Re: ATM Re: Catch a Wave

[Aplanatic@aol.com]

I wrote:

>The real problem with mirror cells is performance when the scope
>is pointed at some large angle with respect to the zenith.  This can
>cause quite a bit astigmatism for thin mirrors.  In my opinion this
>effect is typically much larger than 1/60th wave RMS for a large Dob.
>To ask for great performance when the mirror is pointed at the zenith
>and then ignore the performance when pointed near the horizon is
>not so intelligent.  It's really this aspect that needs addressing.


Richard wrote:

> PLOP cannot handle this problem, but I have done some NASTRAN models on flat
>  plates.  When the tangential forces are equal over the flotation points (a
>  result of compliant silicone blobs), the in-plane deformations are and 
order
>  of magnitude less than the axial deformations with the telescope pointed at
>  the zenith.  These deformations must be multiplied by the slope of the 
glass
>  (a small number) to get the impact on optical performance.

Yes, that result seems quite reasonable if the forces of gravity are 
taken up by many points.  The problem is getting an even support 
for the mirror so that each of the compliant blobs is doing its job.  

Here I am assuming that we have discarded the Sling idea, although 
many may be using this approach.  I only consider blobs attaching 
the mirror to the floatation elements.

In a typical mirror cell there are three adjustment bolts that connect 
to three fancy, but independent, floatation systems.  It is not obvious 
to me how the forces due to gravity are shared by these three bolts.  
It seems quite possible, at least with some floatation designs, that 
most of the tangential force could be taken up by one bolt.  If so, what 
then is the deformation of the surface due to gravity?  Or, give me an 
argument that shows that all three bolts must share the load equally.

In general, how is the tangential load shared by our blobs?  Just 
because they are compliant does not mean that they share the load 
equally.

An unrelated, but important question:

If the floatation system is composed of plates made out of a 
material that has a different TCE than the mirror, what is the effect 
of differential expansion on the surface profile under changes in the 
ambient temperature?  This will depend on the diameter and thickness 
of the blobs, of course.  Is this effect important?
  
>  Recently I have started playing with FLEX PDE, which you can get from the
>  web.  I want to model the eigenfunctions of a muscal saw.   In a musical
>  saw, the steel is bent into an "S", and the bow is applied to the 
inflection
>  point-- that is the hot spot that sings.   This is a non-trivial problem 
for
>  me.  Modeling as a drumhead or thin plate does not really represent what is
>  going on with the shape of the bend.   But that is off topic here...

Maybe off topic, but interesting.  In a similar vein, can anyone 
explain why a mirror sings sometimes during polishing?   When 
I use a very thin mixture of CeO on a microchanneled lap and let 
it almost dry out, the mirror will resonate and make a high-pitched
 note.  A note with a very high Q.  I assume that the mirror is vibrating 
in some fundamental mode, but why it requires just the right conditions 
of pitch, water, and CeO is a mystery.

Dave Rowe.