I belong to several on-line amateur telescope making groups. In them I saw many posts from Horace Davis about his attempts at casting mirror blanks in his kiln. He began his efforts in January '08. Almost all ended in failure, but he did seem to be getting closer to success. When he offered to cast anyone a mirror that could send him some Pyrex, I decided to get in touch with Horace.
I learned he had basically two problems. One was simply keeping his casting mold intact throughout the process. The other was getting the right annealing schedule. His kiln is computer controlled to follow any schedule of temperature vs time. I couldn't help him with the former, but I could the latter. I had an idle mirror grinding machine (a modified MOM 20), so I volunteered to grind and polish his blanks and check for strain using the cross-polarized light test.
Soon, three plate glass blanks arrived, two 10 inch and one 7.5 inch. They looked pretty bad. They showed a lot of strain where Horace had taken a belt sander to them. He had removed lumps of glass caused by a leaking mold, but they showed no strain due to poor annealing.
Soon after that, Horace managed to obtain a sizable amount of Pyrex. He cast a 15" x 1.5" thick blank and sent it to me. Compared to his others, this was a masterpiece. (Note - most of the full size images that follow are around 2 megabytes.)
| This shot and the next are how the 15" Pyrex blank looked when it arrived. There is some surface divitrification on the top. Divitrification is caused by the tendency of the glass to return to it's crystalline state. There was a small positive meniscus that showed up near the edge. It does not show up well here. | |
| The roughness on the back is caused by the casting material. Some of it sticks to the glass when it comes out of the mold. Horace did remove most of it with #80 grit and a pipe flange. |  |
|
The blank is on the turntable of the modified MOM 20. The tool is 11 inches in diameter (75 percent). It is made of 1" porcelain tiles glued to Hydro-Stone. I started off gently with the position of the tool about 1" offset from the center. The mirror and the tool look thicker than they actually are because they are both taped to same diameter 3/4" plywood discs. |
 |
| As I moved the tool outboard, I began to hear a loud and almost scary thunp-thump-thump. I recognized it immediately as a high spot and tracked it down to this (dark spot). I had not noticed it before. Whatever it is, it caused the immediate area to rise above the mean surface. Got rid of the noise in about 20 minutes. Be sure to take a close-up look at this. |  |
|
Took this shot in late evening when the sunlight was coming through the garage door. In only about 20 minutes, the divitrification began to disappear. This photo gives a pretty good idea of what divitrification looks like. Take a close look at the tool. It was an experiment in gluing tiles to the backing using polyurethane glue (Gorilla Glue). It looks kind of like a sponge. Many places for bad things to hide. Use it again? No. |
 |
|
End of 9 micron white Alox. High/Low incidence angles. Mundane but I like the pictures. My fancy CCD camera has auto-focus only and I'm never sure what it is focusing on. |
 |
 |
| The surface tension of Pyrex causes it to form a positive meniscus. A natural bevel if you will. On this blank, the bevel varied from 0" to 1/4". The 0" occurred at the same place where the surface was raised. |  |
| The front is nearly polished. Most of what you see is the back side. The cross- hatched stuff is shelf liner. It lies on top of a 15" x 3/4" plywood disc. It is time to flip the mirror and work the other side. |  |
| The back is shaped by the casting material. It is rough and irregular. The black stuff is #80 Silicon Carbide. The mold material is somewhat fibrous and the imprint can be seen in this picture. The blank is wet. |  |
|
After about 7 hours of #80 SiC, I decided to call it on rough grinding. This should be good enough for a strain test. The large un-ground area appears to be caused by a place where Horace had patched the mold. There is another un-ground area along the edge at about the 1:00 o'clock position. Here, the edge is rolled down. |
 |
|
At the end of 15 micron, the area of the depressions had shrunk considerably, but there are some spots that are still quite deep. The place where the edge is rolled down is now at the 8:00 o'clock position. There is some roll down in the entire southeast quadrant. |
 |
|
While I was at it, I experimented with another type polyurethane glue (see 2nd pic). What better time than when pits and scratches don't matter that much? In the first pic, I am near the end of #80 and spreading the glue for another layer of porcelain tiles. In the second pic it is cured. The third pic is what it looked like at the end of 9 micron. It held up well but it took about 15 hours to cure. I did not add any additional moisture. I was afraid expansion might cause it to rise above the surface of the tiles. That is what happened with Gorilla Glue (see above). |
 |
 |
 |
| Both sides are now polished. Neither side is completely polished. |  |
 |
|
The strain test stand. The rack for the mirror is for hanging file folders. I am using a flood light to illuminate the wall. The polarizing film is framed with Foam Board. The second photo is mirror on stand and no polarizing filter.
|
 |
 |
|
Finally the answer comes. There is little to no strain. The four photos are of the mirror rotated to angles of 0, 90, 180, and 270 degrees. The bright white spots are flaws in the filter, not the glass. Also the wall is not evenly illuminated which may confuse the issue a tad. These pictures just do not convey the same message that you get visually. |
 |
 |
 |
 |
|
There are a few blemishes worth noting. This a macro shot of what appears to be kiln wash floating up from the bottom and penetrating through the top. Most of this is subsurface. There are perhaps a few dozen tiny specs that made it through the top. You really need to see this with both eyes to get the depth effect. The scale on this is about the same as the mirror. That is 1" in the photo is about 1" on the mirror. |
 |
|
This is another strange one and I cannot adequately capture it in a photo. You can actually see a small distortion in the glass that seems to be a change in refractive index. The best analogy I can come up with is that it looks like a tiny drop of Karo syrup on clear glass. However, the distortion is subsurface. Again, the scale is 1" in the photo equals about 1" on the mirror. I don't see either of these blemishes affecting the figure or the coating. |
 |
You may e-mail me by clicking here or Horace Davis by clicking here. Horace also has a toll free number (888) 362-2323.
Tom Stokes - June of '09. Sayonara.