Building the Giant Magellan Telescope

On a scorching summer day in Tucson, Arizona, I decided to take a trip to a furnace. Ordinarily, I might have eschewed such a proposal, but 24 August 2013 was not any ordinary 106 °F day. It was casting day at the University of Arizona Mirror Lab, where the third of seven 8.4-meter mirrors was being spun and cast to become part of the Giant Magellan Telescope (GMT).

When this optical and near-IR telescope takes first light in 2019, it will be the world’s largest and most powerful cosmological instrument, with a resolving power 10 times greater than that of the Hubble Space Telescope, 368 square meters of light-collecting area, and the ability to produce images so sharp that you could to spot a dime 200 miles away.

Inside the extremely loud and cavernous Mirror Lab, tucked under a corner of the UA football stadium, a giant furnace filled with 20 tons of glass and alumina-silica cores was whirling around at 1165 °C. Spinning the glass produces the optimal shape for a telescope mirror, and the cores, which are hexagonal columns, help create its skeleton. As the glass liquefied, it flowed between the columns, producing a sandwich with a layer of glass on the bottom and top about 5 cm thick, and vertical walls between the cores about 2.5 cm thick.

The resulting structure of the mirror, which resembles a honeycomb, is substantially lighter and more easily controlled thermally than a solid piece of glass is. Those two factors are important to the astronomers who will ultimately harness the telescope’s power to better understand our universe.

The GMT will be located at the Las Campanas Observatory in the Chilean Andes near the Atacama Desert. At an altitude of approximately 2500 m and at least two hours from the nearest town, the site is almost completely untouched by human light pollution. The 22-story high, 1100-ton telescope will be able to pierce the night sky to cosmic depths never before reached.

“All the things we surmise from theoretical simulations that must have happened early on [in the universe]—we’ll actually witness them directly for the first time,” explains Wendy Freedman, chair of the board of directors of the Giant Magellan Telescope Organization (GMTO) and Crawford H. Greenewalt Chair and Director of the Carnegie Observatories. “There’s an epoch that astronomers refer to as the ‘dark ages’ because we just don’t know anything about it, and the GMT will have the opportunity to open up that window.”

Subjects like dark energy, dark matter, and star formation are items on some astronomers’ “to-do” lists when they get a crack at the instrument. Other astronomers will image exoplanets. The telescope and its instruments will be able to provide more detailed data concerning the light and chemical makeup of distant worlds than was previously possible.

Not surprisingly, the GMT project is costly and expansive. The initial bill is $500 million, but once in operation for 15 years, the eventual cost could reach $1 billion. Currently, the endeavor has garnered 10 international academic partners, including UA, Carnegie Institution for Science, Smithsonian Institution, Harvard University, the University of Chicago, the Australian National University, the Korea Astronomy and Space Science Institute, Astronomy Australia Ltd., Texas A&M University and the University of Texas at Austin.

The financial and technological collaboration extends to the private and public sectors as well, with business allies providing strategic components of the project. The mirrors are the largest element. Crafted by the UA Mirror Lab, a world-famous facility that has been producing mirrors for telescopes around the world since the mid-1980s, the GMT mirrors presented an innovation challenge to the team. They are designed to work together as one parabolic-shaped mirror, with six outer mirrors surrounding a central mirror in a flower shape. The outer mirrors are asymmetrically shaped—off-axis—similar to the curve of a potato chip.

“We thought the biggest issue was demonstrating we could do an off-axis mirror like this, which was incredibly challenging,” says Jeff Kingsley, associate director of Steward Observatory, the UA’s astronomy research division, and a senior member of the Mirror Lab. “Since that was the highest-risk part of the project, [the team] thought they should get that started early.” And, in fact, initial discussion and planning of the project’s technical requirements date to the early 2000s. The difficulty of casting any mirror lies in its size and its deviation from a sphere, he adds. “Most mirrors deviate from a sphere by microns, but this deviates from a sphere by about 14 mm. That’s a huge difference and it’s considerably more difficult than anything that the Mirror Lab or anybody else has really done.”

On that blisteringly hot day, I had the opportunity to observe only one step in the development of this amazing astronomical apparatus. The process is complicated, and it takes years to produce and finish just one mirror. Part of that time requirement is devoted to sourcing glass, says Brion Hoffman, president of Ohara Corporation USA , the company that has been providing glass for mirrors at the Lab almost since it started spinning them. It takes 18 months to manufacture enough glass for one mirror.

The E6 Ohara low-expansion borosilicate glass is produced in Japan in 15.24-cm chunks, each of which is hand-inspected using a polariscope to ensure the absence of internal fissures. According to Hoffman, the type of glass is ideally suited for use in the spin-cast process because it melts and flows together smoothly and uniformly without trapping bubbles that could otherwise cause stress fractures. The supply chain demands continuity. As I witnessed the glass spinning for Mirror 3, Ohara was preparing to ship its newly manufactured glass for Mirror 4.

It takes one week to melt all the glass, and the cooling process that follows involves several cycles of slowly reduced temperature. “We want a nice stress-free piece of glass where we’re finished,” says Kingsley. His team will drop the temperature as slowly as about 0.1 °C/hour over the next few months to reach that goal.

After the mirror finishes cooling, the technicians and engineers remove the lid and sides of the tub walls, bond a lifting fixture to the front surface of mirror to remove it from the furnace, and then use water jets to remove any refractories still remaining. The jet gun is another innovation customized in-house. It was developed in collaboration with a mining student from the UA College of Engineering.

It will take another year to do rear-surface processes, which include edging, polishing, and loading those devices on the back that will serve as the mirror’s support system. “Then we lift it, flip it over, and set it in a polishing cell,” says Kingsley. Completing front-surface polishing can take an additional year.

There are many other GMT business partners. Rex Materials Corporation , based in Michigan, produces the high-temperature refractory molds between which the melted glass oozes. Several companies are now bidding to design and build the observatory. Access roads and other vital summit support infrastructure will also be sourced in the near future, says Keith Raybould, GMT project manager.

And then, of course, there’s the instrumentation that will be used with the mirrors. Astronomers at the partner institutions are developing these “The telescope isn’t much use unless you can collect light on the other end and analyze the signal,” jokes Freedman. “One of the instruments that is being planned now is a very wide-field spectrograph that will let you take hundreds, if not thousands, of spectra at the same time.”

Another device, employing a high-resolution spectrograph, will be used to calculate radial velocities of stars and orbiting exoplanets, as well as the chemistry of the earliest stars that formed in the Milky Way. In the years to come, the groups will also create an IR spectrograph and an optical fiber spectrograph, as well as an adaptive optics component using a “constellation” of sodium lasers that will correct for the turbulence in Earth’s atmosphere.

That evening, after I had emerged from the UA mirror foundry, I attended a gala dinner in celebration of the GMT’s third mirror milestone. My parting gift? A piece of Ohara etched glass from the same batch used for the telescope mirror I had seen being cast. As a representation of an international endeavor that heralds both humanity’s technical achievements and its quest to learn more about its cosmic surroundings, what else would do?

Alaina G. Levine is a science and engineering writer, career consultant, and professional speaker and comedian. Networking for Nerds, her new book on networking strategies for scientists and engineers, will be published by Wiley later this year. She can be reached through her website or on Twitter at @AlainaGLevine .