Re: ATM Pressure distribution between discs

[mdholm@telerama.com]

Quoting Dale Eason <atmpob@yahoo.com>:

> I belive that this has already been done.  I seem to
> remember a discussion about it a few years ago.  Check
> the archives.
> 
> Perhaps other will remember the details.
> 
> Dale

Jim Burrows did a program that designs tapered laps to produce a particular 
wear pattern when used with a certain center over center stroke.  I used it a 
bit about a year ago but ran into two problems: 1.  I never did get any ability 
to control what happened at the edge of the mirror.  2.  I tended to get very 
zoney results.  I think this was at least in part because the laps usually had 
very little contact area in some portion, and because the fairly short COC 
stroke required didn't blend things very well.

Jim's code, which he supplied me, and discussed with me, is clever, perhaps a 
bit too clever for my brain because I never fully understood it, though I did 
make out what he was doing in outline.  Jim's code made no attempt to deal with 
the pressure distribution problem when one disc overhangs.  He just kept the 
strokes short and ignored the pressure problem.  Jim's code also made no 
attempt to allow different size discs, i.e. the lap smaller or larger than the 
mirror.

I want to do a range of strokes including chordal, W and COC.  I want to do 
significantly off center strokes, where the pressure distribution problem can't 
be ignored (at least I think it can't).  I want to do subdiameter laps and 
maybe superdiameter laps.  I want to do MOT and TOT.

Jim used some elegant code based on a geometrical analysis of the situation.  I 
think, for my purpose, it will be best to stick to a rather plodding approach 
where I just step across the whole surface, calculate the instantaneous 
situation at each area element and sum over the stroke on the surface of 
interest.  Then perhaps, having completed one stroke's worth of calculation, it 
will be safe to use symmetry and rotate the wear around the whole surface.

My way will do a lot more arithmetic, but, the way I see it, all these 
megahertz and gigahertz might as well be doing something useful instead of 
chasing a silly screensaver around my monitor.

Mark Holm
mdholm@telerama.com