Giant Magellan Telescope Starts Making its Final Mirror

11 thoughts on “Giant Magellan Telescope Starts Making its Final Mirror”

1. What we need is a lot more medium size space telescopes. When SpaceX Superheavy is ready we should put about 20 medium-sized telescopes at about $billion each into orbit to monitor the entire universe for changes. When a new phenomenon occurs we can then point the larger telescopes at it. These telescopes would be able to find supernovas, exoplanets, asteroids, and comets.

2. Wow, be the man who drops that mirror after all that :/
   Maybe somebody has to figure out 3D mirror/glass printing, should be possible…

   • And whoever drops that sucker is gonna have more than 7 years bad luck.

3. Didn’t know of this giant telescope named after the great Portuguese navigator Fernão de Magalhães. Thanks, Brian.

4. I’m curious what are the advantages of larger segments in a multi-segment mirror assembly, given the use of adaptive optics in secondary or tertiary mirrors? Otherwise, why not just mass produce 2m segments and set up a grinding line?

   • It’s a tradeoff between manufacturability and light-gathering potential.

5. “The Giant Magellan Telescope begins the four-year process to fabricate and polish its seventh and final primary mirror”

   4 years? Those things take wayy to much time, it’s ridiculous how long it takes.
   We should be manufacturing such mirrors in a matter of weeks, even days.

   • Sheesh … sez who? You, because you have a 30 year background in extra-large scale glass melting, cooling, annealing and subsequent recrystallization, then figuring?

     You, because the department you work at has a need for a 25 meter telescope, or at least time on one very soon in order to confirm or discover something about Planet X which no one yet has?

     You, because the 3-D printed automatic hobbyist astronomical mirror grinder gets the job done in hours, which formerly took High School kids since oh, the 1920s days to do so in their parents’ back yard?
     ________________________________________

     Melting alone is pretty fast, limited mostly by hoping to not damage the mold which the glass sits in. Following the initial heat, the melt needs to cycled through a half-dozen critical steps to first crystallize the composition, then just as carefully ‘vitrify’ it back to a precisely homogenous state. THEN, after those months have slowly ticked by, then the mirror blank needs to be totally stabilized mechanically by annealing, heating to high temperatures, enough to cause the glass to become more viscous and plastic-like than brittle, and then cooling it a few degrees A DAY down to where few (if any) significant sources of thermal expansion distortion and stress build up. Below some lower critical temperature, the blank can again be relatively quickly cooled. Hundreds of degrees a day.

     Now, a year or year and a half since starting, you have a working blank. It is then roughly milled (within tenths of a millimeter) to the exact figure it needs to become, PLUS a millimeter or two of to-be-ground-away glass on top. This has to go pretty slowly, since the quite smooth diamond conic grinding machines can only proceed a few millimeters a minute on average. And there are a LOT of millimeters around the blank.

     Then the polishing itself. Various grits, again to perfect the figure (shape, mathematically determined) down to sub-µm precision and with a surface roughness that looks almost-but-not-quite transparent. A wee film of roughness.

     After that, some months later, comes the testing-and-final-figuring itself. Testing — outside the cutting-and-grinding machinery feedback loops — actual optical testing to see how closely to the ideal the figure has become. That can take months simply due to having to change, and change, and change-some-more the edge and center forces acting on the mirror (as it will encounter ‘mounted’ in the final telescope). The precision boggles (this metrologist’s) mind. Tens of nanometers, across the whole 8.8 meter blank. Almost exactly the same as a BandAid in relation to the Earth’s crust.

     As if another step is needed, now the mirror must be ‘fine polished and metallized’. All those nanometers ‘have to go’, and the whole thing tested again and again. It is now sparklingly clear, absolutely shiny. Ready to receive its top-coat of whatever metals is deemed provident for the final telescope optical path design. Silver, gold, aluminum, tantalum, niobium and so on. THEN!!! Tested yet again. To ensure that the nanometers of metals haven’t crinkled the thing in any meaningful way.

     Finally, some 3–5 years from the start, the mirror is ready to package and ship. We don’t even want to get into how long that business takes. Very long.

     ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
     ⋅-=≡ GoatGuy ✓ ≡=-⋅

     • You showed him.

     • I would hope that in a decade or two most new telescopes are space based and BIG. At which point might (nasty toxic) beryllium make more sense to shorten manufacturing lead times? Or would if make sense to have a production line of blanks being manufactured even now? perhaps in some low-labour cost country?

     • Wondering your take on the star link satellites’ effect on these ground based telescopes- especially in 5-10 years. Will these huge/expensive ground based telescopes be worth it with the huge number of satellites orbiting earth in the near future? Seems space-based telescopes may be the only way. Thanks!

Comments are closed.