Herb's Glass and Scratches at fine grinding

During 2005 I'm working on an 8-inch f/5.5 mirror, as part of the STAR Astronnomy ATM group. This page is about my grief over scratches on my nice little mirror during fine grinding with a tile tool; followed by some homework and sage advice on how to avoid these problems, and how to fix them. Much of the content is from discussion in November 2005 with Gordon Waite. I've quoted him with permission.

This Web page is PRELIMINARY PENDING REVIEW, and was last updated Nov 17 2005.

  • Introduction
  • Considerations
  • Searching the Web on fine grinding scratches
  • Avoiding Scratches
  • The Nature of Scratches
  • Lessons learned


    During 2005 I'm working on an 8-inch f/5.5 mirror from start, under Gordon Waite's ATM sessions, as described in the linked document. At this stage of work, I'm using a dental plaster and tile tool. On November 7th 2005 in the course of fine grinding from 12 micron to 5 and then to 3 microns, I got some small scratches which I removed along the way. But at the end of 3 microns I had a small ding. I took it out with 5 micron, went back to 3 micron, and then got many ugly scratches. I got a large scratch down the center of the mirror, almost 2 inches long. I decided to use 5 micron to remove it, but after that the scratch was not much better AND there were two smaller scratches. I knew at that point to quit for the evening.

    During my drive home from that ATM session, I concluded that I was scratching during removal of the mirror at the end of the wet. In talking to Gordon Waite about this, and reading on the Web and in the usual classic books, I found it was a common experience to scratch during fine grinding with hard tools. This document summarizes that discussion and those readings.


    As I was getting longer wets with finer grits, I probably let the grit get too dry, and the channels too shallow. At the end of the last wet, when I decided to clean and evaluate the mirror surface, the mirror could not be lifted off the tool, it had to be slid off the side! This is tolerable with coarser grits, but with this fine stuff there is more suction (and so more pressure) AND also a better fit between tool and mirror. These all encourage excessive pressure. Any slight bit of broken tile or glass, or possibly a sharp tile corner, could create a scratch. That long scratch on my mirror was across the center and in the direction I pulling off the mirror.

    What annoyed me that evening, was that I had hints that this might happen. Gordon asked if my tool was channeled: if it was that would have reduced the suction. When I started 3 micron, Gordon specifically mentioned "stop work at the first seisure". I did bevel again the edge of the tool but I'm not sure that prevented more scratches. And I did not tell Gordon about my troubles in pulling the mirror off the tool at the end of the wet. I thought I was careful enough, it had not caused scratches at coarser grits - and I did not consider the small scratches I had earlier in the evening.

    That evening, all I could think about was that half an evening's effort was WASTED, I'll have to go back a few grits and fine grind again. But I waited for a better mood to check that scratch against my surface gauge blocks, see if I can determine what size of grit I'll need to remove it. And I resolved to do my homework, talk to Gordon Waite, and avoid these problems in my next session.

    Searching the Web on fine grinding scratches

    A bit of reading the "classics" on mirror making (Texereau), and surfing the Web the next day for "fine grinding scratch", informed me further. Some posts on the ATM mailing list suggest that one should bevel ALL the tiles at this point; some say just the edge tiles. Some posts suggest filling in most of the channels with wax, or epoxy; some (including Gordon) say this is not necessary or even encourages suction/sticking problems.

    Beveling is mentioned, but some say it can CAUSE scratches due to the debris left. Many sites say it's important to clean EVERYTHING with grit changes. Gordon cautions to AVOID scratches during removal by changing these fine grits WITHOUT removing the mirror, as he'll note later on this page.

    Some ATM'ers (like Mel Bartels) say there is too much risk of scratches to use the hard tool at 5 micron and 3 micron. This caution may be due to some brands of grits which have larger-grained contaminants.

    Sticky tools

    In reading Web posts about fine grinding, it's been noted several times the DRAMATIC change in cohesion when using fine grits, concurrent with the increased time that a "wet" is wet. A common error of practice seems to be that longer-lasting wets and the lack of scratches during prior, rougher grits encourages a number of risky attitudes and actions.

    One is an attitude to extend the wet past the point where the tool dries or is about to dry. A number of people point out that the WORST mistake when finishing the wet is to let the tool stand motionless on the mirror for any period of time. The consequence is greatly increased adhesion/cohesion. That's my experience and others as well: when the wet finally dries, it dries FAST.

    Now that the tool is stuck, the next attitude is that "I slid the tool off before without damage, I can do that again". My own experience is that the forces of contact are much greater than at rougher grits. Consistent with that is the greater contact between tool and mirror. More force and more contact are recipes for scratches from any contaminant or sharp edged tile. The apparent common workaround when tools are stuck, is to dunk tool and mirror in water and seperate them several minutes later. It's tedious but it's safe.

    Removing scratches

    I found many people reported scratches appear at several micron (3u or 5u) apparently due to this kind of handling when they are stuck. Those big scratches don't seem to come out with further polishing at that grit. So they go back to 12 micron or coarser to remove the scratch. There is some tradeoff between faster removal at coarse grit, and time spent getting back to a finer surface again through smaller grits. One ATM'er reported working at 5 micron for 45 minutes, or 90 minutes, with little success in removing deep scratches! These seem inordinately long efforts, but it's a consequence of trying to remove scratches much deeper than the grit's particle size. (Rules of thumb suggest a grit can remove material of the size of 3 or 4 times the particle size.)

    Avoiding Scratches


    I discussed my circumstances and my findings with Gordon. He offered several considerations:

    "The edges of the tiles get sharp if they don't have an adequate bevel, and unless the mirror is slid off exactly parallel, the edges of the tiles will give you a gouge." And, as you mentioned, you want to:

    1. Keep the channels deep enough between the tiles to prevent suction.
    2. Keep the edge of both the mirror and tool beveled.
    3. Don't separate the two between wets. Just slide the top piece to the side and squirt on some new abrasive.

    Gordon added from later discussion] that when changing these 12/5/3 micron grits, just offset the mirror, flood the tool with a squirt of water, then put on the next grit.

    To further protect from a sharpened rim, he suggested that the tool should have a good margin of plaster along the outer edges - for my 8" tool he suggested 1/4 inch. He's observed that the plaster rim will become fragile as the tool is ground down. Gordon suggested I "note the edge with the widest margin: keep that side on the mirror when sliding the tool off. Dental plaster is softer than glass and will not scratch it."

    Handling the tool

    When removing the tool, he says: "...offset it and give it a little squirt of the same abrasive, then swirl it a touch and THEN slide it off. At the end of a wet the interface may be a little dry...". A little dry is an understatement - I found it will lock up in less than a minute!

    Regarding cleaning the tool, Gordon cautions to "be really careful at this late point if you try to clean out the channels between the tiles. Your tiles are now pretty thin, and if you pop out too much plaster, you will loosen or remove a tile. That's a Very Bad Thing, because if this happens you only have two choices: A: build a new tool and go back to 220 to mate it up; or B: live with the surface as it is, and start polishing."

    The Nature of Scratches

    In our correspondence, I told Gordon I thought my scratches were due to interior tiles, as I rounded the edge tiles on the tool. Gordon responded at length about the source and the nature of scratches, how to reduce the risks and how to remove scratches at fine grinding.

    Kinds of scratches

    He says "I really doubt that your centrally-located scratch was caused by an interior tile. That's just about impossible. If a scratch is caused by a tile, it just about has to be at the edge of the tool. Nothing else sticks up high enough to make a scratch. Usually it's easy to tell the difference between a tile scratch and a particle scratch.

    A particle scratch will leave a "bumpy" track. As the particle rolls and chips the surface, it will alternately chip/skid/chip/skid and that kind of track is easy to identify. A tile edge scratch is usually a steady scrape. These come when you're pulling the top piece off the bottom piece, and you don't keep them parallel. A third kind of scratch is made during actual grinding (or polishing) when a person hangs the top piece out too far over the bottom piece, and it tips over the edge."

    Reducing risks

    "There's really no "trick" to avoiding all of these scratches. One just has to be careful. The hard part is to avoid developing bad habits. With mirror making, you'll get away with the bad habit 199 times out of 200. But eventually you hit that 200th wet, and your luck runs out.

    "Personally, I try to organize my procedures to minimize the chances where I can screw up. For example, when I'm doing the 12/5/3 sequence of grits, I never take the mirror off the tool. When 12 is finished, I just offset the mirror and flood the tool with a squirt of water. Then I put some 5-micron on there and continue on. No major cleanup. I do the same after the 5, just water flooding and going right to 3 micron. Also, I do far fewer wets than most people.

    There is one notable point about 3-micron, and even 5-micron work. A lot of people avoid 3-micron because of "the increased chance of scratches." Actually, if you are using tile tools instead of glass tools, and you don't stop to clean up between grades, there is almost no chance at all of scratching. Glass tools aren't channeled, and they tend to stick and grab a lot, even if you have a nearly perfect sphere. So people are constantly separating the two pieces, and eventually they get scratches.

    But most of all, people don't mix the abrasives with water and apply them with a squirt bottle. With the squirt bottle, all is good. But when people put the dry abrasive on the tool and then wet it, and then put the mirror on it, you're almost guaranteed to get scratches. The abrasive clumps up a bit with that technique, and it will create scratches. (The same thing happens with red rouge. It clumps and gives you sleeks.)"


    "Wow, I can't believe I've written all this about scratches! I've had maybe two or three scratches in all the mirrors I've ever made. Maybe I'm just lucky, I don't know. But I don't really sweat them, as they're easy to take out, and a scratch isn't as bad a mistake as is otherwise possible to make!!"

    As he said earlier: "Even the worst scratch can be taken out with 12-micron, or 25-micron at the worst. Figure 30 minutes with 12-micron, and 15 minutes each with 5- and 3-micron and you're ready to polish. So you've only added an hour of work. A cheap lesson, right?!"

    Lessons learned

    In retrospect, here is what I've learned. I should have avoided removing the mirror from the tool. When it's necessary to remove it, wet the tool again, and remove it promptly. When I could not lift the mirror vertically from the tool, I should have put both in a pan of water until they seperated. Then I should have reviewed the channels of the tool and rechanneled and ground the edges if necessary. While I scrubbed the tool with each grit change, I was lucky not to lose another tile. Also a personal rule for myself is to avoid work after midnight. Generally, I make many more errors at that time as mostly I am a "day person".

    I've been cautioned at other times not to do so much testing of the mirror: while it's an error of sorts my prime consideration is that I'm "testing the glass" as well as my own process of working glass. If I'm to make mistakes I want to make them with my FIRST mirrors, not with EVERY mirror.

    Mostly, I forgot the lessons of my prior work on refiguring my other mirror. But that did not involve coarse griding or a change to fine grinding; I repolished a poorly figured and surfaced mirror. So this was new territory for me. Many say that errors are educational. That thought did not fill me with joy that night, or the next day. But Gordon's discussions and advice inform me more about my own errors than I'd learn from mere success. The mark of a skilled crafter is not just in being successful, or in avoiding failure, but in knowing how to respond to errors and to learn from them.

    Web Links

    Herb Johnson

    Contact information:
    Herb Johnson
    New Jersey, USA
    To email @ me, look at my commercial Web page for address

    Copyright © 2005 Herb Johnson