by: John M. Hill
It has been a very busy day, but appears that we now have the world's
largest --- albeit still slightly soft --- mirror blank.
Sorry this email took so long to come out. Not only has the control
room been a very busy place today, but I'm sufficiently tired that
typing coherent sentences is difficult.
We held the furnace schedule at 1180 °C for two hours this morning.
Then, at 10:15 we started to cool fairly rapidly (~50 °C per hour) by
ramping down the applied heat and raising the furnace lid slightly.
The lid is attached to the crane with swivels so we can lift it while
spinning. The lifted lid produced a nice cheery orange blackbody glow
for most of the afternoon. The oven temperature is now 780 °C while the
bottom of the mold is still up around 1000 °C. The interior of the mold
spent a total of about 2.5 hours above 1150 °C. (This was less than we
cooked the Magellan 6.5m --- more like the MMT 6.5m, but there don't
seem to be many bubbles in the surface. Probably the lower number of
bubbles is due to the slower heat up rate.)
The reasons that we cooled the furnace sooner rather than later are
that the surface seemed very clean and free of bubbles and that the
glass level as seen on the close-up video camera (B) seemed lower than
we expected. The video camera only shows the inner tub wall which
forms the Cassegrain hole, so we only have a glass depth estimate at
the center of the furnace. Of course the decision about when to cool
came after the traditional debate between Roger and John. The official
Angel-estimate of the glass depth at the center is 1.15 inches with a
possible uncertainty of 0.1 inches. It is a tricky business measuring
the glass depth when the camera producing the images is very hot
itself, but we get a little better at it with each casting. This time
Roger worked out a scheme that uses our depth marks and their
reflections in the liquid glass surface. I've also got a program that
measures edge positions in the video images. The faceplate has enough
glass to make the polished mirror surface, but the polishing crew won't
have to spend much extra time grinding off the extra glass before the
final finishing.
If the faceplate were the same thickness all across the mirror, then we
would be missing about a ton of glass. Thus, there is a good
probability that the faceplate of the outer part of the mirror is
significantly thicker than 1.15 inches. It could also be that "missing
glass" is accounted for by some combination of small deviations in core
height, ceramic fiber shrinkage, oven bending, thermal expansion,
rotation speed, etc. It certainly is not due to neutrinos or MACHOS.
My favorite theory is that one of our techs had his thumb on the scale
while weighing the glass.
We have been able to view almost the entire surface of the molten glass
during the cooling -- either via the video cameras or by direct viewing
through the gap under the lid. All the 1662 cores appear to be in
their proper locations and we definitely HAVE AVOIDED the classical
casting catastrophe of buoyant flotation of the cores in the molten
glass.
After many attempts we have finally solved the engineering problem of
how to keep hot oven vapors from condensing on the video camera windows
and fogging them. Even though these sapphire windows may be at 800 C
during the casting, they are the coldest thing in the furnace they
condense all the vapors. The trick is to put a second hot sapphire
window in front of the first one. This second window accounts for the
beautiful crisp image you see in Camera C. Camera A without the extra
window has quite a bit of ceramic frost on the window.
At the top of this email I declared only "conditional" victory. We
still have to carefully anneal and cool the blank over the next 90
days, before we get to see and touch it in April. A crew of 5 oven
pilots will be watching the furnace and its control system
round-the-clock during that time.
You may have heard about a "24-foot" crack on Friday. This crack did
NOT (repeat NOT) occur at the Mirror Lab or even in Arizona. The story
as I understand it is that about an hour before a Delta 2 rocket blew
up on launch at Kennedy Space Flight Center, the Mobile Launch Platform
used to transport the shuttle developed a 24 foot crack in it while on
its way to the Pad with Discovery on it. The crack occured with a loud
bang heard for several miles. It was determined it was safe to proceed
to the pad and Discovery is now on Pad 39A. The crack will be repaired
at the pad. NASA terms the crack a "superficial flaw", they also
called the Delta explosion an "anomaly" so who knows. The word is that
it should not cause a delay in the launch of NICMOS scheduled for
February 11.
Yup, "anomalies" like that I can do without.
Changing to my LBT Project Director's hat, I'd like to thank all those who
helped make this casting a success. Thanks also to the many people
who sent their best wishes by email and kept their fingers crossed.
Good Evening to all. I'm off to catch up on my sleep.
SOML declares conditional victory on LBT!
Back to LBT Casting News