Friday, November 4, 2011

Return to Polish

I had been blowing glass for several years before I became vaguely aware of idea of coldworking glass. In the MIT Glass Lab there was old ~12”-14” flat lap at the off the shop which I’m fairly certain began it’s life as a battered potter's wheel. The grit in the tray underneath the lap was fill with some grade of “mystery grit” that was filled with bits of glass and whatever else fell in within the last year. I had used the wheel to flatten tippy pieces and make them less-tippy. To remove any amount of material took forever and, at best, the finish looked as if I had dragged the glass behind my car. One day, my friend Nick showed up with a little glass piece that he had cut in half with a diamond saw and had polished the cut surfaces. My mind was blown for two reasons:

1.    You can cut things in half
2.    You can make rough surfaces smooth and glassy again

This might seem like no big deal, but having never seen the process it all seemed like magic.  Last night, I was in the Cold Shop preparing a blank for next weeks demo for my coldworking class and I thought it might be interesting to go through the process of flattening and polishing a relatively large piece of glass.

Before I get into the how to, I should mention a bit about the process shaping and polishing. Polishing is a progression of actions from course, medium, to fine. This relates to the size of abrasive grits that move from larger particle sizes to smaller particle size, but also the nature of process, more specifically, the inversely proportional relationship of speed and control.

The first tools we use will be the coarsest and also the most aggressive. These tools will be used to quickly remove waste and save time in the process, but one must be careful not to get carried away, the rate of removable can also speed the process of making large errors in a hurry. The tools we will be medium and at this point we are still in to business of removing material, but are now concerned with refinement of the overall shape. At this stage we trade speed for control. We are now concerned with the refinement of shapes and the trying and truing of our surfaces. Also, we will begin to pay attention to the condition of the surface and are weary of scratches and defects that would be laborious to remove in later stages. At last, we find ourselves in the finally in fine phase where our only focus is the refinement of the quality of our surfaces.

First Course:
Now, it only occurred to me to start taking pictures of the process after I had already made the first cuts (sorry). So we find ourselves, at the diamond saw after having cut the checked and cracked edges off a hot-cast billet of glass. After the first straight cut was made, the other edges were roughly marked for square and cut without using the fence on the sliding tray (the fence is only square-ish). I’m not really concerned with squareness or cleanliness as we are just hogging off waste and saving time grinding downstream.
The saw leaves the surfaces filled with deep saw marks and jagged edges. I always expect that I will have to remove at worst 1/8” of material beyond the kerf of the blade. Also I take the ragged edges to a very course 80 grit belt and “seam” the edge before I go any further. The seamed edge is more finger friendly and wont chip when you bring it to the flatlap. Seaming is a very important operation that must be repeated the entire process as your previous seams are ground away. I should mention that but as your surfaces get finer and finer, your seams too will get smaller and smaller (with the least aggressive grit that is still efficient). The flatlap is simply a very flat rotating steel plate in which a slurry of silicon carbide and water is fed. After we leave the 80 grit lap all marks of the saw should be removed and we should have relatively flat, squarish surfaces. This is mostly an eyeballing situation.

Middle Ground:

At this stage we really paying attention to the refinement of shape and accuracy. For my blank I am shooting for surfaces mostly square and totally flat on five sides ( I am keeping the air side of the blank as is, and as a result it will have a gentle curve). At the 80 grit I used a combination square to get the sides square within about ~1/16”-1/8”, now at the 220 grit wheel I’ll be looking for squareness from ~0-1/32”. Also I will be looking for a surface that is completely flat without crowning (a concave curve) or faceting. Both are accomplished by placing the blank on the wheel and establishing a consistent flat and then adjusting the squareness with adjustment of downward pressure and the orientation of the piece relative to the rotation of the flatlap. 
To understand how to adjust pressure we first have to understand the dynamics of the grinding. 
Here’s what’s happening on the surface of the wheel. Firstly the surface of the wheel is moving fast from middle to the edge, therefore the closer to the edge is more aggressive and will remove material faster. Secondly, the edge that is “upstream” will tend to be worn faster. This is partially hydrodynamic and partially human error. To keep the glass from flying away your natural tendency is to bear down on the front edge. There are two strategies to correct this error, one is to alternate is to regularly flip the piece back and forth working in equal amounts (always a good idea). The other is to angle the piece such that the piece is oriented to the rotation of the wheel at an angle (~45-60 depending on the size of the wheel) to balance the rates of removal from side to side. Also at this point you should be working away your seams just to the point removal and then adding them back with a very fine belt or pad. For curve surfaces I use a 600 grit SiC belt sander for flat I use a 325-600 prepolish resin pad on a magnetic flat lap Either of theses processes are smooth enough that they are ready for final polishing. Before going any further, the surfaces should be free of chips, scratches, checks, and should be perfectly uniform.

Final Countdown

Home stretch! At this stage we are fast approaching a crawl. We are trading our powered tools for the bacon powered one. Once you get finer than 220 grit, powered lapping (with a non-segmented plate) gets a bit dicey. Wheels tend to grip the glass and are usually less trouble than they are worth. Handlapping is simply the process of grinding with loose grit and a flat surface. In this case, a sheet of 3/8” flat glass (which is as flat as the curvature of the earth, which is flat enough for me). The grit is mixed with a bit of water and drop of soap (for lubrication and to break surface tension) and the glass is pushed over the surface in a random motion (circles, figure-eights, and zig-zags).

At this point we are simply refining the surface of the glass, reducing the size of the highs and lows of a rough surface with successively finer grits. On a larger surface the quickest most consistent results are gained by adding more stages of grits. I use 400, 600 and 800. You’ll find that the benefits of less stages are outweighed by the inconsistency of the work and drastically increased amount of time in the finer stages. 

There is one more stage before final polishing is to create is to smooth out the soft grainy lapped surface with a glossy “pre-polish”. This can be accomplished with loose pumice on a cork or felt wheel, by my preferred method is to use the belt sander 1200 grit aluminum oxide 3M trizact belt. Since the amount of material that is being remove is so insignificant, flatness is no longer a concern. The last stage is to use a cerium oxide compound to bring the glass back to a pristine polish. At this point the dry surface should be indistinguishable from a wet surface.

One the piece is polished and free of defects, once again the exterior surfaces of the object become almost immaterial allowing your gaze to enter the volume and look at glass itself- Magic!


Stereoette said...

GREAT post. I've been blowing glass a lot over the last few months and at least for my basic skills, cold working is a super necessary step. I think your suggestion of hand lapping anything about 220 grit will help me a lot, considering I find the 240 grit flat lap wheel to be my nemesis (I get crazy bevels!)

So, now, I'm excited about coming up with ways to improve my fused glass work with cold working...

Jeanette West said...

It was quite a long process, but totally worth the wait. Anything you want to get polished is really possible if you have a good polishing wheel to do the work. The glass is smooth again. Well done!

nielscosman said...

Hi Jeanette,


With the right tools and taking the appropriate steps the process to polish is actually fairly quick and relatively painless. I am a big advocate of making do with what you have. The same job could be done far fewer tools. In fact, with a sheet of flat glass and a full range of grits one could accomplish the same work with donut-power (by hand). Or you could just plop the glass down in a reciprolap, set a timer and come back when it's done.

Whatever the tools you are using,
the biggest struggle that people have with polishing is the attention to detail, and making sure that you are doing the work that was intended at each step. learning how to see good work is as hard as doing good work. Folks blow into a shop, I hurry to get things done as quickly as possible. They try to rush or skip steps but more often than not, it comes back to bite them in the butt many times over.
Slow is steady, steady is fast!