Re: [ATM] lurie-houghton progress

[Mike Lockwood <melockwo@uiuc.edu>]

Guy,

Guy Brandenburg wrote:
> Yes, convex on concave! ;-) I thought that from one
> side to the other (7.5 inches) you might cut 2 or 3
> fringes if you followed a straight line that would cut
> as few as possible, which is what I think you mean.

OK, so the surfaces agree to 1.5 waves or so.

> With the float/plate glass, all you need is 2 crossed
> polarized plates to see if they are free of
> inhomogeneities (if I spelt that correkly) (;-)) and
> strain and such.

That will let you check for strain, but not inhomogeneity, I believe. 
  Hopefully the glass melt was well mixed.

> suppose that after we pushed a little bit and released
> the pressure, we saw lines curving slightly like the
> lines of wood in a quarter-sawn tree. (as shown here:
> http://www.woodfloorsonline.com/techtalk/properties.html)
> Then what would one conclude, and what action would
> one take?

I would conclude that the surfaces did not match exactly.  Rotate both 
pieces so that the fringes are vertical from your viewpoint.  Put a 
ruler across the pieces, aligned with the ends of the fringe that is 
nearest to the center of the pieces (not at the left edge, not at the 
right edge - right down the middle).

Move to a viewpoint at least times the diameter of the pieces under 
test away from the test apparatus (70" for your 7" pieces).  Count the 
number of fringes the ruler crosses at the center of the disk.  Let's 
say the fringe bends enough that the ruler crosses one fringe and just 
touches another at the center.  This means that the curvature is two 
fringes, and the difference between the surfaces is one wavelength.

To determine what kind of mismatch is present, use one finger to press 
on the center of the top piece and observe the direction that the 
fringes move in.  The fringes move towards the high side, which means 
a larger air gap.  Release the pressure.

Make the fringes shift slightly to the RIGHT when you press on the top 
piece.  (Rotate the two pieces 180 degrees in the tester if this is 
not the case.)  Now the air gap is larger on the right.  Now,

   1) If the fringes curve like (, the air gap at the center is larger 
than the air gap at the edges, and the convex piece is not convex enough.

   2) If the fringes curve to the left like ), then the air gap at the 
edges is larger, and the convex piece is too convex.

So, we can also say that wherever the fringes bend to the right 
(compared to a straight line), the convex piece has a high spot.  For 
1), it's at the edge, for 2), it's at the center.

> Would one conclude that one of the pieces of
> glass has astigmatic cylinder?

It's possible, but not likely.  If you get straight fringes that are 
not parallel, then you definitely have astigmatism.  FringeXP can also 
be used to analyze the difference between the pieces.

What would be good is to take a photo of the pieces under test, with 
the fringes aligned as mentioned above, and so that they shift to the 
right when you put pressure on the top piece.  Post the image and we 
can tell you what's going on.

	Mike Lockwood


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