RE: [ATM] RE:anealing

["Scott Milligan" <starzkey@charter.net>]

Bob:

A perfect paraboloid when tested in autocollimation doesn't show a toroidal
shape because this is a null test for the paraboloid.  This is still true
when the flat is sub-aperture, but only if the auto-collimation test is
properly aligned (i.e. if the source is truly at the focal point, and not
some other point).  Other conic sections are not nulled using a flat, and
hence show various amounts of residual spherical aberration, which, to a
first approximation, appears toroidal when the flat used is sub-aperture.
Stated another way, I was asking how does the optician tell the difference
between a miss-aligned null test using a perfect paraboloid, from a test of
another conic section, perhaps close to the paraboloid but not so close that
you would not try to make it better if you could resolve this ambiguity.
David wrote back saying that all you would need to do is to "square on the
flat" to align the test, and in principal, I agree with this statement.  

But as usual, the devil is in the details; how would one square on the flat?
When the flat is full aperture, this task is readily accomplished by
adjusting the tip tilt of the flat and mirror under test until coma is
removed from the test results.  But in a subaperture test, every problem (to
first order) looks like astigmatism, making it difficult to separate out
mis-alignment of the test cavity from residual under or over correction in
the mirror.  I guess one could take the time to ensure that the mirror under
test had its back side ground accurately normal to the mechanical axis (i.e.
obtain a wedge free condition), then square the flat by making the flat
parallel to this surface, using an autocollimator or perhaps just a simple
laser beam.  But I would personally want to analyze such a plan in some
detail to prove that the required accuracy in the test alignment could be
met before relying on the results.

Scott Milligan

-----Original Message-----
From: atm-bounces@atmlist.net [mailto:atm-bounces@atmlist.net] On Behalf Of
Bob May
Sent: Wednesday, May 25, 2005 12:44 PM
To: atm@atmlist.net
Subject: Re: [ATM] RE:anealing

Scott, I'm kind of lost in space with your post!
How is it that a parabola is the only surface that doesn't show a toroid
shape?  Any off-axis aspheric will show not that it is a toroid but rather
that it is an odd durface.  The paraboloid is only an artificial point along
the scale of the oddity as you go along the shape of the surface.
When you try to test an off-axis surface of any shape, you have to take into
consideration that you are indeed looking at a part of the whole surface of
a figure of revolution or you will end up with all kinds of wierdness in the
surface shape.  Align the surface so that it produces a particular direction
of the offset of the surface to the figure of revolution and you can only
test as a 1D surface shape and that is only what the Foucault test does
although you can rotate the surface and gain further info in other
directions, unlike the off-axis shape which will obviouusly show a different
shape in different orientations.
I really haven't played with an off-axis surface with an interfrometer but
I'd suspect that you can't obtain a result of the surface unless you use a
compensator of some kind to return an image that is the spherical wavefront
of the surface.
Bob May
bobmay@nethere.com
http://nav.to/bobmay
http://bobmay.astronomy.net

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

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