[ATM] TraceXP, Ronchi Simulation, etc.

[Aplanatic@aol.com]

Greetings,
Among other things, I've been working on a new ray tracing program called  
TraceXP.  It works under the windows environment and is very easy to use  and 
understand (I hope).  I'm almost done with the first version and will  be 
looking for a few folks that would like to beta test it.
Along with standard ray tracing and optimization, TraceXP performs several  
physically-correct simulations including the point spread function and the  
Ronchi-Foucault test.  
I'd like to share with you something that I find interesting.  The  following 
JPG image shows a comparison between the simulated Ronchi test of a  mirror 
that Wolfgang Rohr is testing and an actual photograph of the Ronchi  pattern.  
The test was performed in autocollimation on a 250 mm diameter  f/4.8 mirror. 
 The ray trace was set up to mimic as carefully as  possible the actual test 
system, and included the flat with its central  hole.
There is one significant and obvious difference between the actual test and  
the simulation.  In the simulation a point source was used, and in the  actual 
test an extended source was used.  The extended source tends to  (thankfully) 
smear and obscure some of the fine diffraction structure.  
If you'd wondered in the past why your Ronchi test looks so  complicated and 
confusing, the answer lies in the very complicated diffraction  pattern that's 
formed by the grating, as shown in the simulation.  It gets  even more 
complicated when a paraboloidal mirror is tested at its center of  curvature.
Here's the image:
_http://members.aol.com/aplanatic/Ronchi_test.jpg_ 
(http://members.aol.com/aplanatic/Ronchi_test.jpg) 
The image on the right is the photograph that Wolfgang took of the Ronchi  
test.  The two images on the left are the full physical-optic simulation of  the 
Ronchi test of a perfect mirror of the same size and fratio using a grating  
with the same pitch.  One simulation is inside of focus and the other is  
outside of focus.
There are many interesting extensions of this work.  Perhaps the most  
interesting at this time is the full diffraction simulation of the Foucault test  
using masks.
Your comments are welcome.
Dave Rowe
_______________________________________________
ATM mailing list http://www.atmlist.net/