ogg-potluckHow do you give non-casters a taste of pate de verre-making..in less than four hours?

That was the assignment, anyway. It was the Portland chapter’s turn to host the Oregon Glass Guild’s annual state meeting, and we wanted to do something a bit special.

We decided on a theme of Stretch Your Wings, and gave it multiple meanings. First, we meant “stretch your wings by reaching out to the community.” Instead of focusing on personal enrichment, this time we’d make art for the community, a glass quilt to be installed in a local hospital. Everyone who came would make at least one 6×6 inch tile for the quilt.

Second, Stretch Your Wings would also be the theme for the quilt.

Third, we meant “stretch your glassist wings,” as in “try something different.” Since the vast majority of our members focus on fusing, no fusing was allowed. Instead, we set up workstations for sandblasting, stained glass, glass painting and casting, and invited attendees to make a tile using a technique they DIDN’T know.

Fourth, we wanted to stretch our own organizational wings. OGG has a lot of incredibly talented artists who are sometimes overlooked when we bring in outside instructors. It was about time we asked our own experts to coach the membership, with the idea that–if there was enough interest–a workshop could develop into a full class.

ogg-stainedglassWe held it in Chuck Franklin’s gorgeous big studio (thanks again Chuck!), OGG supplied BBQ and drinks, and we asked members to bring potluck.

It was an all-day blast and I was very glad I went. Everyone else seemed to enjoy it, too, and from the looks of things we got some beautiful tiles out of it. We even got families in on the act.

Three guesses as to who ran the casting workshop? Me, of course.

Developing the workshop

Objective: Take non-casters from raw clay to a well-filled mold in less than 4 hours.

Obviously, participants in a free 4-hour workshop won’t receive the full experience of a week-long pate de verre class with a skilled instructor. Still, I wanted to give them an idea of the processes involved, what kinds of problems routinely crop up when you’re designing a pate de verre piece…and in the end, ensure that they could contribute a beautiful tile to our quilt.

Another problem: The other workshops would produce uniformly quarter-inch thick tiles, ready for installation. It’s notoriously tough for beginners to get frit casting glass volumes right even when they’re NOT in a hurry. Yet if the tiles were too thick, or had relief designs at the edges, we’d wind up doing a LOT of coldwork–and possibly damaging the design–to make them fit.

I decided to make slightly smaller tiles, 5.5×5.5 inches, which could be attached to 6×6 inch pieces of float. The resulting tiles would be front-heavy, but they would fit. To keep the combined tiles from being too heavy (and also to save money on glass), they’d be kept low-relief and rather thin.

We shared out space for each workstation in Chuck’s studio and I decided that six students was about as much as I could handle there. I’d make a tile along with the students, as a demonstration, which would give us a total of seven tiles, exactly the number that would fit in my kiln in a single firing.

I had the local plastics place cut 7 12×12 sheets of acrylic for workareas and picked up some inexpensive adhesive spreaders (they’re cheaper than the metal ribs I use to shape and smooth clay). Then I bought 50 pounds of cheap, grogless clay. (Grog, i.e., particles of fired clay added to new clay to strengthen it, makes it harder to sculpt a smooth surface)

I didn’t have time to mix up my own plaster/silica, so I bought a couple of 40-lb tubs of the stuff from Bullseye, and a full sheet of clear 3mm glass. I’d use my own frit, and bring the maybe three gazillion clay brushes and tools I use for sculpting, along with tarpaper and miscellaneous stuff.

Modifying the pate de verre method

workshop-moldplan1With our time and money constraints, it wasn’t practical to have students completely pack their molds with powder and fine frit. Instead, I’d have them use frit to fill in the raised areas and surface of the tile to maybe 3mm.

Then we’d drop in two squares of clear glass to ensure we had enough glass volume in the finished tile.

By doing “pate de verre” (well, really drypack frit casting) only at the surface of the tile, where the color and shape matters, and then backing with clear glass sheet, we’d still get the pate de verre look.

We’d also ensure that participants could use opaque colors if they liked but–thanks to the thinner layers–the pieces would still have some translucency.


Box casting vs. handbuilding

ogg-fritpackI’d use a box method to make the molds instead of my usual handbuilding. In handbuilding, you slowly shape the wet refractory around the mold, creating a shell of even thickness. Handbuilding keeps your heat exposure as even as possible; otherwise you’ll have hotspots where the mold is too thin, and coldspots where it’s thicker, which can increase your chance of cracking as the mold cools.

Handbuilding also lets you surround the model with layers of refractory that do different things. For example, you can use a facecoat, next to the model, which lifts off easily from the glass and takes extremely fine detail well.

Or a strengthcoat with reinforcing fiber to prevent cracking and flashing.

Or a “failcoat” of plaster mixed with organic material; the organics will burn out early in the firing, leaving you with a swiss cheese-like, weakened mold layer that will crush as the glass contracts around an inner curve, preventing cracking.

Or filler coats of cheaper plaster/silica will save on expensive materials.

Stuff like that is much harder to do with poured boxmolds.

However, handbuilt molds also take longer to make and aren’t really necessary for small, relatively uniform relief tiles. A box mold would increase the risk of cracking and flashing (fins of glass that flow into mold cracks), but could be ready to fill with glass in 45 minutes, even counting clay removal.

So…box molds were the obvious choice for this workshop.

Dam those molds

tarpaper box

A simple moldbox of roofing tarpaper, with embedded wax models. Box molds aren’t ideal, but this one is fast and easy. Once the mold is set, the wax models are steamed out.

Box molds are typically built by constructing clay dams around the model, high enough to hold the refractory plus a couple of inches, and sealing them to the worksurface so they don’t leak and then pouring in the plaster.

Problem is, clay dams take time (and a lot of clay) to construct. Worse, I’ve never been in a new casting group yet where one or two clay walls didn’t breach, getting refractory all over the place.

We’d save time and reduce breakout by building our moldboxes from roofing tarpaper. Linda Ethier turned me on to roofing paper years ago, in her casting class, using it to construct open cylinders that she sealed to the work surface with clay.

I’ve taken it a step farther; instead of open cylinders, I set the model in the center of a piece of tarpaper, slit and fold up the sides, and staple it together into a moldbox (above, left). Dab a little clay into the corners, and you have a cheap, leakproof box for refractory.

The tension in the stapled sides will hold the mold together as it sets…but you save on refractory and get a nicer mold if you reinforce the tarpaper sides with boards.


This was the easy part; I’d simply dump 6-7 pounds of clay on pieces of roofing paper (which would prevent the clay from sticking to worksurfaces), give one to every participant and tell them to shape a clay cube about 5.5×5.5 inches by 2-3 inches thick.

The extra thickness would become the reservoir of the mold, giving us headroom to pack in additional glass. Powder can lose 50% of its volume as it compacts and smooshes out the air, so the mold walls needed to be high.

I’d encourage participants to build UP rather than carve into the block, to create easy-to-fill cavities.

ogg-windwomanThe whole positive/negative thing with moldmaking can be difficult to grasp until you’re hit upside the head with a negative space to fill, so this would definitely give them a taste of the real thing.

And to make sure we could actually get the clay out quickly, I’d prohibit undercuts or wispies in the model (unlike the first 6-inch pate de verre tile I ever did, below, which took FOREVER to clean.)

Color palette blues

Just when I should have been finalizing prep, I caught my cousin’s cold or flu or whatever and was pretty much down for the count, right up until class. I’d taken three days off work to prepare…and wound up spending all three in bed.

As a result, I jettisoned the instruction sheets, participant kits and other niceties I’d planned to include and got a lot more casual. (In fact, if my friend Becky hadn’t agreed to TA and squire me around town picking things up, I’d still be in bed)

Gratifyingly, one of the participants elected to take notes and distribute them, but if I’d been better prepared she wouldn’t have needed to. This post is sort of an attempt to make up for that.

ogg-claymakingBeing sick also made for a rather intriguing frit selection. Apparently out of time and my head, I grabbed a box and just started dumping frit jars in. When I stumbled into the workshop, still sick, I checked the box and found an extremely, er, eclectic palette:

  • Giant full jars of Olive Green and Blue-Grey Opal
  • That daggone jar of Cranberry Pink I’ve been trying to get rid of for 10 years (Cranberry will TAKE OVER a pate de verre composition)
  • Lots of greys and whites, including my last jar of Dense White (I was apparently thinking of angel wings)
  • Honeymix (well, at least I did SOMETHING right–this 50-50 mix of Light Peach Cream and Crystal Clear powder makes one of the most beautiful alabasters I’ve ever seen)
  • Plenty of sulfur and selenium glasses to react wonderfully with equally plentiful copper blues.

Now THERE’s a challenging palette, one that gave me a really nice segue lecture on “How to discover if you’ve got a reactive color in your frit mix.” ūüėČ

So…there was GONNA be a part II to this post, where I showed the results, explained what the students chose to make¬†(for example, how Jamie’s innocent robin redbreast somehow¬†morphed into an evil, scimitar-beaked raptor¬†ripping apart a screaming, skull-headed worm with blood dripping down into the fiery pits of Hell…until I reminded her that this¬†glass quilt was for a CHILDREN’S HOSPITAL…). Unfortunately, my pneumonia got worse, and the post became history. Sorry about that.