[ATM] Test stand astig., edge support

[Mike Lockwood <melockwo@uiuc.edu>]

Hi,

In response to my theory about test stand astigmatism altering 
Foucault test reading results, Jim Burrows wrote:
> With an on-the-glass perfect mirror sitting on a test stand, the first 
> and obvious astigmatism generated by the test stand would be primary 
> astigmatism, A22 in Born & Wolf's notation.  The trouble is that this 
> astigmatism doesn't change the correction along any diameter, just the 
> radius of curvature, so Foucault would say that the mirror is perfect - 
> however....
> 
> To have the test stand change the correction, it would have to generate 
> higher order astigmatisms, A31, A33, A42, A44, etc.  To predict such 
> things, you'd have to dive into FE stuff, an extension of PLOP.

Very interesting.  Thanks for the insight.

What I am curious of is whether absolutely ALL the test stand 
astigmatism can be modeled with A22, or if there are smaller amounts 
of higher order astigmatism underlying it.  While I respect the 
analysis of the data, I find it hard to believe that all of the 
deformation is merely a change in ROC, and I don't think the data in 
the report proves it, and certainly not for thinner mirrors.

In other threads there are discussions (which have happenned before) 
about ball bearings, where to place edge support points, slings, etc. 
  All these techniques will probably distort the mirror in different 
ways when the scope is pointed near the horizon, or even well away 
from the zenith.

Is there really any hard data to back up the performance of any of 
these edge support systems?

I believe that Mel quoted some computed results that justified his 
decision to support mirrors with posts 90 degrees apart by saying this 
helped prevent astigmatism.  I've been using this method (with Teflon 
contact points) with good results, but I doubt it will suffice for 
very thin mirrors.

It seems like it's time for someone to take some spherical mirrors of 
varying thicknesses and place them on different test stands and do a 
serious interferometric analysis of the distortions induced.  This 
would shed some light on support methods for mirror testing, and the 
expected amplitude of the effects.  Also, if this test could be 
performed with the mirror tilted back (at 30 degrees or so), then we 
could get an idea of what edge support techniques would least affect 
optical quality when a telescope was aimed fairly low.

With more of us thinking about working thinner, bigger glass for 
reasons of weight and temperature equilibration, I think we need to 
know a bit more about edge support.  PLOP has largely solved the back 
of the mirror support problem; now it's time to solve the problem of 
what happens when the telescope isn't aimed at the zenith.

I have a feeling that a different solution is optimal if we want the 
best support of the primary over a range of angles, rather than the 
much simpler case in which the primary is flat on its back.

	Mike Lockwood


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