Ed Grinds A Mirror:
A Serial Article
By Ed Ting
Latest Installment: 10/30/00
Index To Mirror Grinding Section:
Page 1 (This Page): Grinding, Polishing, Testing (5/00-10/30/00)
Page 2: Ed Builds a Telescope (11/00)
Introduction:
This summer (2000) I joined a mirror-grinding class hosted by our local club's
ATM committee. Despite having looked through hundreds of telescopes through
my observing career, I had to admit I had never ground my own mirror; in fact,
my knowledge of the whole subject was rather sketchy at best. It was time
to remedy this.
The 6" mirror blanks
(Note: This article will appear in installments over the next few months. Check
back here from time to time to see how I'm progressing. Good or bad, I promise
you will hear all the details! -Ed 5/29/00)
Chapter One: Rough Grinding, May 28, 2000
I arrived on a cool Sunday afternoon in May, and started grinding with five
other club members in a local barn. Right away, there was pressure on me;
I had missed the first two-hour grinding session and everyone was well ahead
of me on the coarse (80) grit. As a result, I used somewhat more pressure
on the blank, and took fewer breaks than the others.
On the spherometer we were using, the correct depth to be reached was 20.8
thousandths of an inch on the coarse grit. Most of the others were already
in the 15-17 range, and one person finished within the first hour of the
second session. Since I was starting at ground zero, this put even more
pressure on me to catch up.
Measuring the depth of the center.
The mirror is currently at f/55.
Anyone have a 27 foot-long tube?
Mirror grinding is a noisy, messy, wet activity. It doesn't take up enough of
your attention to fully occupy you, but it does take up just enough that you
cannot do something else like watch TV, listen to the radio, etc. You're pretty
much stuck there looking at your hands, and adding water or grit once in a while.
This goes on for hours at a time.
The whole session was kept under the watchful eye of our Great Grinding Grand-
father, a man who's been to 50 Stellafanes and has ground more mirrors than most
of us will ever see in our lifetimes. Every astronomy club has at least one guy
like this, and if you need a mental picture, just substitute the one in your club
whenever I mention him here.
The Great Old Grinder looked over my mirror after the first hour and pronounced
that I was "pretty much average" in terms of rate and the quality of my work, which
was fine by me.
Two more people finished up during the second hour, but I wasn't so nervous
anymore. My rate had increased steadily, and I had already registered 17.5
on the spherometer. In fact, I had already passed one person who had been
there for both sessions. I stayed a little late, and actually overshot my f/8
target a little when all was said and done. I also passed another grinder by
then.
Next time, it's on to the 120 grit. As I was cleaning up my work station, I
thought about something I once heard John Dobson say about the rough-grinding
stage: "This is a Caveman's work. You eat well, you sleep well, and in between
you work like hell."
Working Like Hell:
My Mirror Grinding Station
Our Great Guru looked at my work, and said, "Say, you're pretty fast." Then,
ominously, he waved the blank at me. "Next time we'll see if this is a half-
assed job."
Chapter Two: Grinding With the 120 Grit, June 11, 2000
My mirror is hyperbolic. In my zeal to achieve an f/8 curve so quickly, I
ground out the center of the mirror far more than the required depth. I'm
not alone in this; a couple of other grinders also have hyperbolic mirrors,
although mine is the worst off. I work carefully, grinding center-over-
center in short strokes to take down the edges.
After two hours of this, though, there's not much progress. While grinding
the blank, there's a huge air bubble in the center that's visible even from
the next grinding station. The bubble shrinks in size as time goes on but
never completely disappears. Through the din, I think I hear someone refer
to me as "Air-Bubble Ed."
At the end of the session, our Great Grinding Guru looks over the surface
of the mirror with a loupe and says the 80 grit pits are all gone, the surface
looks smooth, and that the focal length now sits at 45 inches.
So what I have here, as of June 11th, is a quickly-manufactured, smooth-
surfaced 6" f/7.5 hyperboloid.
Chapter Three: Finishing the 120 Grit Stage, July 9, 2000
By now we have settled into a comfortable routine. We show up around 3PM
on these Sundays, talk a bit amongst ourselves, take our direction from the
Grinding Guru, and set off to work.
Things have undergone a subtle change, however. There is more talk, more
direction given than before, and somewhat less action. The Grinding Guy
points out what seems like an endless stream of cautions as to what can
go wrong: Clean your station thoroughly, don't wear long sleeves (grit
can collect in them) don't hold the salt shaker over the mirror (you don't
know where it's been) use less pressure, test after every few wets, etc, etc.
As if to reinforce this, our Grinding Guy stays very close to the tables
this time, issuing small corrections, evaluating our strokes, looking
over our work. The message, subtle or not, is clear: you can goof off
all you want during the rough grinding stage, but now things are starting
to get a little more more serious. Pay attention.
Through careful work (or perhaps blind luck) my blank returns from its
hyperboloid and is now a rough sphere again. At around the same time,
the 120 grit pits seem to have dissipated from the surface of the mirror.
Not wanting to push my luck, I call myself finished on this stage and
clean my workstation.
We carefully spray off the tables outside afterwards. There are to be no
120 grits left over for the next phase. I look up and it starts to rain. I
don't mind; it's been a remarkably cool summer, the coolest in recent
memory. I have this to be thankful for. We could have been sweating
in the barn these past few weeks.
Chapter Four: The 240 Grit Stage, August 13, 2000
This has got to be the coolest summer on record.
As the grit stages progress, we spend more and more time preparing, and
less time actually grinding. The work area must be cleaned, dusted,
vacuumed, and wetted down before each use. The tables are power-washed
outside and wiped down with cloth before use as well. All of this takes
time, of course, and the actual grinding part of these sessions is slowly
diminishing.
The work area, just before wet-down.
Tables are outside, drying
We've settled into a routine that's comfortable for us, and with this a
degree of boredom has crept in as well. Things aren't going to get too
interesting until we get to the pitch lap stage, which seems far, far
away at this point.
In fact, there are still three grit levels left, and it's been taking us two
sessions for everyone to get through each level. Throw in the fact that
we meet every two weeks, and this project could go well into winter at
this pace. When I point this out to the group, I see a few shoulders sag.
As a result, we've agreed to increase the length of these sessions. We
intend to meet slightly more often as well. The goal on the upcoming
Sundays is to get everyone through one entire grit level on one session.
What began as a leisurely mirror-grinding exercise is turning into a
race against time. Hopefully we have caught this soon enough.
Chapter Five: The 30 Micron and 12 Micron Stages, August 20, 2000
This was the first of our double-length sessions. Four of us showed up
early to get through these grit stages more quickly. I had stayed up
all night with a group of friends evaluating the Tak FS152 and the AP155
and I was dragging. The prospect of grinding for 4-6 hours wasn't
terribly appealing.
But in a strange turn of events, I had a blast. These smaller grits look
like baking soda or flour, and when you grind with them it is hard to
believe they are having any effect at all. The noisy scraping sounds
from the 80 grit stage have been replaced by occasional near-silent
squeegee-like peeps.
With the microscope, you can measure your progress through the grits.
The pits slowly become smaller and smaller, and when they stop getting
smaller you are ready to move on. It takes me two hours to get through
the 30 micron grit, and I get my second wind. Instead of going home, I
power-wash the table, break out the 12 micron grit, and start again.
The 12 micron stuff feels like finely-ground, sifted flour, and again I can't
believe the stuff has any effect on the glass. But it does. The microscope
records my progress and it takes me a mere thirty minutes to finish up this
stage.
My mirror feels sooooo smooth now. The Grinding Guy takes note of
my rapid progress (I am now two stages ahead of the last person) and
remarks that I seem to be in a hurry to finish my mirror. As if on cue,
I ask him what would happen if I were to simply go out and get the mirror
aluminized right now, skipping the polishing stage altogether. What kind
of images would it form?
For the first (and possibly last) time since we have started this project,
the Grinding Guru has no response. I am not sure if this is good or bad.
I feel like I've accomplished a lot today. One more grit to go, and then
it's onto to pouring the pitch lap. Many of you have already written in to
tell me how much you hate the polishing stage. I'll see for myself in a
few weeks...
Chapter Six: Finishing up the Grits, September 10, 2000
If the 30 and 12 micron grits seemed finely-ground, the 5 micron stuff is
even more so, almost like a white liquid. It's so fine that the only way I
can use the stuff is to mix it up in a slurry in the water bottle. Then I
just shake up the bottle and squirt the white liquid on the mirror. While
grinding, it feels like it's not doing anything at all, but the microscope
tells me otherwise - real progress is being made here, and in about 2
hours I'm done.
The end of the grinding (or the beginning of the polishing, depending on
your point of view) arrives uneventfully - for me. Others have not been
so lucky. One member discovered a crack near the edge of his mirror.
Cracks have a tendency to expand and propogate (think about the cracks
you get in your windshield) so he went to the grindstone and ground down
his mirror, past the crack. The problem is, he had to take so much off,
the surface of the mirror is now shaped like a "D" rather than an "O".
Another member lightly dropped his mirror on the tool while grinding -
something I had done dozens of times myself to my own mirror - only to
discover that he had chipped off a tablespoon-sized hunk right off the
surface of his mirror. Ouch. He is examining his options right now and
may or may not continue with the rest of us.
Chapter Seven: The Pitch Lap, October 8, 2000
Pitch is a mysterious substance made up of rosin, bees-wax, turpentine,
and "other stuff" that sits in a small cardboard container looking like
brown glass. Pitch is not a solid, it's a liquid with a very high viscosity.
Add a little heat and it turns into a fluid.
The usual way to handle this is to heat the pitch in a small used coffee
can and pour it onto your mirror or tool. But the Grinding Guy has other
ideas, and his solution is simple and elegant. I wrap my tool with 2" wide
masking tape, making a small reservoir for the pitch, which was crushed to
pieces with a hammer. The whole thing gets tossed in a small oven and is
heated until the pitch melts.
Crushed pitch on the tool, ready
to heat (mirror in back)
Twenty minutes later the tool is taken out. The pitch is allowed to cool,
and I add a small slurry of cerium oxide to the top that will act as the
polishing compound. When everything is dry, I peel off the tape and the
pitch/cerium oxide layers stay put. The pitch looks like chocolate pudding,
while the cerium oxide layer looks a little like frozen chocolate milk on top.
Together, they look like one of those double-layer Jell-O instant pudding
mixes (Hmmm....when I get done uploading this segment, I think I'm going
to get myself something to eat.)
Soup's on! The melted pitch,
20 minutes later
Facets are cut into the still-cooling pitch. I used the edge of a chopstick,
but anything sharp and long would have done the job. After the facets are
cut, the tool takes on the "brown squares" look that you have all seen in
the textbooks.
From here on, it works just like the grinding segments. You work the mirror
and tool over each other and test periodically. Polishing takes a long
time - I was told to figure 5 or 6 hours for this - and takes more effort
as well. The pitch/cerium oxide don't exactly glide over the mirror. It
feels a bit like pushing and pulling two tightly-sealed plungers against
each other. My arms get tired after only a few minutes, and I have to
take frequent breaks.
The Grinding Guru takes a
few strokes on my mirror
"It's not bad," he says
After a few hours of this, I get my first taste of what this work is all
about. We take my mirror outside and reflect an image of the sun onto
a piece of paper. I see a white round disc with clearly-imaged clouds
rapidly moving across the surface. I'm thrilled in a way that I cannot
describe. I go back to my polishing with renewed vigor. The end is in
sight!
Chapter Eight: Fine-Polishing and Testing, October 30, 2000
Polishing is hard work. The mirror and pitch/tool feel as if they're vacuum
sealed against each other, and I find I have to use my upper arms to get
them to move. Pitch is strange stuff. It's brittle and chips easily when
cool, but glides and shapes itself like hard caramel when worked. After
about an hour and a half, the channels I cut close up and the pitch lap
has to be trimmed again. I'm pretty bad at trimming the lap - I made
several chips in it. Fortunately, it doesn't matter once polishing begins
again in earnest.
Polishing the mirror
Note the channels are still wide open
This takes a long time. It takes over five hours to polish the mirror
until its surface glows with a shiny sheen. The center polishes out
first, the edges last. Watching the "frosty" look at the edges slowly
go away was like watching paint dry. Every time I thought I was done,
I'd have to go back for another half hour or more. Ugh.
Testing
A Foucault tester is a very simple device. A light shines through a
slit, hits the mirror, and reflects back towards a knife-edge where
you look. As you move the knife edge across the field of view, you
magically see the surface of your mirror magnified many, many times
over. Little hills, valleys, and zones become obvious. The test is
so sensitive that a medium-sized cat walking near the tester or mirror
will induce heat waves that obliterate the view for several seconds.
We spend close to half an hour setting up the mirror and Foucault
tester -everything has to be aligned perfectly, which is a major pain-
and, in the darkened room, I get my first look at the condition of the
mirror which I've been working on for over five months now. Slowly,
with my heart in my throat, I slide the knife edge into place and take
a peek.
Judge, Jury, and Executioner:
The Foucault Tester
The surface looks smooth, and the edges are clean. There is no turned-
down edge, the bugaboo of first-time mirror makers. The mirror is also
not hyperbolic, which would be difficult or impossible to correct. The
only aberration I can see is a half wave hill, or bump, in the center of
the surface, which I am told is very easy to remove.
Figuring
A 6" f/8 sphere departs from a 6" f/8 paraboloid only very slightly. How
slightly? One method suggested is to take seventy-two 3/4 length "W"
strokes on your spherical mirror, and you should be close to a paraboloid.
That's right - of the tens of thousands of strokes taken on the mirror over
all these months, it comes down to a mere seventy-two strokes on the
tiniest grit of all.
This takes all of two to three minutes. We let the mirror cool for several
hours and a reading is taken. The hill has been cut in half, so more fine
figuring is necessary. I take another one hundred "W" strokes and leave
until the following morning. The waiting is necessary to allow the mirror
to fully cool. Even the heat from your hands, or the friction from the
pitch lap, can throw off your results. We confirm this by looking at the
mirror right after figuring. It's a mess. It "shows" a huge depression
and severe astigmatism.
But the following morning, the mirror looks completely different. Running
the numbers through "Tex", the program developed around Jean Texereau's
classic test methods shows the mirror deviating only .10-.13 waves O-P
center to edge. There is a minor zone about 2.45 inches out from the
center of about .10 waves. This zone is small in magnitude, but rather
steep and far enough out to be of minor concern.
We talk for a bit about this zone. It may be possible to fix it, but Sir
Isaac's pesky laws come into play - for every action, there is an equal
and opposite reaction. Taking out the zone may induce another aberration
in another part of the mirror. Due to this, the fact that this is my first
effort, and the apparent excellent condition of the mirror otherwise, I
decide to leave well-enough alone.
First Light With the Uncoated Mirror
Early the next morning, I install the mirror in the cell of a discarded 6"
f/8 Newtonian owned by the club. An uncoated piece of polished glass
reflects 5%-10% of incoming light, so I'm effectively looking through
a 1.5" - 2" telescope. It's enough.
The star test reveals slight undercorrection, perhaps a bit more than the
Foucault tester showed, but quite acceptable nonetheless (I was warned in
advance to mentally double the wavefront error calculated by the Foucault
tester to be on the safe side - this turned out to be good advice.) The
last quarter moon looks pinpoint sharp and I'm thrilled like a kid on
Christmas morning. I pump up the power on Jupiter and Saturn - silly
considering the dim images - but the planets hold up even at 190X.
I'm satisfied with the figure. Off to the aluminizer it goes!
Aluminization, October 19, 2000
I visited Research Service in Seabrook, NH and dropped off my mirror
for aluminization. Visible were about ten evaporation bell jars in
various states. The aluminization process is simple in concept, but
the equipment is extremely expensive.
"Cool! Can I watch it work?"
The Evaporation Tank
These guys did a great job (I had my mirror the next day.) Research
Service is run by Don and Steve Jaynes at (603)-474-9332. Much of
their business is done in the institutional/commercial market; their
work with amateur astronomers is done as a courtesy to us, at
rock-bottom prices. If you do wind up using them for your own
project, please keep this in mind and do your best to be a "low-
maintenance" customer. We want these guys around for a long
time.
The completed mirror, October 20, 2000
Like marathoners at the end of a long race, our ATM'ers are spread
out all over the field - I'm finished, but one very careful member is
still patiently removing grits at the 30 micron stage. For me, this
may not be the end of the line, as I am considering building a Dob
mount for the mirror and using the whole assembly as a star party
telescope. If I decide to go ahead with the project, you'll be the
first to know.
Click Here to see the conclusion!
Back to Home Page