From: "Tom Krajci" 
To: 
Reply-To: "Tom Krajci" 


>From: Charles Mitchard 

I'd prefer that more people were involved in armchair research, rather than
theorizing.  One starting point is the ATM archives at:
http://astro.umsystem.edu/atm/
(One of these days our email posts to this list will, hopefully, have added
to them this URL.)
Try a search on phrases such as
"mirror support"
"mirror flexure"
"edge support"

>There is a lot of effort into producing in some instances rather
complex
>engineering solutions to solve the problems of flexure or distortion in
the
>support of the mirror when in the telescope....

>...If this is the case then why not a million point support in the form
of say
>short pile bristle carpet where each bristle will only place an
extremely
>small pressure point before deforming to spread the load further?

You need to cool the back of the mirror.  Your proposal makes it difficult.

Also, you may have internal friction/stiction effects with a million
bristles rubbing against each other.  Rubbing implies friction. Friction
implies stiction.  Stiction implies imperfections in the ability of the
mirror support to respond to the support needs of the mirror.  Can you use
compliant hinges in your support scheme?  It would help eliminate
friction/stiction.

>Maybe that is where the complex engineering solution could come into
play
>as the support for the carpet would also have to be very flat and non
>flexible so as not to allow the mirror to follow said flex.....

You are substituting one 'rather complex engineering solution' for another.
  Can you show us how your proposal is simpler, or more
accurate, easier to fabricate, or more likely to succeed?  That would make
it more attractive.

>If the mirror is free to "slide" on its supports....

Sliding implies friction/stiction/inaccuracies with support. Investigate
compliant hinges that considerably reduce this problem:
http://overton.tamu.edu/aset/krajci/cell-bearing.htm

>I do not want to get into the argument of thermal equilibrium yet as I
feel
>that can be worked upon later.

You may not want to get into the argument just yet...but if you fail to
take thermal equilibrium/cooling into account...you are merely guessing
blindly as to whether or not your design will properly address the
problem....which gives you ample opportunity to re-build or re-design what
you did not thoroughly work through in the initial design process.

>Mirror grinding.

>...If we say place a constricting band around a cylinder of rubber from
its
>base to approx half way up and apply more pressure the cylinder will
deform
>into a for want of a better description a mushroom shape, that is the
>center would rise with a drop in the edge and an outward curve.
>If a similar exercise were applied to a mirror is this what would
happen >also?

>Maybe the only answers are to try it and find out....

No.  Model it in software and save the labor of polishing what may be a
lousy initial figure.

Download a free copy of GRAPE at:
http://www.grapesoftware.mb.ca/

I've, so far, modeled things with up to 500 nodes and 1500 elements.  I
think the software can handle more, but it then takes longer to compute a
solution.

Good luck,
Tom Krajci
Tashkent, Uzbekistan


-+-
Outgoing mail is certified Virus Free. Checked by AVG anti-virus system
(http://www.grisoft.com). Version: 6.0.434 / Virus Database: 243 - Release
Date: 12/25/2002

--- BBBS/NT v4.00 MP