it might affect accuracy.

I agree, even if you could get some fringing on the plate (and I have my doubts it will be flat enough, but we will see) I can't see much point. In addition to the above methods, I believe autocollimation was the main way to check plates in the past, however that has largely been surpassed with (relatively!) cheap differential levels that are much easier and faster to use, and can input directly into a PC/software to semi-automate the task to produce the map. Autocollimators can be purchased second hand for "reasonable" money, but are rarely found with mirrors. Indeed I have no idea what happens to all the world's optical tooling mirrors, as they're seen for sale relatively infrequently, yet are an essential component in optical tooling if this type. I presume a cleaner somewhere is going to open a cupboard to find 35 tonnes of mirrors come crashing down around him! While it's possible to collimate off other surfaces it would be quite difficult to check the condition of a surface plate without proper mirrors. I will buy two first surface mirrors and make up cubes to mount them on for my autocollimator, just the same it's not a trivial task, and not something I've yet had the time to do.

For your own optical flat John, in addition to checking the condition of micrometer faces, they work well with any lapped surface, so it can be handy for checking gauge blocks for any burrs or crap that might have prevented them from wringing. Just the same, it's unlikely to be something you will wear out!

Pete

[Techo1]

I can answer a couple of your questions John, first of all the surface of the optical flat and the part must be clean and dry to produce a slight air gap ie. no interposing fluid, secondly the surface of the part needs to be slightly reflective so a granite plate may or may not show any light bands.

Lex.

[Pete F]

Sodium is the preferred light source because the two frequencies the light comes out at are close together and there are only 2. To accurately count fringes you need to be as close to monochromatic as you can otherwise there is interference in unwanted places.

Michael

Michael I'm curious as to what you mean by interference in unwanted places? While I'm aware of the process, I've never used my flat for anything other than surface condition (ie "flat"/concave/convex, depending on the fringe pattern), and assume you're talking about absolute measurement circumstances?

[Michael G]

Not a very precise comment on my part really - but if the spectrum bands are separated by too much distance or there are too many of them they could interfere with the other light frequencies so that clear well defined fringes are not as easy to get. Sodium is the common (cheap) solution.

Michael

[jhovel]

Peter,
have a look at
interferometer | eBay (http://www.ebay.com/sch/Electrical-Test-Equipment-/92074/i.html?_sop=15&_nkw=interferometer&_frs=1)
and see what goodies you might find there.....

[Pete F]

Peter,
have a look at
interferometer | eBay (http://www.ebay.com/sch/Electrical-Test-Equipment-/92074/i.html?_sop=15&_nkw=interferometer&_frs=1)
and see what goodies you might find there.....

Nothing there Joe. These are the mirrors used fir autocollimation

Nikon | Instruments Products | Industrial : Autocollimator Accessories (http://www.nikon.com/products/instruments/lineup/industrial/autocollimator/accessories/autocollimator_accessories/index.htm)

My original intention was to scrape a cube and attach a first surface mirror to that. By flipping the cube and autocollimating on it I could auto-generate a precise 90 degree surface. However I'm also considering instead making it an adjustable 3 point stand, with one if the points a very sensitive micrometer arrangement so I can take my angular readings at the mirror end instead of the autocollimator end as is normally done. I have a very good reason for wanting that, but need to rush to work so can't explain now.

pete

[Pete F]

Hi all,
I have just acquired (at the right price) a 3 inch optical flat about 1/2 inch thick.
I remember from university physics laboratory days using a sodium light (I think) to check the flatness of a surface by observing interference fringes.

Is this approach workable for a granite surface plate ?
If it is, what is the work method and suggestions on a source of a suitable monochromatic light source?

So far, all I have tried was a squirt of WD40, the optical flat floats effortlessly across the surface.

Getting into totally unknown territory is fun !

John.

Sorry for the delayed response, I was just heading off to bed and realised I hadn't checked this as promised so raced down and had a look.

My little Mitutoyo plate is filthy as I use that all the time and I didn't think I'd get much to work off that, so wasn't exactly surprised when I couldn't get fringes. Instead I whipped the cover of the bigger Chinese delight but also couldn't get any fringes. I grabbed a mic and that works just fine without fail, so there wasn't anything lighting wise etc to spoil the party.

While flatness may be an issue, I think a bigger problem may be that the colour and texture of a granite plate means the light isn't reflected very well, and that's something that's essential for the whole interference caper to work. So while floating the optical flat around isn't a problem, seeing anything meaningful with it is.

No great surprises there, but I thought I'd just confirm what we all suspected.