Making a transom window

Thickness isn’t the only consideration when engineering a project in kilnformed (fused) glass. The shape of the object can be important, too. If you’re making a long, thin panel that’s rather thick, such as the transom window I recently made for my parents’ bathroom, you’ll need to understand how the glass moves and responds throughout the firing cycle.

I don’t typically do scenic fusing stuff, but the problems posed by making a thickish, long, thin, rectangle were interesting and (I thought) worth sharing. The window shown above, which is installed in a transom over my parents’ bathroom door, measures about 10″x36″ and is about 3/8 inch (10-11mm) thick.

Note: I orginally published this post in 2006, about a year after I made the window. Somewhere along the way it was lost in all my site updates and host-switching. There’s good stuff in here, though, so I’ve retrieved the lost content and republished it.

Designing the window

Mom and Dad had moved to the rainy, misty Northwest but didn’t really care for its continuing, gloomy drizzle. The master bathroom in their new home was spacious, but dark and rather depressing. They needed more light, so my brother-in-law Jerald opened the area above the door for a transom window, and I planned to surprise them with a piece of glass art to fit in the opening.

My parents’ old house had had a huge flower garden spanning nearly a half-acre and Mom missed being surrounded by flowers and sunshine. I decided to bring sunny gardens indoors with the window, and Jerald would add lights behind it to mimic sun. All told, it took a couple of weeks to make, with a lot of other projects and dayjob happening in between.

Concept. Mom loved agapanthus (Lily of the Nile); growing up in central California, our garden was full of these glowing, purply-blue flowers. I wanted to give the effect of an overgrown jungle filled with agapanthus. The window would be viewed from the floor, about 8 feet off the ground, and needed to be as transparent and pass as much light as possible. With luck, the installed window would give the appearance of a giant flower garden behind a very tall fence.

MOST artists start with beautiful sketches of the proposed work, on paper, which the client approves. I can’t draw to save my life, and this was a surprise for Mom and Dad, so–as usual–the only sketches were in my head, and rough ROUGH outline on the base glass. I drew up a blueprint in AutoCAD, but aside from that, this would be a build-as-you-go project.

I’m adding my project notes to this post, so that you can see exactly how I did this, the glass I used, some of the issues I encountered, and the firing schedules. If you really want to dive into the nitty-gritty, download and review this PDF.

CAUTION!!! YOUR MILEAGE MAY VARY!!! Depending on your kiln, the materials you use, size, a WHOLE BUNCH of parameters, these schedules may or may not work for you.

Download my project notes. And please, let me know if you’d like to see more of this kind of thing…

Choosing the glass

The light coming into the bathroom was a cold greyish-blue (hey, it’s Portland!), which needed warming, so I used its complement, Bullseye’s Light Coral Tint, for background. Colors in nature are rarely just one, pure tint, though, so I wanted to intersperse Pale Yellow and Crystal Clear to create a sun-dapped effect. I’d add Light Amber and (VERY sparingly) Medium Amber to punch it up. I chose to use coarse frit for maximum control over these pale, transparent tints, and added a little crushed clear irid and dichroic glass–not much–for sparkle.

You must be careful with sparkly irids and dichros in a frit painting. Their beautiful iridescent effects come from metallic coatings vaporized onto the glass; heated long enough, the coatings will rise to the surface of the glass, creating a melted plastic sludge effect. I was planning to make this piece about 3/8 inch thick; by putting these chunks on top of the bottom-most layer I hoped to seal them into the glass and keep them from reaching the surface.

The window would be viewed from two sides; the kilnside would have a different, less glossy texture than the glass on top in the kiln, and I would most likely be firing both sides face-up at some point. I used 6mm Tekta as a base glass for my “painting,” hoping it would give enough thickness to retard irid breakthrough when I was fusing bottom-side up.

Most of the decorative elements–colored glass petals, leaves, whatever–would be sandwiched between the two panels, a sure-fire recipe for trapping bubbles as the top glass melts down on all those air spaces. To prevent that, I would fuse everything flat and THEN do the sandwiching in a final fuse. I chose a 3mm piece of clear Bullseye glass for the top/front panel.

I’d fuse decorative elements to both top and bottom of each piece. The clear glass between the elements would space out the elements, giving an illusion of depth. I’d try to exaggerate that by putting the darkest colors on the inner-most layers. Then I’d put successively lighter colors on top and add the very lightest colors on what would be the outside surfaces of the glass.

Dark colors recede; light colors come forward. If it worked, you’d be able to view some depth from EITHER side of the panel, a necessity for a two-sided window. Bullseye’s got a wide range of blue, green, and purple glasses, so I’d have enough choices to add my desired dimensioning.

Planning the firings

I set it up in two panels, a back panel with 6mm Tekta, with elements fired on both sides to give some dimensions to the piece, and 3mm front panel, again with elements on both sides. I’d fuse the base panel with decoration, turn it over, add decoration, and fuse it again. Then I’d decorate the second/top panel in a third firing. I’d put the two panels together, add decoration to the top surface, and fire a final time to bring the panels together.

That meant at least four firings. Bullseye tests its glasses to just three firings; the more the glass is heated and cooled, the more chance you have for the glass chemistry to shift and potentially lead to cracks, devitrification, or off-coloring.

I wasn’t worried about that on this project. In my experience Bullseye glass is pretty stable 8-10 firings in the cooler colors I was using, and in the warmer tones. My biggest problem, again, would be the irid/dichro coatings rising through the glass. With every firing, the coatings would rise. I didn’t want to spend a lot of time drilling out sludge and refiring, so I would try to apply the least amount of heat energy (AKA heatwork) necessary to achieve the fuse. And I’d keep my sparkly glass particles small, so that if a few did end up on the surface there wouldn’t be great honking pieces changing reflectivity.

OOOPS!! How do you get this thing into the kiln?

Now for Tricky Part #1: I can now definitively state that 10×36 inches is the absolute, penultimate size limit for shoehorning a glass panel in a Skutt GM1414 kiln. Or, probably, most kilns of the bathtub ilk. That’s the problem with an oval kiln: You THINK you have an enormous amount of room, but the longer the glass, the narrower it must be.

The GM1414 has an interior footprint of 41.5×24.5 inches. Take away a couple of inches for air circulation, and the practical size is about 39×23. Realistically, you’ll be limited to fusing a glass circle of about 22 inches. A 10×36 inch piece of glass needing dams will literally be scraping the sides of the kiln at either end.

I tested the setup, and it was in fact so close to the side elements that I worried about annealing issues with cracking. On top of that, this was a loooooong piece of glass, with different colors and styles of glass on the top and bottom, resting on a kilnshelf. The shelf would insulate the glass, causing it to heat and cool a bit more slowly than the surface, and the dams/edges would do the same. The elements on the side would mitigate insulation to some extent, but more than likely the top surface of the glass would have a faster heat-up/cool-down rate than the rest.

Still, the fit was so tight that I couldn’t do my usual mullite brick dams around the edges. Instead, I cut long vermiculite board strips, and literally wedged them against the kiln walls to hold all that glass in place. It was only partially successful; I had minor glass breakthroughs in a couple of places.

Tricky Part #2

…and there came another issue: Glass expands as it heats, and contracts as it cools (that’s what the whole COE “coefficient of expansion” explains). The underside of the piece, on the shelf, would cool (and contract) more slowly than the top.

Normally, in more symmetrical pieces, this isn’t a huge deal. A 10×36 inch glass, though, has a nearly 4:1 ratio, length to width, i.e., it’s really long and skinny so there’s a lot more to contract in one direction than the other. You can have problems with the glass bowing up, off the kilnshelf, as the edges cool and contract faster than the center. If it gets too extreme, you could actually add enough warp to stress the glass and cause a crack. Hmmmm.

Fortunately, Bullseye Glass’ BEcon glass artist conference happened right in the middle of this project, so I had my pick of many really knowledgeable glass artists for advice.

Richard LaLonde provided the best hints. He advised me to add a “relax” cycle to the firing cool-down. I’d start the annealing process, but instead of going straight down to the annealing temperature for a long hold, I’d drop the glass temperature briefly below the strain point, then heat it back up to 50 degrees above strain point and hold for 30 minutes or so. That would allow the glass to relax into its final flat position with at least some of the stress relieved.

My test firings of small samples had definitely shown some bowing, so I tried it. Must have worked–the final piece was strong, without cracks, and perfectly straight. The glass did poke through its dam on one side, bleeding a bit, but that was easily fixed with the Alpha portable grinder. Ten minutes and an 80-grit diamond wheel worked wonders.

Transported the whole piece to Mom and Dad’s, hefted it into place…and discovered that transom openings aren’t always straight (surprise, surprise, thanks a lot, Jerald).

I’d brought the Alpha along, so I ground and fit and ground and fit, until I realized that there was a LOT of glass to remove on one edge. My friend John Groth trimmed it to the exact size I needed. (Yay for waterjet cutters!)

I eased the finished window into the space so that it still had a little room all around. (You do NOT want to wedge glass into a too-small opening; as the wood expands and contracts, you can stress and crack the glass). I left about an eighth of an inch of space on each side, using trim to cover the gap and fasten in the glass. I didn’t want to glue it in place; Mom wants to take it with her if they ever move. This way, they simply remove one side of the trim, pull the window out, and replace it with plain old glass.

I wouldn’t just call it the greatest art I’ve ever made; like I said, I’m not really into scenic glass windows stuff (although I love other people’s work in the same vein), but it fulfills the intent. I’m on the hook with my sister’s family to produce a couple more of these for the house they’re building. Designs to be determined (I think one of them will have pumpkins in it), but they’ll be coming up in the next year (or seven, at the rate my sister is designing that house).

And now back to sculpting and casting…