[ATM] FW: Large Thin Mirror

[Mel Bartels]

>>>
I'm currently lapping the back of the glass with the aluminum cell using
30-mesh sand (works well). When I get contact out to the edge, I'll go to 80
grit SiC, and then to 120 grit SiC. I think this should be sufficient in my
case, since I'm not trying to achieve a frictionless interface, and I'll be
bonding to the aluminum eventually, so a course finish my work better. 
<<<


A few thoughts...

Many of us have tried to break the mold.  The state of the art of amateur
large thin mirror making is the direct result.  Innovations like plaster
tools, tiles on edges, ring grinders, evolved techniques to avoid 'stig are
all a consequence.

Bob Kestner in particular spent much time with me circa 1980 teaching me
while I was doing my 24 inch.  Kestner is a real hero of mine, and his story
from ATM to eventually heading the Hubble corrective optics project at
Tinsley is worth knowing.

The result is that very few stones are unturned today.  There might be a
stone that we haven't seen, or a stone that can be turned in a new way, but
the odds are against.  Best strategy is to learn the state of the art on the
first mirror, then experiment on the second.

When you tie the mirror to a backing substrate like aluminum, you may end up
with the benefits of none and the complications of both.  Consider what it
means to work the glass back and aluminum front until both match to optical
tolerances at the nanometer level.  Even fine grinding with 3 micron
aluminum oxide will get you at best to within a micron, which is a 1000x too
coarse.  If you try to go finer than 3 micron, you will walk into a
disaster, which many of us have endured.  The nanometer level is because
that's the resolution at optical wavelengths, and unless you design special
tooling that polishes with relieved pressure, you're likely to sag the glass
during polishing.

If you glue the glass to the aluminum, now you have a 3 layer stack: glass,
glue, and aluminum.  One idea that I think worth exploring is the gluing of
glass to substrate via a thin ring of glue at the 70% zone as indicated by
PLOP.  This has been tried with some success and disastrous failure, but the
failures were possibly due to the implementation.

Another idea is to discard the glass completely and work with aluminum.
There's much in the archives.  Aluminum is not fully resolved, IMHO.

Mel Bartels