What's new
What's new

Who can do inside surface finishes in single-digit microns?

n2zon

Cast Iron
Joined
Apr 29, 2014
Location
Dutchess County NY
I know someone working on a prototype that requires an aluminum part with a very smooth finish to reflect light.

The part requires six radial, stepped holes with 45* steps in a center-drilled disc. Smallest ID is .250". The surface finish inside the stepped holes likely needs to be in single digit microns, perhaps as fine as 2 microns. The ID tolerance is quite loose (he says +/- .010" is better than he needs). Flatness is not critical, either. I'm waiting on confirmation of the wavelength of the light to be reflected.

He knows it won't be cheap.

I know some of you work in microinches...who can do this sort of work? I'd like to put him in touch.

Personally, I would be fascinated to even know how it could be done!

Edits:

- Correct number of radial holes (6, not 4; my error).
- Add info about flatness.
- Added placeholder for info about wavelength.
 
Last edited:
I think you need to explain the project in better detail. The 2micro surface finish is one thing but your description sounds like it also has optical properties. That would imply that the surface flatness is probably also specified or needs to be. You also need to specify the light frequency that is being reflected.

If this is for some sort of optical device, everything gets to be very complicated and just going for a surface finish is useless. If it is just for cosmetic purposes then, it is much simpler.
 
I think you need to explain the project in better detail. The 2micro surface finish is one thing but your description sounds like it also has optical properties. That would imply that the surface flatness is probably also specified or needs to be. You also need to specify the light frequency that is being reflected.

If this is for some sort of optical device, everything gets to be very complicated and just going for a surface finish is useless. If it is just for cosmetic purposes then, it is much simpler.

This is not just for cosmetics. It's intended to be part of an optical device.

I'll ask him about flatness (clearly it's a curved surface). I asked about wavelength but don't have an answer back yet. My guess is around 10 microns.

Does your shop do this sort of thing?
 
Opticians/lens makers. And yes, aluminum has been used for mirrors.

Telescope makers.

Aluminum can be polished using ordinary, abrasive and polishing compounds. But if it needs to have certain optical properties, if it needs to be part of an image forming system, then there are additional requirements and the specifications for that imaging system will be the determining factor. The overall form of the surface may be a lot more important than the degree of polish given to it.

This sounds like one of those half-baked ideas for a device that does bla-bla-bla. And the customer did not develop a proper set of specifications but expects you to think it out and come up with a working part. It is HIS job to design the part and to provide a proper set of specifications, not yours. I will bet that he does not even have a way of inspecting the part other than assembling it in his device and seeing if it works.

Good luck.
 
Opticians/lens makers. And yes, aluminum has been used for mirrors.

Telescope makers.

Aluminum can be polished using ordinary, abrasive and polishing compounds. But if it needs to have certain optical properties, if it needs to be part of an image forming system, then there are additional requirements and the specifications for that imaging system will be the determining factor. The overall form of the surface may be a lot more important than the degree of polish given to it.

This sounds like one of those half-baked ideas for a device that does bla-bla-bla. And the customer did not develop a proper set of specifications but expects you to think it out and come up with a working part. It is HIS job to design the part and to provide a proper set of specifications, not yours. I will bet that he does not even have a way of inspecting the part other than assembling it in his device and seeing if it works.

Good luck.

I'm not going to try to make the part. For one thing, this is (far) beyond my capabilities.

He has been using a microscope for inspection. He was initially focused on surface finish, but I started to ask questions about other stuff, and he's thinking about the answers. One thing I told him is that some time soon he will need a fully-dimensioned drawing with callouts for material, surface finish, tolerances, flatness, and so on for someone to quote the work. As far as I know, imaging is not involved.

What I'm looking for is a shop that does this kind of stuff and is capable of doing the work.

Some additional info: Flatness is not critical. I'll add this to the OP above.
 
If this is an optical application flatness is very much a key parameter.

In optics that are used for mirrors or reflecting surfaces, even though the surface is curved, the surface needs to be optically flat. If it is not, you end up with several issues that will make the image blurry.

I suspect that the end customer does not have a grasp of what he is after. If all they need is just a polished surface to a specific Ra, then that is very easy to do.
 
you’ve mentioned needing possibly 2 microns or possibly 10 microns. I can’t imagine any optical application using such corse surfaces. It would be shiny but not very reflective. Could you possibly mean micro inches? ( you did mention people here that work in micro inches) If that is what you mean then Conrad Hoffman probably has the best answer, preferably with a very good air bearing spindle.
 
If 10 microns is the light wavelength, infrared, 2 microns finish is not out of reason. In the 60s I made concave mirrors for spectrophotometers from aluminum, but accurate imaging was not that critical because they focused light from a slit on a diffraction grating. I used 1 micron synthetic sapphire from Microabrasives Co. with a pitch lap and got mirrors that performed well but under bright light you could see fine scratches. Finer grits clumped and made wide shallow scratches known as sleeks. My biochemist boss kept telling me how incompetent I was considering that in the early days of astronomy all they had was metal mirrors and they worked so why couldn't I do it? When I finally had all I could stand of his harassment, I quit and toured Europe. In a museum in London I saw a metal mirror used by Lord Rosse with exactly the same scratches. They just put up with them.

I haven't followed developments in recent years, but back then the aluminum mirrors were coated with an alloy called Kanagin (spelling may be wrong) which was harder and polished better.

Without a lot more information, this is really a pointless discussion.

Bill
 
If this is an optical application flatness is very much a key parameter.

In optics that are used for mirrors or reflecting surfaces, even though the surface is curved, the surface needs to be optically flat. If it is not, you end up with several issues that will make the image blurry.

I suspect that the end customer does not have a grasp of what he is after. If all they need is just a polished surface to a specific Ra, then that is very easy to do.

Again, this is not an imaging application....
 
Again, this is not an imaging application....

From what I can understand so far, I'd suggest this be an assembly of ground glass disks, with each thickness being it's own diameter with angled aperture, polished and with an aluminum or silver reflectance coating vapor deposited using a bell jar.

Assembly would be a common housing, with jacking screws at each stack point to adjust the mirrors into alignment. This would allow testing, changes, and retest with minimal pain and bomb tossing between fabricator and client.

I may be able to help, but as EPAIII says a lot of additional information and specification is needed, otherwise it's a problematic venture.
 
As Milland notes, lots more info required. The component performance spec has to accurately drive the piece-part specs, with some optical engineering deeply involved, assuming the finish values already discussed are actually required.

My company has used several machine shops in the Rochester NY area for opto-mechanical components such as lens barrels. Rochester has such expertise in place because of the optics and optical system mfrs that used to be located there, such as Bausch and Lomb, Kodak, etc. Milland may also have very good inputs and recommendations, once the project gets better defined.
 
I suspect what 9100 says about the coating is important simply because unless the unit is in a vacuum, the aluminum will tarnish quickly.

Tom

Aluminum tarnishes almost instantly. By tarnish I mean, oxidize. Fresh aluminum surfaces form a thin (about 5 angstroms) layer of aluminum oxide. The
layer is tough, gas tight, and will re-form whenver the surface is damaged. The reason it stops growing at about 5 angstroms is that's when the layer becomes
completely gas-tight. The surface is self-passivating. An aluminum PVD coating on a mirror will be fully covered with oxide on its surface within a minute
or so, as soon as it is removed from the vacuum chamber where the metal was put on.

Aluminum oxide is extremely tough, non-soluble, and an excellent electrical insulator. The self-passivating property of aluminum films was one of the reasons
it became a huge technological trick when integrated circuits were developed.
 
When you think about it, one micron is about 40 millionths of an inch, not very smooth. I would get it as smooth as I could with a hone or lap, then electropolish it. Astronomical mirrors have aluminum evaporated on them, then coated with silicon monoxide.

Of course, I don't even know why we are worrying about it. One of the most irritating features of these forums is people who come in with a problem, then refuse to furnish enough information to make a meaningful recommendation.

Bill
 
When you think about it, one micron is about 40 millionths of an inch, not very smooth. I would get it as smooth as I could with a hone or lap, then electropolish it. Astronomical mirrors have aluminum evaporated on them, then coated with silicon monoxide.

Of course, I don't even know why we are worrying about it. One of the most irritating features of these forums is people who come in with a problem, then refuse to furnish enough information to make a meaningful recommendation.

Bill


I find it equally irritating that a simple question can not be responded to.

Can anyone provide a 2 micron or better surface roughness on an aluminum id? |Nothing more was asked.
 








 
Back
Top