What's the best glass for mirrors?

[wimvb]

Always on the lookout for new gear, I've been comparing reflectors. It serms that most mirrors are either made from pyrex or quartz. Yesterday I saw this thread with an ad for an altair astro newt with schott glass mirror:

I know that Schott glass is good in a refractor, but what are its benefits in a reflector?

So, what glass is best for a mirror? Pyrex, quartz, Schott or other?

[andrew s]

You want low expansion and stable substrate as it is just holding the coating. So Zerodur (and similar ceramic glass) and quartz are best followed by Pyrex type glass.

Regards Andrew 

[wimvb]

Ok, did some quick homework:

Zerodur: 0.1*10^-6 /K (grade 2) ... 0.007*10^-6 /K (grade 0) (Schott)

Schott B270: 9.4*10^-6 /K, (Schott) Ouch, 100 times more than zerodur.

Quartz: 0.55*10^-6 /K (wikipedia)

Pyrex: 3.3*10^-6 /K, (Schott 8330), considerably more than quartz

All have roughly the same density, so not much difference in weight, for the same mirror thickness.

What about hardness and workability? Do they measure in? Not for the finished product, I would guess. But it would be a cost factor if one type of glass is substantially more difficult to polish into a mirror of high class.

[Astrobits]

The type of glass is irrelevant to the performance of a finished mirror.

The biggest factor in poor performance of a mirror is the temperature difference between the mirror and the ambient air. If the mirror is just a little warmer than the air, thermal plumes of air arise from the mirror and destroy the image. It is not until the whole mirror and surrounding air are in equilibrium do the thermal plumes cease and the mirror performs to it's best. As there is then no thermal difference across the mirror it's thermal expansion properties are of no consequence.

It's a different story if you are making a mirror as  low thermal expansion glass allows a shorter time between polishing and testing stints. Polishing generates heat and higher thermal expansion glass needs longer to cool before testing. Zero expansion glass needs no time to cool before testing so it benefits the mirror maker.

Nigel

[andrew s]

1 hour ago, Astrobits said:

The type of glass is irrelevant to the performance of a finished mirror.

I don't agree with this. The maintenance of the figure of a mirror, in use, depends on the physical properties of the mirror in terms of e.g. Young's modulus and thermal expansion.

It was very common before the introduction of low expansion glass for the figure of a mirror to change during the night by a noticeable amount.

1 hour ago, Astrobits said:

The biggest factor in poor performance of a mirror is the temperature difference between the mirror and the ambient air.

I agree with this but it can be mitigated by various methods.

Regards Andrew

[wimvb]

Thanks Nigel, very informative. I was aware that once a mirror is finished, the material shouldn't affect the optical performance, other than through thermal effects. Cooldown time should be determined by heat capacity, rather than thermal expansion coefficient, assuming same mirror dimensions. Are mirrors of a certain material also generally equally thick? I can imagine thinner mirrors to bend more under their own weight.

I once measured the bending of a semiconductor wafer (slice) under its own weight. The wafer in question was 2" in diameter and only 0.3 mm thick. In all fairness, I used xray diffraction to measure it.

But I can imagine bending being an issue for a high quality mirror.

Btw, I'm not planning on grinding my own mirrors. I'm just being a curious customer.

[davidc135]

Might be a consideration if imaging along with tube etc materials.  David

[Astrobits]

3 hours ago, andrew s said:

I don't agree with this. The maintenance of the figure of a mirror, in use, depends on the physical properties of the mirror in terms of e.g. Young's modulus and thermal expansion.

It was very common before the introduction of low expansion glass for the figure of a mirror to change during the night by a noticeable amount.

I

Regards Andrew

Before the introduction of low expansion glass it was standard for mirrors to be very thick, a situation which continued probably more out of dogma that actual scientific reason, and is still seen today. Because they were very thick they failed to keep up with falling air temperatures and hence thermal plumes appeared in the air above the mirror which caused a degradation of the image during the night. I am not aware of any measurements that showed it was distortion of the mirror due to differential thermal expansion primarily causing this degradation. However, there has been work showing that a 2" thick lump of glass cannot keep up with the air temperature falling steadily, by relatively small amounts, during the night. Hence a differential temperature grew from zero during an observing session and we all know that a telescope has to cool to ambient temperatures before it can give good images. Even those with 1/10th wave Zerodur glass optics. Many professional observatories refrigerate their domes to the expected night-time temperature so the telescope is ready to use as soon as night arrives.

The best cure is to make thin mirrors and, yes, flexibility of the glass then becomes important and they need careful support.

Nigel

[John]

Orion Optics primary mirrors tend to be thinner than some other brands I've noticed. I believe they use Suprax glass from Schott in Germany.

 

[andrew s]

12 minutes ago, Astrobits said:

I am not aware of any measurements that showed it was distortion of the mirror due to differential thermal expansion primarily causing this degradation.

Nigel, I think you are confusing two different issues.

I agree that the residual thermal lag between mirror and atmospheric temperature causes distortion. This is mainly down to the turbulent boundary layer that can be removed with suitable fans.

However, this is different from a mirror going having a perfect figure to one being under corrected or over corrected as the temperature changes. This was documented for example in Amateur Telescope Makes Handbook(s) where for plate glass mirrors they advised makes to under correct their mirrors so that they would be perfect under falling temperatures. (Or it may have been the other way round as I no longer have the books to check.) 

Regards Andrew

[wimvb]

14 minutes ago, davidc135 said:

Might be a consideration if imaging along with tube etc materials.  David

I've been window shopping (as far as that's possible on the interweb). Imaging scopes that look most interesting are the 10" TS ONTC and 10" Lacerta photo newton at f/5 "-ish". The two glasses I've seen most in these or similar scopes are pyrex and (fewer) quartz. The use of Schott B270 by Altair Astro was the odd one out. Linear expansion of this glass seems at least twice that of pyrex. Mirror thickness is never specified, for some reason. Scopes that I'm interested in all have carbon tubes.

 

[25585]

TS offer quartz mirrors. Questar used to & may still. 

One advantage of thicker mirrors is reduced brittleness. The larger & thinner a mirror, the more prone it is to damage from bumps & jolts. 

I will be interested to compare my 300P DS mirror with my 1980s Oldham/David Hinds 12 inch mirror for edge thickness. The latter has about 2 inch edge thickness and weighs roughly 25lbs. SW not much difference in weight but I did not want to handle or breath on it too much so did not weigh it. 

@John can you give any estimates from your 12" OO mirror?

 

[John]

The OO 12" F/5.3 primary mirror is about 37mm / 1.46 inches thick at the edge and weighs around 5.6 kg / 12.3 lbs.

 

 

 

[25585]

My DH mirror is a lump. It is made from Pyrex. Lasted over 30 years no edge chips or cracks so definitely durable. 

I would probably personally take it to & collect from Orion for recoating though. Interestingly both old & new are almost the same focal length. 

[John]

Is it Oldham Optics or David Hinds ?

They were different companies as I recall

Both produce / produced good mirrors though.

[25585]

It was my Dark Star dob mirror. Best quality option. I am not sure even Dark Star knew when they assembled the scope. Oldham seems to ring a distant bell more.  

[John]

1 hour ago, 25585 said:

It was my Dark Star dob mirror. Best quality option. I am not sure even Dark Star knew when they assembled the scope. Oldham seems to ring a distant bell more.  

I've still got a Dark Star catalogue somewhere. I think they used mirrors from more than one source. David Hinds always produced at least 1/8th wave ones I believe although it was probably plate glass. My Astro Systems 6" F/6 (1st "proper" astro scope) and had DH mirrors in it.

[25585]

5 minutes ago, John said:

I've still got a Dark Star catalogue somewhere. I think they used mirrors from more than one source. David Hinds always produced at least 1/8th wave ones I believe although it was probably plate glass. My Astro Systems 6" F/6 (1st "proper" astro scope) and had DH mirrors in it.

Mine was definitely Pyrex. After the excellent A grade ones in my Fullerscope 8 inch, I insisted & the order took longer as only plate glass were in stock, I have a photo of the mirror's back with its focal length in handwriting on the back. Invoice may be in the loft somewhere. The scope cost me £525, most was probably for the primary. 

[Astrobits]

Andrew, I am not confusing anything.

If I remember correctly the early ATM writers did indeed advocate such under/over correction to counteract the perceived distortion you are concerned about. I also seem to remember that they recommended over correction for viewing in the cooling air of evening and under correction for the warming air of morning depending on when your favourite viewing time was. This assumed that the edge of the mirror reacted to the change in temperatures before the center. Although it appears a logical conclusion based on expectations, I am still not aware that there was any scientific testing done to validate that view.  And what happens during those evenings when the air doesn't cool at the expected rate?

All mirror testing during manufacture is done in a lab environment. i.e. stable air currents and relatively warm ( about 20C ). I am not aware that ANY manufacturer has made under or over corrected mirrors deliberately and if they did they would not be able to quote the deviation from the perfect parabola that they now do routinely.

If you want a mirror made from low expansion glass then they are readily available at a premium to plate/float glass. They are usually thicker and will take longer to cool so, as you mention, you will need fans for dealing with the heat plumes ( aka boundary layer) and to speed up the cooling of the mirror. Compared to a thin float glass mirror you might get a short time of an evening when your thick mirror + fans performs better, but once they have reached thermal equilibrium there will be no difference between them. A thin mirror will follow changes in the ambient air temperature more quickly than the thicker glass so you will need to use the fans on the thicker mirror as required.

In my view using PYREX, ZERODUR, QUARTZ and others as a mirror substrate for most telescope applications is marketing hype to jack up the price. It is only when a certain size is required that we must go to the more exotic glass as float is now only available up to 1" thick which is O.K. to mid teen's diameter. I have a 16" in 1" thick float and have no problems with it's image and one of my 8" x 3/4" thick mirrors out-performed a Meade SCT on Mars many years ago.

Nigel

P.S. Just to throw you another googly, that 8" x 3/4" mirror that out-performed the Meade was mounted with three blobs of Silicone.?

 

[andrew s]

Hi Nigel, we will must have to disagree on this. I would and have always opted for low expansion mirrors. Thin is better but needs excellent mirror support.

Regards Andrew 

[andrew s]

Doing some google searches came up with this 1918 paper which comments on both focal length change with temperature and tube currents. Slightly out of date but it does show it was know to science.

http://adsabs.harvard.edu/full/1918PASP...30...55P

Regards Andrew

[John]

6 minutes ago, andrew s said:

Doing some google searches came up with this 1918 paper which comments on both focal length change with temperature and tube currents. Slightly out of date but it does show it was know to science.

http://adsabs.harvard.edu/full/1918PASP...30...55P

Regards Andrew

"slightly out of date" ?????

Probably best to agree to differ I reckon

[rl]

To put some real numbers on this, the free program GuiPlop gives some interesting (sometimes scary) results.

http://www.davidlewistoronto.com/guiplop_3d.htm

This program gives you a theoretical mirror profile after inputting scope tilt angle and cell format. 

Any large mirror of realistic thickness claimed to be 1/10 wave p-v on the wavefront depends very much on the  supports to maintain the profile as the scope goes towards the horizontal. 

I opened up my OO CT8 and didn't like the look of the 9-point cell..the distribution looked intuitivley wrong. So I measured what they had made and compared it with the modelling results....pretty much perfect agreement!

Note this program gives you the choice of glass and so builds in the material properties, and the dimensions refer to the mirror, not the wavefront errors, so double them up.

Over the years OO have moved from a pyrex 12" mirror 19mm thick on a 3-point suspension to a pyrex 12" mirror 37mm thick on a 9-point suspension...I've owned both and there is all the difference in the world in their performance. 

If you are into Newts it is worth a bored hour on a cloudy night just playing around with the program doing some speculative designs...RL