CNC telescope mirror grinder.

[ynneb]

Recently I bought a new telescope for my son. It has a 12 inch diameter mirror.
While the telescope looks expensive because of its size (6 feet long), essentially its a tin can with a big mirror and a few lenses.
The most expensive part of the telescope is the mirror. It probably makes 80% of the cost.

I am told these mirrors so ground accurate that even cleaning the dust off then should be kept to a minimum so as to avoid damaging their tolerances.
I am also told that they must be ground by hand by counting and using a certain pattern.

I am wondering is they could be ground using CNC methods, and at the last stages be ground using the hand method.

Does any one know much about this process, or know of any good links ?

[InspirationTool]

My understanding is that there is such thing as CNC machines for mirrors. The good ones have a built in tester. This would be *extremely* difficult to build a fully automated one.

Google mirror-o-matic for references on automated, amateur built, but non-CNC machines. The mirror is then tested, and re-figured. Lather, rinse, repeat.

-Jeff

[HuFlungDung]

Rough and fine grinding the mirror is about half the work. Polishing is the other half. Testing and figuring makes the third half

I've ground a few by hand, not to say I know anything about machine grinding. Machines are definitely built for the job, but from what I have read, the repetitive nature of machine motions actually works against you when trying to polish a perfect sphere. A slight amount of random motion is eminently desirable to prevent zoning of the surface. Zoning arises during the polishing stages, and has to do with the spacing of channels in the pitch lap. A certain stroke length repeated exactly can result in certain areas of the surface being polished more than others, which makes a wavy surface, one that is quite difficult to restore to a smooth one.

The conversion of the spherical surface to the parabaloid will not fix smoothness problems in the original sphere.

So the short of it is this: build a machine, but plan in some kind of randomizer. Some method of stroke length that varies randomly, plus or minus 1/2" and everything in between. Rotation of the mirror and tool should also be randomized, ie., accurately indexing the two is not desirable. You could, with cnc, create a program with randomized movements, but even that might be a bit of a feat, because what the human mind believes is random usually ends up being quite systematic.

[Kipper]

I had this link posted to me by a fellow astronut (I only have a small one ) I was fancying grinding my own but time has a way of running away and at present i'm a little too busy to even play on the net http://www.sidewalkastronomers.com/index.html
take a look there and theres quite a few tutorials on the net.

[ViperTX]

Well I done the grinding using a pyrex blank and a sacrificial piece.....you rub 2 pieces of glass together in a W pattern....with grinding grit between them....changing the value of the grit until you reach the cerium oxide (I believe that was it....might be confusing my jewelery stuff with the mirror grinding stuff).

There was also a step where we used pitch to create a pattern on the polishing blank.....hmmmmmm.....

Did this back in the early 60's.....silvered the mirror.....that was back before you could get them aluminized. The largest was an 8 inch diameter.....takes a gorrilla or rub 2 larger pieces of glass with a slury mix between them...

Today, it's much easier to pay $200 and get a complete mirror set in the 8 to 10 inch reflector range.

Ft Davis Observatory and someone else did a 100 inch mirror.....they had a somewhat automatic grinder.....some university in Arizona....I believe still provides the grinding service for those huge mirrors.

[bowentw]

If you want to see a really big mirror grinding tool; here is a link

http://mirrorlab.as.arizona.edu/MISC.php?navi=histo