[ATM] Spherometer Accuracy

["Mutalib Abdallah" <omegatroid@hotmail.com>]

Ultimately, the accuracy of a spherometer depends on one consideration that 
nobody has mentioned: the measured surface must be a sphere!   Sometimes it 
is not.

The spherometer will give the same reading everywhere on the surface, at all 
possible rotation angles of the instrument, if, and only if the surface is a 
sphere.   Furthermore, the curvature measurement will not depend at all on 
the size of the spherometer base if the surface is a perfect sphere.

To characterize a more complex curved surface by sampling it at only three 
points is nonsense, just like a two-zone foucault test would be.   We 
frequently encounter seriously non-spherical surfaces during initial 
hog-out, when we first dig a hole in the center of a relatively flat 
surface.   Spherometry is not accurate there.   Also, since we have been 
talking about extremely accurate instruments, we might need to be careful 
about the difference between a sphere and a parabola on deeply curved 
mirrors.

(My apologies.  I'm getting out my fire extinguisher.)


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