Re: [ATM] Ultrathin glass and mirrors

[Stephen Koehler <s.c.koehler@gmail.com>]

Vladimir,

> David, you can not do this. Ronchi tests you intend to do will never
> come even close to what Dale was talking about. That means that
> you will be not able to do any serious comparison.
> Minimum what you should do is switch from Ronchi to Foucault.

One thing, at least, David can do with Ronchi is to confirm the strong
undercorrection of that mirror.  We can compare his Ronchi images to
simulated and references sets here:

http://www.visi.com/~mkoehler/eason_davis_misc/davis_ronchi_sim.png
http://www.visi.com/~mkoehler/eason_davis_misc/davis_ronchi_ref.png

The set simulated from Dale's interferometry shows very little bulls
eye pattern after the circle of the "phi" pattern reaches the outer
zone.  The reference set shows substantial bulls eye.

I agree with you, however, that David should be using Foucault in
preference to Ronchi to measure the mirror's correction.

One other thing David might want to do is look for the  astigmatism
signature visible in the simulation.  Comparing the images marked
1.000 to 4.000, the middle part of the center line leans one way and
then the other.  Normally, you would do this test when a mirror is
spherical, because the effect disappears after correction is applied. 
In this case, the center of the mirror is spherical enough that the
effect shows up.  To see it, the axis of the astigmatism must be
angled at 45 degrees.

David,

Forgive me for speaking indirectly to you in my response to Vladimir. 
If you capture a good set of Ronchi images through focus, I think we
can do an interesting comparison to the simulation.  I realize that
your grating is coarser than the simulation (50 lpi, right?), but I
can easily rerun the simulation with the proper grating once you have
the images.  Dale has marks on the edge of the mirror, so you can put
it in the same orientation as the simulation.

--
Steve Koehler
_______________________________________________
ATM mailing list http://www.atmlist.net/