Re: ATM AO and Youngs Modulus for glass

[GregJ888@aol.com]

The Bi-morph works on curvature, the Shack Hartman in tip/tilt (basicly
measures the change in the perpendicular).  Sort of like polar and cartessian
coordinate systems.  There are several ways to attack the problem, but there
is a transform required between the two (as near as I can tell).

Probably the best way is to up the number of sense points in the Hartman to
allow a better model of the wave front then generate a correction matrix to
feed back to the bi-morph.

Another way is to simply measure the change in the sensor output for
correction voltages to each electrode, then use this data to build the
correction matrix.  Not as mathmatically correct, but likely as effective with
the quallity of the bi-morph I can assemble.

Since this system is for a fairly small telescope, .5m, and only 5 modes of
correction I hope to run pairs of electrodes from each D to A channel.  I'm
also looking for improvement, not perfection.  The D to A has channels
available and plan to use 1 for the calibration source.  There are several
configurations for the Bi-morph under investigation: 
1) central area, and 2 rings with 4 sections each; 
2) central area split into 4 pie shaped sections with an outer band with 4
sections;
3) center section with 2 rings with 3 sections each.  

The outer band is for edge control (read tip/tilt) with the correction done in
the inner areas. 

Either of the first two, with electrodes connected in pairs will allow a 6
channel A to D card to do the job.  In the long run that would be a nice fit
for available hardware.  The third option is almost sure to work well, but may
not provide the best correction and none of the current mathamatical
discriptions are base on 120 degree axis shifts.

Just have to change a single ring in the Hartmann sensor to change its
configuration and rest of the hardware will work unchanged.

Greg Jones