Firing schedules, untangled (part I)

Firing schedules are probably the single biggest source of confusion in kilnforming glass. Over the years (and a bunch of research, testing and listening to smarter-than-me glassists), I’ve developed strategies for schedule management; this series will share a bit.

As always, this stuff is based on MY experience, with my designs, my technical considerations and my understanding. I don’t promise that everyone will agree with all (or any) of it, or that it will work exactly the same way for you. Instead, consider it a starting point for your own explorations.

Join any gathering of glassists, and the alwayses and neverses just start a-rollin’ in. NEVER do this, or ALWAYS do that…

I call them “The Rules,” and you should forget about ’em. “The Rules” will get you into trouble, stop you from thinking, and just generally mess up your day. Breaking them will NOT earn a visit from the kiln police, no matter what horrified glass communities say.

Here are a few very common “rules” that I could happily bury (and feel free to add your own in the comments):

NEVER use pre-programmed firing schedules? Not true.

If you’re a noobie glassist who just took delivery of a brand-new kiln, you’ll discover it offers pre-programmed schedules for glassmaking, kinda like that “baked potato” setting on a microwave oven. WOW!!! Can it be THAT easy? You just carve into a couple sheets of glass, arrange them on your virgin (but kilnwashed) kilnshelf, hit the button and–a few hours later–out comes a piece of art?

Yup. Go right ahead and do it…for awhile.

Experienced glassists, stop screaming: Remember how scary (and expensive) glassmaking seemed when you started out? If you’ve never fired a kiln before, you’ve got a lot to learn, and a not-inconsiderable safety concern (1500 degrees in my KITCHEN? Are you KIDDING?). Pre-programmed schedules take some of that worry out of first firings.

If a pre-programmed schedule helps new glassists get over that scary hump, why not? They eliminate one huge variable (designing a custom firing schedule), allowing a newcomer to get a few successful pieces underbelt and fall in love with glassmaking. Blind someone with science, and we’ll scare them off.

Besides, there’s a good chance that a pre-programmed schedule WON’T fail on those first projects. Small hobbyist kilns hold small hobbyist pieces, so there’s just not that much that can go wrong with the schedule. By using a pre-programmed schedule, the noobie can focus on the basics: Glass compatibility, pattern, cutting, assembly, mold releases, etc.

When is it time to leave pre-programmed schedules behind and learn how to program a kiln? When it stops working.

If you’ve successfully fused nine plates and the tenth blows up in the kiln…congratulations! You’re becoming an experienced glassist! Your work has progressed to the point that your controller’s one-size-fits-all schedule needs to change. Now you can begin learning the difference between what you WANT to happen, versus what’s really happening, for each segment of your firing program.

At that point, dear instructors and glass pundits, you can start telling them to stop using those bloody, brainless preprograms. Not before.

The next firing of a piece is ALWAYS longer than the last? Nay.

Annealed is annealed and, last time I checked, glass can’t count. Why would it care whether it was being fired once, twice, or twenty times…as long as each previous firing left it well-annealed?

There ARE reasons to lengthen successive firing schedules, but they have more to do with the fact that the first firing usually combines lots of smaller pieces of glass into a single, thicker whole. Thicker or oddly shaped glass needs a longer anneal. Sometimes the glassist suspects that something–too many firings, abusive coldwork, previous firing misses–added new stress to the fired piece, and will tiptoe a firing schedule to process temps to avoid a break.

But past firing history by itself should have diddly-squat to do with picking a firing schedule.

There IS one thing to worry about with successive firings: Compatibility. Glass formulations can change with high or excessive heat and being heated and reheated more than the manufacturer recommends. Warm colors are more prone to this than cool, opaque glass can be fussier about it than transparent…but all glass will eventually change just enough that it no longer plays nicely with other glass.

Overfired glass may shrink/expand at a different rate than promised by the manufacturer, or its viscosity could change from the expected. It will no longer live in the “compatibility” range that allows it to be fused to other glasses without adding stress, cracking, or even exploding in the kiln. (And yep, I’ve seen that happen, with enough stress).

Bullseye guarantees its glass to three normal firings (you can look up the details on their website). Doesn’t mean that it won’t be perfectly happy being fired 20 times, but the odds of success go ‘way down.

Here’s the thing about compatibility, though: If the glass has truly become incompatible with its mates in a piece, no amount of schedule-lengthening will fix it. The only way to restore compatibility is to remelt it, adjust the chemistry, cool it into sheet/frit/noodle/stringer…or just go get another piece of glass and consider this a learning experience. (I have a lot of those)

ALWAYS add a hold at every firing segment? Not necessarily.

I was solemnly told this by my very first fusing instructor, “…because the glass must rest in between heating cycles to avoid cracking.”

Sidenote: This was the same instructor who told us that kilnwash wasn’t necessary if you used slumping molds designed especially for glass, and that ANY glass in his stained glass shop would successfully fuse to any other glass. The look on his face, as he pried glass shards from a dozen expensive molds and the walls of his brand new kiln, was kinda priceless.

The idea is that an extra five minutes will allow all parts of the glass to get within the required 5 degrees C (a bit less than 10F) to avoid adding stress or cracking the glass. This is kinda like stopping twice at a stop sign to make up for crashing through the last one.

It’s true that uneven heating (i.e., not maintaining thermal equilibrium, popularly known as “evenivity”) creates a thermal expansion/contraction problem that can shatter your piece. But wouldn’t it be better to simply maintain thermal equilibrium by using the correct schedule? If you need to add a hold at the end of each segment, you probably should slow down, and think about what each firing segment is supposed to accomplish.

There are plenty of good reasons to add a hold in a firing schedule:

• Allowing time on the ramp for the softening glass to drop from the center out, eliminating trapped air (the “bubble squeeze”)
• Processing the glass at full firing temps
• Anneal soak, giving the glass time to relax any potential stresses created by fusing a bunch of different thicknesses/shapes/colors together
• Drying out a mold
• Using certain controllers that require a hold for every segment

Adding a catch-up hold to make up for a bad schedule? Not one of them.

Low and slow is ALWAYS the best schedule? Not always.

Given how often I’ve written “go low and slow,” you’d think I’d agree with this one and, mostly, I do. Violence begets violence in a glass firing schedule: Change temperature too rapidly, and the glass will follow suit. It will trap more air, slump unevenly, flop (and potentially misalign) instead of settling gently where you want it, and you’ll have to add more heat, sometimes for longer periods of time, to achieve the desired result. Those higher temperatures can, as mentioned, shift compatibility, cause undesirable color changes, and exacerbate devitrification.

However, there ARE times when low and slow is simply a waste of time. You don’t need to slowly bring a bunch of frit up to process temperatures: Even if you thermal shock those tiny bits of glass, all shattering will do is make more frit.

There are, indeed, cases where low and slow might actually hurt the piece. Float glass isn’t formulated to resist devit, for example, so a long hold in the “devit zone” is more liable to hurt than help. And if your kiln has some serious hot spots, a very long anneal might actually induce stress (although that’s pretty rare).

Plus, there are times when you actually WANT the glass to flop down quickly, as in this tack-fused bowl. A long, slow softening would have rounded out and deformed those brittle stringers. I wanted just enough heat to tack the components together so that they locked and fell, then I wanted them to get them the heck out of the heat.

You can NEVER fuse/slump/firepolish in the same firing? Gee, sorry to hear that.

If I’d known it was impossible, I wouldn’t have made this plate or that bowl. Both were made using a standard slumping mold, in a single firing that fused, slumped, and firepolished the final piece.

Under normal circumstances, trying to accomplish a fuse and a slump in the same firing is a waste of time and often of glass. If the glass has enough heatwork (i.e., the total heat energy applied by a certain heating rate and temperature level) to fully fuse, it’s way past the point of slumping and will simply flow down to become a puddle in the mold. However…that’s NORMAL circumstances.

There are a few abnormal circumstances where it works. For example, Bullseye’s Square Slumper mold is one of my favorites because it has just enough shape to make a sophisticated-looking plate, but is shallow enough that the glass can’t get very far when it moves.

The green plate in the picture is made of heaps of leftover pate de verre mix, i.e., fine clear frit coated with a thin layer of transparent, colored frit powder. When I’m finished with placing a color in my refractory mold, I slap the excess onto a square slumper. When there’s about a half-inch of leftovers in the mold, I tamp it down hard and edge it with medium black frit, the softest, most forgiving border glass around. Then I fire it in a VERY controlled schedule that fuses, slumps, and firepolishes.

Of course, there’s no particular advantage to doing it this way over just fusing it flat and slumping it later. Combining those functions into a single firing is a lot more persnickety, with a much higher failure rate, and doesn’t save a lot of time.

It would be extremely difficult, however, to flat-fuse the lotus bowl in the picture, and then get it to slump into that particular configuration in a mold. Unless I cast it, the only way I could manage it was to cut and pre-fuse all the pieces, giving them a slight lipping, then balance them carefully in the mold, in the kiln so I didn’t disturb anything. I then fired EXTREMELY carefully, so that gravity slid the pieces into their final positions and fused them in place.

It certainly didn’t save any time–the bowl you see was my second attempt after a LOT of tests with scrap glass. Worse, the annealing schedule was so long and slow that it literally added days to the kilntime. (those little “stamens” in the center, i.e., frit balls, took eight tries to find a firing schedule that worked, and I’m still not entirely sure there isn’t stress in there). If you want the full story, check out the LotusBowl tutorial.

I do combine firepolish and slump in a single firing, mostly because it cuts down on firing exposure for the piece. It’s a fairly easy process; just coldwork the firepolish area to at least 400-grit and then slump a bit hotter/longer than you normally would. Once you’ve learned your kiln’s eccentricies and hotspots, it won’t take long to work out the schedule.

Glass ALWAYS slumps/fuses at X degrees? Nope.

Nope. Glass is an amorphous solid, like chocolate or wax, which means it doesn’t have a well-defined melting point.

A crystalline solid, such as a metal, has its molecules arranged in orderly arrays, all bonded to each other, kinda like prisoners in a cellblock lockdown. They can’t leave their cells, so their movements are limited to a single cellblock, what we call a solid.

Heat to a crystal is like handing a saw in a cake to a jailbird–apply enough of it, and the molecules stage a jailbreak. They roam around the penitentiary, causing all kinds of havoc, i.e., they “melt,” and become a liquid. (Keep on heating, and they can get out of jail entirely, and roam wherever they like, becoming a gas, but let’s just focus on melting).

In a crystalline solid, the bonds are the same, so it takes the same amount of heat energy to break all of them. Apply enough heat, and–whoosh–your solid cellblock turns into a puddle. The temperature at which that happens is the melting point of that solid.

Glass molecules aren’t arranged in an orderly fashion, and the bonds between molecules can be longer, shorter…whatever. Apply the same heat to a block of glass, and only SOME of the bonds will break. Other bonds stay intact, so what you get is softening, not melting.

Apply more heat, and more bonds break. As the energy levels increase, the glass gets progressively softer, flows a bit more…and will keep on doing that until it’s as fully liquid as it’s gonna get. It’ll do it in reverse when you cool it down, i.e., stop adding that energy. It’s why candles may droop a bit on a hot day but only turn liquid with direct flame.

The judicious application of heat over time is what lets us do with glass what we do with glass. Imagine trying to slump a piece of ice. (Admittedly, it’s a lot more complicated than this, but if you get softening vs melting, you’re a long way to figuring out your own firing schedules)

Anyway, that’s why two schedules, one that zooms 500dph to 1350F and holds for 30 minutes, and one that creeps up 200dph to 1225F with a 5-minute hold, might accomplish similar results. The AMOUNT of time the heat was applied can be just as important as the temperature(s).

You can literally shift glass transition from the solid to liquid phase by the way you apply “heatwork,” i.e., total amount of heat energy applied over a period of time. The fabulous thing about this? By carefully controlling kiln temperature AND time, you can dial in the result you want. Forget the “one temperature fits all” notion.

From a thermodynamics perspective, the concept of “heatwork” is a bit fuzzy; I’ve tried shoving X degrees per hour times X hours times whatever to build some sort of heatwork calculus and never managed it. There’s probably someone out there who has (so please PLEASE speak up).

Bottom line: If you ever hear someone say “this glass fuses at 1475F,” they’re either making a LOT of assumptions about the heatwork that led up to that temperature…or they don’t know what they’re talking about.

This schedule ALWAYS works, right? Sadly, no.

“BUT THAT SCHEDULE ALWAYS WORKED BEFORE!!!” may be the saddest words ever spoken by the glassist staring down catastrophe.

Whether it worked before or not, it’s not working now. Your glass cracked/didn’t slump/wound up in a puddle/stuck to the mold/exploded, so clearly, something didn’t work. So…what changed?

When I first started, I was absolutely certain that the disaster in my kiln had NOTHING to do with the schedule. Oh sure, I’d tacked on some frit balls, stuck another couple layers of glass underneath, stacked another kilnshelf on top to fire two pieces at once, used a new mold…but nothing CHANGED…right?

Wrong. Every so often the change is using the wrong glass, or a problem with the kiln. Most of the time, though, you’ve changed what’s inside the kiln. It may not be obvious, but something has changed…so you must change the schedule to compensate. I’m gonna get into that in part II.

So that’s it for mine; how about yours? And if you’ve got data to the contrary please, please share–I love learning this stuff.

In the meantime, Bullseye Glass probably has the best tutorials around for figuring out why things broke, and some absolutely marvelous video tutorials that are better at explaining this stuff than I’ll ever be. If you haven’t subscribed, do.