Shards of tempered glass (the Shards series)

There’s been some interest online lately about doing chunky, frosted glass vessels like the ones pictured here. They’re something I’ve been experimenting with ever since my parents’ big glass display doors came down.

My parents bought a home formerly occupied by a collector of antique dolls and old china. She’d housed about 300 dolls inside a massive mirrored shelving system with huge and dangerously unstable tempered glass sliding doors.

The only way to get them out was to shatter them, which resulted in about 10 garbage bags full of broken tempered glass. What was nice about it was that there was a LOT of self-compatible glass there (you can’t be sure that two separate float glass sheets are compatible).

What was bad about it was that there was a LOT of it, and I did a lot of size-sorting and washing to get it into a usable state.

After that, however, the experimenting (and the fun) started.

I’d never fused with float glass before, and quickly found that it has very (VERY) different characteristics than glass especially made for hot/warm working.

For one thing, it’s a lot stiffer, it doesn’t tend to soften and spread as much, and even though its working temps are only 50-100 degrees above, say, Bullseye glass, it takes a lot more heatwork to make it do stuff.

I found that float has a (much) stronger tendency to devitrify, i.e., begin to crystallize for a frosted look, and there’s also that little compatibility issue.

Float glass fusing artists, i.e., people who work exclusively with float glass for architectural purposes or whatnot, tend to order their material by the case from the manufacturer. They know exactly what they’re getting understand the formulation, and experience many fewer surprises than I’m talking about here.

Fusing with float glass scrap is a whole ‘nother ballgame. It’s a lot of fun, you can make some absolutely gorgeous pieces with the stuff, but there are some rules.

Glass you get from a window or shelf can come from anywhere. It might not even be float glass (there’s also plate glass, for instance, although it’s a lot less likely, especially if you live on the west coast).

Float manufacturers, moreover, are obviously not going to maintain a fusing compatibility standard (well, heck, they don’t do that for stained glass, either… 😉 for something they’re planning on using for a window or coffee table top.

On the other hand, float glass is a LOT cheaper than fusing glass. And in my case, it was free. It’s clear and smooth, and if it’s tempered, as it was here, it’s in interesting little cubes, squares and–where the fractures hadn’t yet completely realized–interesting-looking shark-tooth shapes that caught and bounced the light much better than the usual conchoidal fractures of regular glass.

The 1 sheet rule: 1 sheet = 1 project. Because you don’t know the source, you also don’t know the compatibility. Rather than spend time testing each individual glass shelf/door/whatnot to make sure they’re all compatible, I don’t combine different sheets of glass.

It can limit the size of the pieces you make, but that’s better than losing a piece to compatibility issues. (The ice sculpture on the left was as much as I could get out of an old glass display shelf)

Clean first, shatter later. It is one more pain in the neck to clean shattered tempered glass thoroughly. Ask me how I know this. Cleaning the glass really, really well with Dawn and/or denatured alcohol removes potential nucleation sites from the glass surface, reducing the level of devit.

One firing is better than two. To keep the glass as transparent as possible, I tend to do these pieces in a single firing, which means I lay the shards directly on the mold and fire in place. Multiple layers of float shards fuse together nicely and in the first firing either stay transparent or acquire only enough devitrification to be interesting.

Minimum of THREE contact points in a tack fuse. I lay up the glass so that every piece has at least three (preferably four) contact points with adjacent pieces. I’ve found that pieces with one or two contacts tend to break off with use.

The smaller cube/crumbles can be piled into a mold and leveled, which makes a nice, faceted look, but I tend to prefer combining smaller and larger shards in a kind of mosaic-and-grout style. I’ll often build “bridges” of rectangular pieces underneath the larger shards, which helps make the piece more stable.

A more sharply vertical mold either needs to be done in two firings (fuse and slump) or I use dual molds.

The cup at the top of this post, for example, has an inner and an outer mold (thanks to my mother the ceramicist). I carefully assemble the shards inside the outer mold, something like building a jigsaw puzzle, put a second layer in the bottom for stability, then place the inner mold in the glass.

In the picture, I’ve got a piece of eighth-inch fiber paper between the inner mold and the glass, to take care of expansion issues. I still had a fairly high rate of cracking on these, though, so I’m now using a thinfire liner with a sand/plaster core.

It still presses the glass in place, but isn’t as abusive on the cooling tackfuses. Big deal with that technique is keeping sand from getting trapped between the shards.

The single-layer cups are my favorite in this series–they remind me of ancient Roman glass. But they are EXTREMELY fragile unless very well sintered together, and probably more curiosities than actual saleable items.

The alternative with this method is to use multiple layers of glass all the way up the cup. It produces a different look, heavier, icier and less delicate, but you could probably also bounce this on the floor with relatively little damage (except to the floor).

I did this with nested stainless steel mixing bowls, kilnwashed, with about an inch in diameter between the inner and outer bowls. Since the stainless is so much lighter than the ceramic, I filled the inner bowl with sand to keep the glass compressed against the outer mold.

Taking this to extremes, I’ve also been making what I call “glassnests,” like the bowl at the top of this blog, which can be 5-6 layers of shards in between the inner and outer molds. The technique produces a very cool crushed ice look with lots of edge detail, as much as two inches thick on the sides.

These probably aren’t to everyone’s tastes, but there’s something very seductive about the solid mass of glass.
The Shards have been really popular with friends and relatives; the smaller trays make nice thank you gifts and the big, 20-inch round trays are a hit as holiday presents.

I’m just starting to mess around with adding colors to this style, and figuring out how to get a similar look with Bullseye glass.

On firing schedules
Many people have asked me to post firing schedules, but I’ve found there’s a lot of “your mileage may vary” in tack-fusing tempered float. I can’t control a whole bunch of variables, such as:

• What kind of float you’re using
• Whether you’re obtaining float from a consistent source or doing what I do, i.e., getting free batches of scrap from wherever (and therefore have absolutely no control from one sheet to the next)
• How big the tack-fuse components might be (it really matters)
• How your kiln heats (and even how big your kiln is or how far the float will be from the elements)
• The desired final texture of the glass (sharp edges? rounded edges?)
• How well you can clean the shards (again, it really matters)
• The degree of clarity/frost you want in the end result

I tend to prefer slow ramps with lower process temps, and prefer the “frosted” look that results, i.e., I get some devit. If you go faster through the devit zone your glass will be more transparent, but I’ve found that you also wind up with less component adhesion if you want to preserve the cubed/faceted look of the original glass.

Letting things cook at a lower temperature for awhile seems to help the components settle and fit into each other better, so that you need less melting-in to get things to stick.

My top process temps range from about 1380F to 1480F. 1380F keeps the pieces very sharp (you could cut yourself on the edges of big pieces) and barely tacked together. That look requires a long hold with a slow upramp, and works best with smaller components.

In my Skutt bathtub kiln, a top process temp of 30 minutes at around 1480F will produce a lot of softening and a very stable piece that’s relatively clear, but I don’t like its overmelty look as much.

Beyond 1480F I don’t get a whole lot of difference in the look, maybe a little softer; this glass doesn’t seem to move much.

My ideal working range (at the moment) is about 1425F-1440F, depending on the size of pieces I’m using. I tend to ramp initially at about 500dph, initiate a much slower “settle cycle” (in regular fusing you’d call this a bubble squeeze) across 1200F-1300F, and then move 50-200dph up to my process temp.

Addendum: There’s a discussion on annealing in the comments that’s instructive. Please note that these are TACKFUSES, which means that you’ve got cooling glass contracting and pulling against itself in every which way, setting up all kinds of strain opportunities.

Annealing therefore needs to be slow enough to allow all those potential stresses a chance to relax. That means your schedule will be much slower than the actual thickness of the piece would suggest.

Float glass seems to have a greater ability to tolerate strain without breaking than standard fusible glass (in my experience, anyway), but why risk it? If you’re going to play in the float glass fusing world, get yourself some polarizing film, do some tests, and figure out where your float glass comfort zone should be.

Hope that helps, but as I said, your mileage will vary. I guarantee it.