[ATM] Re: Second Thoughts on Mirror Support

[jonah at eecs.berkeley.edu (Jeff Anderson-Lee)]

Here's a scenario that seems promising:

16 by 1+5/8 inch f/4 pyrex mirror
9 RTV pads: 3/4 inch diameter by 3/16 inch thick
1/2 inch steel annulus back plate 12.64in diameter with 2in central hole.

The P-V on the thicker back plate is reduced to 1.22e4-4 mm (4.79e-6 
inches). 
By monitoring the contour lines drawn by plop and shifting the plate 
support
points we can control the relative flexing at the mirror support points 
to give
us another factor of 5 or so versus the P-V, bringing us down to 1e-6 
inches
of deformation.  The net force for a 1e-6" deflection on the smaller but 
thicker
pads  is 0.20 lb, which is safely under our 0.67 lb limit.

It doesn't leave a lot of wiggle room for differential expansion, so I'd 
probably
stick to using this for steel with plate glass, but nonetheless, it just 
might work!

Of course, now you have a huge heat sink sitting 1/8 in behind the 
mirror, but
at least with the central hole you could put in a muffin fan to provide 
a flow of
ambient air to improve cooling.

Jeff