Re: [ATM] Mirror Cells

[mdholm@telerama.com]

>Keep in mind that the mirror and it's support will inevitably have 
>different rates of thermal expansion.


Also, that any support, as well as the mirror itself, flexes as the telescope is
moved.  The flexure may be small, but the requirements are very demanding.

The front of the mirror should not be bent more than about 5-10 nanometers from
the ideal parabola for high resolution observing.  That is 5-10 billionths of a
meter or 5-10 millionths of a millimeter or about 0.0000002 to 0.0000004 inch.

Ordinary machine work, as practiced in home workshops is doing well to hold
0.001 inch tolerances in routine work and flexure is noticible in most
mechanical assemblies at that level.  Ask a machinist about setting up a lathe.
 Despite the generous proportions of steel, it is easy for a lathe bed to be
0.001 inch or more out of alignment.  In fact, if it is not carefully aligned
after installation, it most certainly will be out of alignment by 0.001 inch or
more.  For the mirror surface, we need to come up with four more zeros, a factor
of nearly ten thousand times better than ordinary machine tools.

This is the fundamental reason that many, semingly obvious, ideas for mirror
support do not work out in practice.

Mark Holm
mdholm@telerama.com
_______________________________________________
ATM mailing list http://www.atmlist.net/