[ATM] Round Robin C Foucault profiles compared to Interferometry

[Dale Eason <atmpob@yahoo.com>]

I have uploaded the last segment of my analysis on
mirror C.  In these images I compare the surface
profiles from the data provided by the Foucault
testers to my interferometry by graphing the surface
error heights in nanometers.  The interferometry
profile data was derived from only the spherical terms
since that is about what Foucault can measure. The
inability of Foucualt testing to measure Astigmatism
has been discussed before and I will not address it
here.

I found the profiles fell into three general shapes. 
The first matched the igram closely except at the
center where Foucault did not measure. 

See:
http://www.atmlist.net/contrib/atmpob-at-yahoo-dot-com/Round%20Robin%20C/analysis/igram_Foucault_compared.gif

The second type has a high middle.
http://www.atmlist.net/contrib/atmpob-at-yahoo-dot-com/Round%20Robin%20C/analysis/high_center.gif

The third type has a large bump close to the edge.
http://www.atmlist.net/contrib/atmpob-at-yahoo-dot-com/Round%20Robin%20C/analysis/two_bump.gif
 

At first I thought the two bump versions was just from
the other diameter.  Diameter 2 has that kind of
feature.  That is it has a broader hill at the edge
than diameter 1.  However one tester (Steve Goldman)
has the same two bump feature for both diameters.  

The same is true for the high center version, one
tester (Howard Banich) has similar profiles for both
diameters.  From my Foucault images I know that the
center 1 inch radius is almost a sphere just before
the hill at about 1.5 inches.  So a high center may be
more reasonable than what my interferometry analysis
shows.  The interferometry analysis depends on how
many zernike terms are used to match the surface.  I
think in this case there are not enough spherical
terms used.  I used the max that my program allows (8
spherical terms).  When I create simulated Foucault
images on the data it creates a small dip in the
middle instead of the flat the I see in the real
Foucault images.  The center is not that important
because it will be covered by the diagonal but I
wanted to the analysis to agree with each other.  So
I'm not there yet with the center.

Back to the other parts of the mirror. 
It appears from the documentation that all testers
used the same zone radii for their masks.  So that
does not account for the differences.

Also note that the mask does not measure inside of 1
inch radius or 33 percent radius on the plots.  The x
dimension of the above plots are in percent of the 3
inch radius. 


Dale Eason




		
__________________________________ 
Do you Yahoo!? 
Take Yahoo! Mail with you! Get it on your mobile phone. 
http://mobile.yahoo.com/maildemo 
_______________________________________________
ATM mailing list http://www.atmlist.net/