The Cross Polarizer Test for Stress and Straie in Optical Glass
Background
Often, amateur telescope makers (ATM's) come across optical glass of unknown origin. Only rarely, it seems, do we buy optical glass directly from the manufacturer, with all the melt data and annealing certifications. Other than the obvious examinations for bubbles, inclusions, and color, a bad piece of glass can look just like a good one.
Fortunately, there is a very simple test that will tells us a great deal about the quality of the disk at hand. The first of these test, the classic cross polarizer test, will show stress, areas of the disk not at mechanical equilibrium with surrounding glass, and striations (or striae) areas where the index of refraction (n) changes. You may have seen this effect when wearing polarized sunglasses, and looking at tempered glass, such as the rear window of an automobile. The checkerboard pattern often seen shows the stress purposely put into the glass to make it fracture into small square pieces, rather than large pointed shards.
Polarized Light
Polarizing filters work by "lining up" the waves in a beam of light, something like pushing a pile of sticks against a picket fence. Only the vertical sticks will pass, and the horizontal ones will be blocked. If one were to take another section of the fence, and turn it 90° (so that the slats are horizontal) and place it behind the first, the vertical sticks would then be blocked, and nothing would pass. In a like manner, if turn two stacked polarizers at 90° to each other, all the light will be blocked.
There are two basic varieties, which can't be mixed for this test. The more common of these is a linear polarizer, which is what usually is used in sunglasses. The second variety is a "circular polarizer", which refers not to the shape, but rather how the incoming light is polarized. Circular polarizers are often required with modern cameras, as a linear polarizer can confuse the autofocus and autoexposure sensors.
In practice, two pieces of either type will work for the test, but the circular polarizers add the very minor complication that the filters must be oriented facing in the same direction. Better than a thousand words is a moment's experience. Rotating two cross polarizers with one facing the "wrong" way will not black out. Flip it over, and it will.
(NOTE: I have recently found that using linear polarizers with a large air space between the first polarizer and the glass will sometimes show no apparent defects, but that using circular polarizers in contact with the glass will readily show defects. This is under investigation....)
Performing the Test
Place the disk to be tested between the two polarizers, and rotate one until the light blacks out (as it would if there were no glass present). Lines or bands of color indicate stress in the glass, where the plane of polarization has been changed. Striae shows as ghost-like bands or filaments of white. I often use a CRT glare screen (a large circular polarizer) resting on a transparency film viewer, but almost any source of light will work.
If the glass is not polished, you can rub the fine-ground faces with a few drops of clear oil. I use microscope immersion oil, but I suspect almost any fairly clear oil would work. The second, rotating piece of polarizing material (referred to as the analyzer) need be only be big enough to look through. In fact, even the illuminated piece can be smaller than the disk, as long as it some distance behind, such that it illuminates the entire disk (but see the note above!).
Here is an 8" disk of currently-unknown
flint glass, with both faces polished. This disk shows no stress or striae
at all. The very faint smudges are fingerprints and such from handling,
and though hard to see in the picture, very obviously on the surface of the
element being tested.
Here is a 4"
disk of KZFS glass, and it shows a very obvious line of striae in the glass. This disk
is unusable for a refractive element. Reannealing can relieve stress, but
it cannot change the index of refraction variation. (This is often due to
problems in getting a perfectly even mix during production. KZFS
glasses are rather temperamental; difficult to manufacture to precision specs,
and subject to etching by water or even humid air).
Pictured below is a rectangle of SK6-type glass received from a major manufacturer. The spec called for fine annealing, but this piece is obviously not well annealed. This piece shows the classic "iron cross" pattern of stress in glass (the pattern is centered in the glass, but the glass is a bit larger than my polarizer light source). This piece would not be at all suitable for making a lens, and in fact might crack when cut.
While these tests won't tell you if your glass is grade A or grade B, it will tell you if it is fatally flawed, especially for those making refractors. Generally, striae is cause for rejection of the glass for refractors, but a large, faint pattern of stress can often be "worked around" by localized figuring. For reflectors, striae clearly doesn't matter, but sharp lines of stress are a bad sign, and may well lead to a mirror that will warp during temperature changes. In cases of extreme strain, the glass may shatter from small mechanical stresses.
Pieces of perfectly homogenous and precision annealed glass won't show any color or shadows, even when tested through the edge! This is a severe test, and only the best glass will pass such a test.
Any questions? I'll be glad to answer, just send me an email.