Re: ATM Progress on the 13.1" mirror (figuring round 4)

["Bill T." <twentiethwave@hotmail.com>]

Here's a helpful experiment:

Give Tex the normal parameters for your mirror, but make all the caustic
readings "0.0."

Look at the Tex graph.  That is what a perfectly spherical mirror looks
like.  Note the low spot at 70%.

A perfect parabola would look like a straight line.  Tex biases its graph
representation to remove the part of the curve coming from the parabola.

If you are bringing up the parabola more or less symmetrically, that low
spot at 70% will persist as long as remants of the sphere (spherical
aberrtaion) are present, but gradually get flatter and flatter, until the
line is finally flat when the parabola is fully developed.

You don't have TUE.  The upward bend in the outside zone is just a "remnant
of the sphere" still present in the not yet fully parabolized mirror.   All
it means is that the focal length on the edge is, for a parabola, still
short *in relation to the focal length in the center*.

The basic task of figuring is to change the sphere into a parabola.  We can
think of a parabola as having a shorter focal length at the center than at
the edge, and that our job is to adjust the focal lengths at center and edge
accordingly.  Turns out, making the center shorter also makes the edge seem
longer in comparison, even when the actual focal length of the edge hasn't
changed -- what's changed is that the edge seems longer  in comparison to
the newly shorter center, even when we may have hardly worked the edge at
all.

Important background... polishing applied on top of areas inside the 70%
zone makes them shorter.  Polishing applied on top of areas outside the 70%
zone makes them longer.  Polishing on top of the 70% zone removes
parabolization.  BUT...anytime that sub diameter lap strays over the edge,
the edge starts to roll.  You have a very limited "polishing budget" at the
edge, before it starts to noticeably roll.  That's why working mainly on the
center is so desirable, it avoids messing with the edge.  You COULD bring
down the edge of the graph by polishing directly on the edge, but a rolled
edge would soon follow.

Just keep concentrating on the center, and chances are that the tilted up
part of the graph in the outer zone will gradually start to tilt down.
Bring a good percentage of the strokes right up to the edge, but don't go
over.

Star test before doing more figuring!   A good 1/4.4 wave mirror (where you
are now) with a smooth surface and a good edge might easily beat a mirror
with better Foucault numbers, but a bad edge, rough surface, and astigmatism
from too many figuring attempts!


Bill T.