Is there a faster way to fabricate large telescope mirrors?

[mapreader]

To mix, grind and polish a mirror for a major telescope often takes years, judging by cases like the GMT project. Are there any emerging technologies that would reduce the time, cost and effort to make large aperture mirrors?

I think having such a long period of time between commissioning and first light is the cause of most of the problems that plague big telescope projects. While the mirrors are slowly settling in the blanks, funding can dry up, competing projects emerge etc. I think only the first of the Giant Magellan Telescope's eight mirrors has been cast. At this rate, it probably won't see first light within the decade, if ever. Added to that it now has to compete with the Thirty Meter Telescope (which has major funding and scheduling problems of it's own) and the EELT (which has been whittled down from its grand 100-meter original OWL concept to 39 meters last I checked). I wouldn't be surprised if the EELT, which is relying on several European governments for funding, doesn't get built at all due to the financial mess of the Eurozone. Even if it did, the time it would take to fabricate all those mirror segments puts first light depressingly far away.

I'd be interested to know if there are any new technologies being explored to shorten the time and reduce the cost of this process.

[antoniseb]

I don't know of any revolutionary new ideas. The basics improvements right now include making the glass thinner, so it is faster to anneal and cheaper to maneuver.
I suppose that eventually nanotech will give us better grinding agents, which might speed things up, but I haven't heard of anything happening right now on this front.

[jfribrg]

I don't know much about how the mirrors are formed. Are they poured into a giant hockey puck shape and then ground into a parabolic shape later or are they poured into a parabolic mold to begin with? I know that there is no substitute for the slow annealing process, but my understanding is that most of the time is spent grinding the shape. What about rotating the glass as it is poured and annealed? This will give it a parabolic shape, although I don't know if you can control the focus this way.

[mapreader]

@jfribrg

The way they do it at Steward Observatory Mirror Laboratory is detailed here:

http://www.optics.arizona.edu/opti696/Sp%202007/TermPapers/Spin%20Casting%20Mirrors%20by%20T%20Zobrist.pdf

They heat the glass in a rotating oven to give it a parabolic shape.

Edit: Wrong link. I've corrected it.

[selden]

Most large mirrors are pre-formed to the approximate shape; otherwise it would take far too long to produce them. There have been attempts at innovative methods for producing the initial shape, like spinning the material so that the surface naturally forms a concave surface, but the final process of producing an optical-quality surface is still a very tedious grinding. Spinning a "dish" full of a reflective liquid like mercury to act as a mirror is one technique that's been tried, but it can't be used to view things that are far from the zenith. Simply machining the surface can work if you don't need it to be optically perfect. Segmented mirrors are effective in many circumstances. They can be produced more quickly and less expensively since their components are relatively small and several can be polished at the same time. Still, a single, large, continuous, reflective surface has many benefits which justify the expense in time and money.

[swampyankee]

There is the use of liquid metal mirrors. (http://science.nasa.gov/science-news/science-at-nasa/2008/09oct_liquidmirror/), but they have the disadvantage that they can't be pointed away from the zenith.

[parejkoj]

The University of Arizona has one of the best mirror fabrication plants in the world. They spin-cast mirrors up to 8.4m diameter in a ~1000 degree Celcius furnace. Here's a video of the furnace spinning the LSST mirror (http://www.youtube.com/watch?v=t2TnsPhjbmM), and another talking more about the telescope (http://www.youtube.com/watch?v=XIy5Xo1xalY), with some more video of the mirror lab.

The trick is then cooling it slowly enough to prevent it from cracking. This takes time, as even the best current mirrors still have quite a bit of thermal mass. Not much way around that--we've already drastically reduced mirror mass by using honeycomb designs and fancy materials, and if the mirror is too light, it will be too hard to control its sag.

[profloater]

The University of Arizona has one of the best mirror fabrication plants in the world. They spin-cast mirrors up to 8.4m diameter in a ~1000 degree Celcius furnace. Here's a video of the furnace spinning the LSST mirror (http://www.youtube.com/watch?v=t2TnsPhjbmM), and another talking more about the telescope (http://www.youtube.com/watch?v=XIy5Xo1xalY), with some more video of the mirror lab.

The trick is then cooling it slowly enough to prevent it from cracking. This takes time, as even the best current mirrors still have quite a bit of thermal mass. Not much way around that--we've already drastically reduced mirror mass by using honeycomb designs and fancy materials, and if the mirror is too light, it will be too hard to control its sag. This is also a very old way of making large mirrors with molten alloys. You can make excellent large mirrors with aluminised mylar on a frame, the trick is to heat the mylar with a hair drier to stretch it tight. Unfortunately suction on the back side produces a spherical rather than parabollic surface. However such mirrors are very light and easy to support.

[antoniseb]

This is also a very old way of making large mirrors with molten alloys. You can make excellent large mirrors with aluminised mylar on a frame, the trick is to heat the mylar with a hair drier to stretch it tight. Unfortunately suction on the back side produces a spherical rather than parabollic surface. However such mirrors are very light and easy to support.

With today's technologies, this can work as a concentrator for sunlight and such, but it is not precise enough of a sphere to do optical imaging. Perhaps in the future when we have more control over the thickness and molecular orientation of the polymers that make up the mylar (or whatever substance).

[profloater]

With today's technologies, this can work as a concentrator for sunlight and such, but it is not precise enough of a sphere to do optical imaging. Perhaps in the future when we have more control over the thickness and molecular orientation of the polymers that make up the mylar (or whatever substance).By way of invention, a plastic film mirror with bonded wires backed by a well chosen array of permanent magnets can be made to deform either at frequency or statically and by using optics with feedback such a mirror can be made to correct initial errors. Modern computing power makes this approach ever more interesting.

[antoniseb]

By way of invention, a plastic film mirror with bonded wires backed by a well chosen array of permanent magnets can be made to deform either at frequency or statically and by using optics with feedback such a mirror can be made to correct initial errors. Modern computing power makes this approach ever more interesting.

I'm not against your idea, I'm just curious to get your estimate as to how long you think it will be before a 4-meter hexagonal segment could be built using this approach? ... or perhaps a 100 meter ground based monolithic mirror?