Why glass for primary mirrors?

[Norlesh (Shane)]

My question is why primary mirrors seem to be always made from glass? Since a mirror eventually needs to be coated with a reflector anyway it's obviously not for any inherent optical qualities. So why aren't more versatile substrates used - I would have thought there would be suitable plastics and resins around that could be lighter, easier to work with (lower temperature casting, chemical solubility for finishing, ...), and just as able to hold there end shape.

So why does everyone still start off with a glass blank to fabricate a mirror?

[mental4astro (Alexander)]

A mirror certainly doesn't need to be made of glass. What is required for the mirror are very specific physical properties. Thermal stability is the first. For photo this is paramount. Dimensional stability when the mirror is tiIted - trickier with larger apertures. Lightweight - don't underestimate the ballast that the weight of glass gives to the engineering of a scope. This last point is purpose governed.

Resins can be an interesting alternative. The weight advantage is significant, particularly for eq mounted instruments. Their greatest problem is dimensional stability from thermal variations. Producing stress free large aperture resin blanks is not easy either. I'm sure that there is R&D going on into these. At the moment I can see cost as the main limiting factor. Glass has so far proved to be the most versatile and predictable at the moment. Borosilicate glass types in particular.

[Satchmo]

I Can add there that the transparency and homogeneity of glass all aid in the monitoring/checking of internal stresses after annealing . It is impossible to use a polarising stress guage ( polarimeter ) on an opaque material.

[julianh72 (Julian)]

The ideal material for making a mirror would be Unobtainium, which has a zero coefficient of thermal expansion (no distortion as the temperature changes), an infinite elastic modulus (no deflection under load), is able to be machined to precision of a small fraction of a nanometre, and can be polished to make a highly reflective surface.

In the absence of a good supply of Unobtainium, Pyrex or other glasses are about the best mirror-making materials available to us. The great reflecting telescopes of the 1700s and 1800s typically had polished "speculum metal" mirrors, but these were quickly superseded once the technology developed to make a metallised glass mirror.
https://en.wikipedia.org/wiki/Specul...and_telescopes

[Norlesh (Shane)]

Thanks for the replies - you make some good points. Guess I'll be sticking with glass for my first mirror while I keep dreaming of a cast aerogel substrate with spun primary surface as the follow up

[LewisM]

I had always thought why you could not make the blank from aircraft grade aluminium, highly polished, and then coated and overcoated. Lightweight, thermal expansion should be acceptable (I would think).

Or are there simply some substrates that do not work well with silver coating? I suppose the aluminium would oxidise over time, but probably less than the silver anyway. Anodising would slow this somewhat.

Acrylic is another I would have thought stable and suitable, or perspex. (that does stabilise well at extremes of temperature). Acrylic HAS been used in refractors (cheap and nasty ones), the downside being UV crazing and fogging. Wouldn't matter so much with a mirror.

Interesting ideas. I guess we just traditionally stick to glasses.

[bojan]

Some info on materials can be found here:
http://www.engineeringtoolbox.com/li...ents-d_95.html
http://www.engineeringtoolbox.com/th...ity-d_429.html


Material - thermal coefficient [(10-6 m/(m K))] thermal conductivity [W/(m K)]

Aluminum 22.2 205
Glass, Pyrex 4.0 1.05

So.. in principle glass is 5x better than aluminium in terms of expansion coefficient, but aluminium is 200x better in terms of conductivity - so less internal stresses, due to uniform thermal gradients inside material.

I imagine glass is much easier to figure and polish compared to aluminium... or not?

I think the homogenity of glass is definitelly better controlled.. and there is no issues with chemical stability of glass at melting temperatures.. while aluminium may even burn in normally oxigenated atmosphere.

[mental4astro (Alexander)]

Acrylic is not totally chemically stable. The manufacturing process sees an incomplete chemical reaction that leaves a lot of free monomer in the substrate - part of the reason for UV crazing (monomer is the single molecule that then forms the chains in a POLYmer). Monomer leaching is common in lesser quality acrylic. It also will release some of this free monomer in a vacuum situation, contaminating the metal coating process. I don't know about the contact temperature of the vaporized metal during the coating process. If it is high (over 70degrees C) this will affect the surface quality. It is also excessively soft to maintain a reliable figure in a large aperture. Would be a ba***rd to figure too, being so soft. This would be the same for most composite materials too.

[Peter.M]

http://www.laserfocusworld.com/artic...se-weight.html


There are interesting developments out there for mirror substrates. If this technology gets off the ground large mirrored scopes could be mounted on far smaller mounts.

[Peter.M]

Originally Posted by bojan

Some info on materials can be found here:
http://www.engineeringtoolbox.com/li...ents-d_95.html
http://www.engineeringtoolbox.com/th...ity-d_429.html


Material - thermal coefficient [(10-6 m/(m K))] thermal conductivity [W/(m K)]

Aluminum 22.2 205
Glass, Pyrex 4.0 1.05

So.. in principle aluminium is 5x better than glass in terms of expansion coefficient, but aluminium is 200x better in terms of conductivity - so less internal stresses, due to uniform thermal gradients inside material.

I imagine glass is much easier to figure and polish compared to aluminium... or not?

I think the homogenity of glass is definitelly better controlled.. and there is no issues with chemical stability of glass at melting temperatures.. while aluminium may even burn in normally oxigenated atmosphere.

aluminium is 5x WORSE than glass. You want the mirror to not expand

[bojan]

Originally Posted by Peter.M

aluminium is 5x WORSE than glass. You want the mirror to not expand

Yep sorry, typo..
Corrected.


Another point in favour of borosillicate glass:

I have Coulter Optics 10" F/5.6 parabolic mirror, I purchased it back in 1980... it was never been re-coated, I washed it 3 times so far, and apart from small damage on Al coating (when the secondary fell on it during mishap 6 years ag0), it is like new.

I would like to see plastic mirror that old.. and performing as well as this one

[LewisM]

Hmmm, well there goes aluminium.

How about ceramics?

[bojan]

Alumina is good.. 5.4.
However it is very expensive.

Fused sillica is excellent (0.55). It is used for high end telescope mirrors..

[AstralTraveller (David)]

Back in the 70's I knew a bloke who experimented with plastic mirrors. I'm not sure what type of plastic, I never thought to ask. He claimed to be able to figure them and polish them but they wouldn't hold figure. The plastic was hygroscopic and expanded and contracted with changing humidity. The other issue of course is that the very softness that made them very quick and easy to figure would make them susceptable to cleaning accidents. The solution might be a protective overcoat but he couldn't pursue that idea. In hindight I wonder what you would gain in the end?

BTW, I can confirm that polypropylene is significantly hygroscopic. We sometimes determine the carbonate content of soils by weight loss after acid digestion. The samples are processed in PP centrifuge tubes and we have to ensure that the tubes have the same moisture content at each weighting (dry in oven, cool in desicator, weigh immediately after removing from desicator).

[Bassnut (Fred)]

I think the best solution for a cheap mirror is spinning liquid mercury. As a largish mirror, its 1% the cost of a glass one. The mercury only needs to be a mm or so thickness on top of a rough form work. Mercury is about $200 per kg. I dont know, perhaps you need 2 kg for a 20" mirror?.

It needs a motor rotating a few revs per min to maintain a parabolic shape, thats not hard, it seems to be a no brainer really.

A very small, trivial problem is it can only point to the zenith, but you can pan in post with a large chip cam, so thats covered. The kind of monster apature you can afford with this kind of mirror would make exposures pretty short anyway.

[Steffen]

Aren't Cervit, Zerodur and Astrositall all ceramics?

Originally Posted by Bassnut

Mercury is about $200 per kg. I dont know, perhaps you need 2 kg for a 20" mirror?

One kg of mercury is a ball of about 5cm diameter, do you reckon two of those will be enough for a 20" mirror?

[Rac (Raymond)]

A friend of mine has 10L of Mercury. Neat stuff to play with.

[AstralTraveller (David)]

Originally Posted by Rac

A friend of mine has 10L of Mercury. Neat stuff to play with.

I hope he has it stored properly. It's a bugger to clean up if it spills. We have ours in 1l bottles, about half full and re-enforced with tape. All the bottles sit in a spill tray. I'm always amazed at how heavy it is.

[traveller (Bo)]

Originally Posted by LewisM

Hmmm, well there goes aluminium.

How about ceramics?

Ancient civilisations used polished bronze as mirrors, but not quite for astronomical purposes.

[pixelsaurus (Mike)]

Just for a laugh, back in the early 80's a friend and I decided to make an epoxy mirror. As the epoxy hardened we spun up the disk on an old record turntable. We made a 150mm mirror at about f/2. The surface was rubbish but given the constraints of the material, we concluded that the process might show some promise in the future.