GMT4 Unveil!

About exactly 3 months ago, I posted about the latest high-profile event at the Mirror Lab - the renaming of the Lab to the Richard F. Caris Mirror Lab, and the latest casting for the Giant Magellan Telescope - GMT4! Well, the baking and cooling is completed, and yesterday the casting crew started disassembly of the oven revealing the latest mirror substrate for the first time. Of course, we'd been watching the video image through the high-temperature melting, but watching glass cool, where the anneal schedule has it dropping only .1C per day is worse than watching grass grow! At left is the oven with the lid and one section of the wall removed. With the glass still safely enclosed, the unpacking and mold removal will wait until after the holiday break.


While you would think that casting a disk of molten glass would be straightforward, there are lots of complications learned as the Lab started small, and slowly culminating in these present 8.4 meter mirrors. Between the upper image and this one at left, you can spot the Inconel bands that wrap around the outer tub walls to restrain the centrifugal force of the spinning molten glass. Each band wraps around 180 degrees of the mirror circumference, so there are 4 posts around the oven to ensure an even wrap around the mold. Note also that there are more bands near the bottom of the mirror as the hydrostatic forces of the mass of the glass plus the centrifugal force require more support. At right is shown the top of the cast surface, with marks on the machined edge indicating the thickness of the faceplate - 2 inches in this case. In all these images, the hexagonal shapes of the internal structure can be seen. These mold cores are made of a machined ceramic that are removed after casting to leave air hollows in the substrate, resulting in light-weighting. Also note the electrical resistance heaters in the walls for controlling the internal oven temperature. Heaters are also installed in the floor and top of the oven, controlled by feedback from the thermocouples that poke through some of the holes in the oven walls.

This mirror is GMT4, the 4th cast for the Telescope and is destined for the center position of the mirror array. As a result, it is the only mirror with a cast center hole - and an oversized one at that, 1.5 meters in diameter. All the other substrates are off-axis parabolas, and we're only allowed a small 3" hole in the center to allow drainage of fluid for fabrication stages - generating and polishing. These are machined in with diamond tools rather than cast. A close examination of the left image shows the reflection of the center tub wall off the fire-polished surface. Note that the reflection isn't perfect though! When the glass temperature first froze as the temperature dropped below its softening point, it was a near-perfect surface. However, as the temperature continued to drop from about 800C where that happens, to room temperature, the ribs continue to shrink, while the faceplate, supported by the cores, does not. That is why the ribs are about 1.5 to 2mm shorter than the faceplate...

To my eye the mirror looked perfect, and that was confirmed to me by the casting crew who was starting to unpack it and cleaned the upper surface for quick inspection.

The only other activity in the lab that attracted my attention with camera in hand was the Large Optical Generator working on the GMT3 backplate. Shown at left is a 3-frame mosaic showing the operation in progress. Flooded with coolant is the flat backplate, with each of the holes (each core has a 3.5" hole that supported the ceramic core) plugged with a baby-blue castable mold material that keeps the coolant out of the substrate. Shown in the left distance is the generator spindle enclosed in the ventilation system where the mist (contaminated w/glass particles) is exhausted to HEPA filters on the roof. The yellow frame is a safety rail to keep workers on the backplate from falling the 10 feet to the floor.

At right is a ground-level view of the base of GMT3. The mirror substrate, face down and glued to the spider via 36 annular pads, is mounted to the turntable for backplate fabrication operations. The spider is what lifted the GMT3 mirror out of the oven and held it vertically for mold removal. When this backplate operation is finished, it will be installed in the polishing cell for faceplate work, and the 36 pads, held on with RTV, will be wire-sawed off. I've posted about generating before, check this link for some images. On GMT3, backplate work is nearing completion - they are currently using a fine resin-bond wheel for closing out the backplate work.