Re: ATM 3/4" mirrors, multipoint cells, machine generated optics

["Bob May" <bobmay@nethere.com>]

I would sort of suspect that the half-pound or so of glass sitting on each
point will tend to make it work decently well, especially with its neighbors
wanting it to be that way.  Thus, I would think that a 162 (4 levels of
triangles followed by a layer of beams) point cell should function quite
nicely.
Probably the biggest reason why we haven't done thin mirrors is that the
pros have gone and started doing multisegment mirrors rather than keeping
with a single surface for their large thin mirrors.  The other end is the
builtup (HexTek) or cast (Roger Angel's work) mirrors which are a bit more
difficult to do for a start.  Thin mirrors have the stigmata of astigmatism
attached to them for obvious reasons of flexing while working the glass and
the subsequent lack of accuracy in the shape.  On a really thin mirror this
would be almost like taking some water and beating it into submission to
form a permanent parabolic shape in a pan.
Bob May
http://nav.to/bobmay
bobmay@nethere.com
NEW! http://bobmay.astronomy.net