Re: [ATM] Polishing / Figuring Simulator

["Jerry" <wa4guu@verizon.net>]

I think Matt is on to something. 

Aside from his approach, what would be wrong with a results analysis? If one
could predict the wear if they knew the pressure distribution and
stroke/overhang can't one find the pressure distribution from the measured
results of stroke and overhang?

That seems to me to be what humans do. They learn to figure by testing the
optic. By testing they learn the result of an action. That is not to say
that humans do it the best way. 

Then if learning is a part of the process, perhaps a knowledge of the
pressure distribution is unnecessary. The results on the optic is what
counts in the end.

Some of my misspent youth was spent in poolrooms. I did learn something from
it. I doubt that anyone could program a machine to shoot pool (pocket
billiards) by physical theory and calculations as well as a human learns by
rote.  It is a game of physics that is learned by trial and error by those
who play well. No pool shooter I know carries a slide rule... I mean
computer to calculate the shot he will make. Showing my age there. I know
that mathematicians have analyzed the physics of the game, but none to my
knowledge has won a championship. The poolroom slacker does what the
mathematicians can only describe. Not to disparage the mathematicians, I
have more respect for their abilities. I just wouldn't be betting on them in
the pool game.

I suppose that the engineer would like a measured and calculated answer to
every problem. I would think a programmer would love to create a process
that learns to do the task. Some of both might make a powerful combination.



Jerry

-----Original Message-----
From: atm-bounces@atmlist.net [mailto:atm-bounces@atmlist.net] On Behalf Of
matt
Sent: Monday, May 08, 2006 8:48 PM
To: James Lerch; ATM
Subject: Re: [ATM] Polishing / Figuring Simulator

Hi James,

Instead of calculating the pressure distribution , why not measure it ?
A modest 8-16 channel data acquisition system could do that , and it
wouldn't cost much .
The pros use not only computer controlled polishing but also computer
controlled stressed lap polishing.
This changes exactly the lap pressure distribution in the lap/mirror contact
area.
The actuators that stress the lap could be integrated with the pressure
sensors for (servo PID) closed loop operation .
Even if this solution appears more complex, at least it has the potential of
precisely solving the problem you are describing (determining the lap/mirror
contact area pressure distribution ).
I don't think there's even a remote chance of accurately calculating this
pressure distribution without actual value measurements with multiple
pressure sensors , and without knowing the real pressure distribution , all
your calculations become increasingly complex and the errors increasingly
large to the point I doubt they could be useful at all.

best regards,
matt tudor



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