Re: [ATM] Ritchey-Chrétien

[Jim Burrows <burrjaw@earthlink.net>]

At 2008-01-20 18:13 +0100, Vladimir Galogaza wrote:

>But at the final stages, the more we are
>close to the ideal (sphere or anything else) the less we rely on randomness.
>Just few strokes or few minutes of action before test, are allowed. If there
>will be
>real randomness, than final actions could last minute or hundreds, result
>will be always the sphere.
>But nobody believes in this. People are cautious and do few strokes and lot
>of
>measurements to find out where to put further non-random efforts.

In pursuit of non-randomness, I've been using starlaps on a pottery 
wheel.  Theoretically, if the starlap follows the deviation curve of 
the mirror, the deviation should be reduced.  Of course, the random 
effects, missing contact, lap glazing, etc., get in there.  To 
address that problem, I've written a program that takes the input of 
three curves:  initial deviation a, lap design L, and resulting 
deviation b.  The program says, OK, if lap L transformed a to b, what 
lap design should reduce b to zero?  It has worked, somewhat, 
figuring a 108 mm RC cass secondary (I've gotten the system RMS down 
to 30 nm, according to 3 tests on the secondary - took quite a few 
spells, though).

I don't think starlaps make much sense on big mirrors, so we should 
look at what motions of a machine should reduce the mirror surface 
deviations from a desired surface.

         -- Jim Burrows
         -- http://home.earthlink.net/~burrjaw
         -- mailto:burrjaw@earthlink.net
         -- Seattle N47.4723 W122.3662 (WGS84)

_______________________________________________
ATM mailing list http://www.atmlist.net/