[ATM] Re: Weird Foucault readings

[Vladimir Sacek <vladis.2@juno.com>]

Nils, 

I think your real objection was that it isn't that you "tried" to
determine the most convincing (you may prefer some other attribute) null
location, but that you did actually determine what it is. The reason I
said "tried" is not because I think it's not valid, rather that
determining this point for a perfect mirror has more or less compromised
usefulness for actual (imperfect) mirrors. But, then, it seems to be the
best point of reference we can have.

I figured out what caused "zonal confusion". Replace what I called "zone"
with "zonal opening" or "zonal section" and that will do. I tend to
simplify the terminology, knowing what I mean by it. Wrong, but works for
me, and I don't care for anyone else ') Forgive me - mea culpa.

>> The point is that we don't know the profile before the test; if we
would
>>know it, what would be the testing for? To test the testing method?

>I see no reason to share your nihilism. While Foucault testing, like any
>alternative, will have trouble with small-scale irregularities (as
compared
>to the widths of the zones), it can handle large-scale ones gracefully.

Realism is probably better a word. What guarantees that there won't be
small scale irregularities? Alternative test may have difficulties to
measure the extent of error of small scale irregularities, but it won't
necessarily diminish their ability to determine figure error - which may
happen with the Foucault. The factor is not only size of irregularities,
it is also degree of rotational surface asymmetry. Both need to be
sufficiently controlled by the actual parabolizing process in order to
have reliable Foucault test results. I believe Carl Zambuto, among
others, has good ways of controlling them.

I went over my previous post, and noticed an error. Total longitudinal
spherical aberration of a mirror is 16AwF^2, not 8AwF^2, as I put. 

Vlad


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