Using Undocumented Glasses

Many ATMs are tempted to use undocumented glass to make refractor objectives. This can be done very successfully, but certain precautions must be taken to ensure that the glass is really suitable and that your project is likely to turn out well. As I mentioned earlier on this webpage, you MUST know for sure exactly what kind of glass you are dealing with. Many different crowns and flints exist and they have extremely variable properties. It is not safe to assume that every unmarked blank of optical glass is BK7!

In many cases, though, some record exists concerning the type of glass. Often it is marked on the cylinder wall of the blank. I have in my possession, undocumented specimens of BK7, F4, F15, BaF4 and BSC2 which I picked up second hand. Since the first four disks specified Schott glass types, I was reassured to examine their cylinder walls and find number and letter forms which I know to be used in Germany. This writing not only specifies the glass type, but also gives the melt numbers and other identifying codes. So, in this case I feel confident that I'm dealing with real optical glass of the specified types.

Even so, that doesn't mean that the glass is good enough to make lenses. It must be scrutinized further. One sometimes comes across old "pressings." These are disks of glass that have been molded into predetermined shapes. I have a number of them, both crowns and flints. They may have their glass types molded into the faces as well as their approximate radii of curvature. Recently, I obtained such a pressing of BK7 and it has turned out to be very fine glass suitable for making into the corrector lens of a Schupmann medial refractor.

If you come across pressings, it is probably safe to assume that the glass is good enough for a telescope objective. If it hadn't been, the makers never would have gone to the trouble of molding the original glass.

More trouble comes when the glass is merely a flat disk. Then it may be good, or it may be very bad. The following two tests are sensitive and MUST be performed, if you don't want to waste a lot of time fabricating a failed objective.

The first test is easier to perform, but not decisive. It evaluates the quality of the annealing in the glass. This test is well described and illustrated by Texerau in chapter 7-1 ("Rough Check for Strain"), and I leave you to study his method. Note that the test is easier to perform, if you can get or rig up a photographer's light box and cover the front surface with a large sheet of Polaroid instead of looking into a window glass. But the window glass will work, if you follow Texereau's method.

It's also easiest and best to perform the anneal test if your disk is fine ground or better, polished on both faces. But if it is not polished you should smear oil on both faces of the glass before trying to look through it. IT IS ESSENTIAL, that you be able to see the whole surface of your glass AT THE SAME TIME in polarized light. Don't try to examine it piecemeal.

If the examination in polarized light reveals bright colors of light in the glass, then forget it. That glass is junk and can never be used for a lens. Best of all is if the glass goes evenly dark all at once with no colors as you rotate the polaroid in front of your eye. That glass has an excellent anneal. But often your glass will show some sort of dark gray cross pattern. This should be well centered on the disk and quite hazy in outline. The hazier the better. No colors should be visible. If the cross fills up most of the disk, then the anneal is good enough. If it is strongly marked with lots of white between the arms, then the glass will be a gamble to use. Its figure will constantly change as you rub on it and you may never be able to control what the figure does. But I have successfully used disks that had a large hazy cross and only a small bit of white area.

The second test is exactly what Texerau shows you in chapter 10-7 ("Polishing and Retouching"). This involves grinding and partially polishing both faces of your disk approximately flat and then setting the disk up in front of a good long-focus spherical mirror. A knife-edge test, or better still, a Ronchi test will tell the whole story. You are looking for the homogeneity of the glass. If the disk is not sufficiently even in quality throughout its substance, then it will refract light irregularly, and try as you might, you may not be able to figure your lens even to a 1/4 wave null. I know this from experience.

So I would urge you to study Texereau's method and his picture of the optical defects in a 16" glass plate (figure 94). Here you will see in an exaggerated form most of the defects you are likely to find in a disk of optical glass.

Some notes to help you conduct the test. First, a Ronchi screen is much better for this test than a knife edge. What you are looking for is how symmetrical the bands are across your glass disk. Ideally, looking thoroughly your disk should be like looking through clear, still water: there should be no change in the bands with the disk in place and with it removed.

Usually, though, unless your disk is very good and you've ground good flat surfaces and polished them well, which is a big pain, you'll see some changes. Typically, when I run this test, I grind the two faces only approximately flat (to within 50 microns of true flatness across an 8" diameter spherometer), and only polish them enough to transmit light. I don't figure them. I also don't bother to remove wedge from the disk unless it's over 100 microns. The wedge causes lateral color, but an e-line filter will remove the extra distracting colors.

And because I don't bother to grind and figure the faces, I get spherical aberration and maybe some axially symmetric zones. But none of this matters for an f/15 achromatic lens. If the glass is good, then the bands should be symmetrically shaped as they would be on a sphere or a parabola, and though the disk may show the center or edge rather hazy and dark due to the poor polish, nevertheless, the bands won't be asymmetrically warped, show zig-zags or irregular twistings. If your glass shows such peculiar NON AXIALLY SYMMETRIC features, then forget it. The glass is junk. I've tried to figure glass like that, and it was hopeless. The kind of inhomogeneities that can be figured away are very mild ones like those which Texerau discusses in his Figure 94. They're hardly noticeable by this test. Anything strongly marked in the glass is hopeless, in my opinion. Striae are also bad and good glass shouldn't have any. If yours has some, then they should be few, very week, and preferably curved. Otherwise, I'd advise you to look for some different glass. Making optics is frustrating enough without inviting more trouble on yourself! It goes without saying that if you buy directly from a glass supplier, then you'll get recent glass which should be without any problems. It's the older disks and the unusual glasses that can present difficulties.

If you examine your disk by the knife-edge test, then the cutoff should show only axially symmetric zones, clearly marked. Preferably the cutoff should look just as it does when the disk is not in front of the mirror. I once examined a large disk of special flint glass by this test. It showed a veritable star-burst of irregularities. The knife-edge was nowhere to be seen, even remotely. After much thinning of the disk, I still saw hopeless irregularities. In consequence, it has made a heavy piece of landfill and I have successfully completed other lenses.

Lastly, I should say that you can find many beautiful specimens of undocumented glass out there. The majority of the pieces I've looked at were great, even larger older pieces, like my F15 Schott flint 13" in diameter and 2" thick. So don't despair, but do test!