cabsFun little project, which started with my urgent need for about 600 glass cabochons for a craft fair.

I began cutting up dozens of failed projects, roughly shaping and firing them into cabs ranging from about a quarter-inch to 3×3 inches. Sometimes the results were spectacular, sometimes not…but my absolute favorites came from failed boxcasting experiments. I decided to make them on purpose.

I’d been experimenting with box casting and pate de verre, trying to find a way to get the beautiful, translucent color blends of PdV without all the just-under-the-surface bubbles. Now, it’s the bubbles that make pate de verre sing–they bounce the light in all directions and literally make the piece glow–but bubbles also make it extremely difficult to grind and carve into the glass without creating a lot of light-disrupting pinholes.

So, reasoning that sheet glass would give me the smooth surface I needed, I’d added inner and outer walls of clear sheet glass to the castings. As an added bonus, the clear glass would provide a lens, magnifying the patterns. I’d done this before for wall panels, where I “painted” frit onto clear glass sheets.

This time I’d put clear glass on both sides of the frit pack, so that I could stand the panels up vertically and pour frit into the opening. I got the idea from those sand/sugar layering crafts that hit big in the 80s, the ones where you carefully drip colored sand/sugar into a glass container, then manipulate the layers with a stick to create all kinds of patterns.

I assembled my hollow-wall glass boxes with some scotch tape, dammed them up, then used that 80s technique to layer in colored frit. I filled the center of the boxes with refractory plaster, dammed the outside with kiln furniture, and fired them like castings.

They failed for about seven reasons, but the theory was solid–someday I’ll try it again–and the walls of those boxes were gorgeous. After cutting them up to make beautiful cabochons, I decided to simply try making “sand painting” panels, without the box.

What I got were 4-5mm thick panels resembling a potmelt or hot-raking project, or perhaps a pattern bar gone nuts. These panels cut easily with a standard glasscutter and breaking pliers, and allow both surface carving and incorporation into other projects without much fuss, just like potmelts.

But where potmelts and rakes show solid lines of smooth, fully-melted glass blending into each another, this method produces distinct, textured bands like colored sand paintings. Better still, you can precisely place your color, giving you a lot more control over design than you have with hot-rake or potmelt. This method also helps a little frit to go a very long way and, with good heat control, allows you to dial in smooth, granular, or downright pebbly appearance.


A piece cut from one of my “sand-painting” panels, showing the degree of detail possible with various forms/colors/transparencies of frit.


Coarse transparent frit packed between two thin clear sheets of glass, fired at a tack-fuse, creates this mosaic-like texture with a totally smooth surface.

I’d done something a bit like this in 2003 or so: I’d piled coarse frit onto a sheet of thin clear glass in patterns, smoothed it out and topped it with another thin clear sheet, then fired to a melty tackfuse. I coldworked the edges of the panel and slumped the result into a set of dishes. The dishes were extremely crystalline, graphic and shiny, rather like a complex mosaic, but the coarse frits didn’t give me much control over detail.

Smaller frits solved that problem. If you’re patient (or masochistic) enough, you could probably use this method to create detailed portraits and landscapes, the way they do in the Middle East.

DISCLAIMER: I am NOT claiming to be the inventor of this method. Like most glassmaking techniques, the first inventor probably died 2000 years ago, so I’m just a “rediscoverer.”

Experienced glass fusers probably already have everything needed for this project already, but here’s a list just in case:


  • Long bamboo skewer, stick, pick, just so it’s longer than the glass you’ve cut
  • Superglue (or a hot melt glue, but the superglue holds better (I think))
  • Invisible (scotch) tape, the cloudy kind; other types may not burn out completely
  • Kilnshelf about two inches bigger on all sides than the panel you’re making
  • bargellic10-4Kiln weights (mullite bricks, stainless steel bowl full of sand, whatever)
  • A jeweler’s scoop if you can find one, or a thin plastic spoon


  • Two sheets of clear or crystal clear thin (2mm) glass (mine are usually around 8×10 inches). Cut one sheet a quarter-inch shorter in the long direction–this will be the top glass.
  • Enough 1/4-inch strips of 3mm glass to fit along the bottom and two long sides of your base (back) glass sheet. These will be your bottom and side spacers.
  • A half-inch glass strip that fits between the two side spacers, across the top of the envelope, to seal the frit into the envelope and prevent it from leaking during firing.
  • Frit–powder, fine and medium frit

Let’s get started.

Step 1: Make the “sandbox” 


The completed glass “envelope,” with tape across the bottom and up the sides about 2 inches, plus a reinforcing band around the top, to keep the filled box from coming apart. I’ve used clear spaces for the 3 edges of my envelope, but you can use any color of scrap you like, since these edges will be cut off.

Cut your glass pieces to any size–use thin clear glass for the sides of the envelope and 3mm for the spacers. BE Crystal Clear shows off colors best, but you can get neat special effects using thin tinted glass for top and back.

Clean your glass thoroughly and let it dry. Set the longest sheet on your worktable, with whichever will be the inside of the “envelope” facing up. Cut a 1/4 inch 3mm strip to fit across one of the shorter sides–this will be the bottom of the envelope. Place it exactly across the bottom, matching edges, and superglue it in place.

Cut two 3mm side spacers to fit from the bottom of your end piece to the top of the back glass and superglue them to the edges of the back glass. Finally, cut a half-inch wide 3mm strip to fit between the side spacers, and set it aside. (this is the top/sealing spacer)

Lay the back glass assembly flat on your worktable, and apply superglue to the three glued-on spacers. Lay the top glass sheet on the spacers, matching the bottom and side edges. The top piece should be a quarter-inch or so shorter than the back glass. Press the pieces together firmly until the superglue sets.

You should now have a glass box, the “envelope,” with the long sides 3mm apart, and the front a bit shorter than the other.

Problem: Superglue isn’t all that strong. While it will probably hold your glass together throughout the filling process, I’m a suspenders-and-belt kinda gal, so I reinforce the bottom and top. If one of those glue joints gives way during filling, you’ll waste all your work, messily.

It’s easiest to run a strip of tape across the bottom of the envelope and about two inches up the sides, pressing firmly to seal. Add another band of tape around the top, about an inch from the edge, to make sure the glass stays together during filling.

If you’re worried about incomplete burnout, remove the tape when the envelope is filled, sealed and ready for the kiln. Most of the time, though, you can leave it on; the tape will burn off cleanly. (Of course, the one time you really NEED it to burn off…)

If I know I’m going to be making a few of these, I’ll make all the envelopes at once.

Filling the envelope

Start by picking your colors–frit is notorious for looking pastel in the jar but turning out dark, dark, dark after firing, so I usually just grab up the catalog and choose my colors that way.

I prefer mostly monochromatic or analogous palettes with maybe one or two surprises, but really, anything goes. I’d suggest restricting yourself to no more than 10 or 12 colors/textures for an 8×10 inch envelope, about 60-70 percent powder and the rest fine and medium frit. Coarse frit will probably be too big, and even medium frit may sometimes need one or two pieces picked out to ensure it fits between the glass.

Depending on how exacting you are about logging your colors, you can write down all the colors you’re using and if there’s a specific pattern to your fill. I prefer, though, to simply keep the stack of frit jars on my worktable, then photograph them next to the filled envelope when I’m done.

Filling is the fun part. Let yourself go wild, but be careful with scooping and transporting frit over open jars. I haven’t (yet) accidentally spilled the wrong frit into a jar, contaminating an expensive jar of glass, but…

You can scoop out some frit into a separate little cup to prevent jar contamination, but if you pour it back, you’ll risk contamination again, and so you wind up wasting glass. I lessen the chances of contamination by arranging my frit jars around the work, only one layer deep. That way I can reach straight into the jar without hovering over any other container.

bargellic-driftI fill the envelope using a jeweler’s scoop, a small, flat shovel on tweezers, which scoops up tiny gemstones. It’s perfect for sliding frit into a 3mm opening. If you can’t find one, you can also use a thin plastic spoon, or make a scoop by cutting out roofing tin into a parallelogram, and folding the sides into an open box.

(Just FYI, the jeweler’s scoop is also one of the best tools I’ve seen for applying fine lines of frit or powder to just about anything–you can buy them at any jeweler’s or lapidary supply store and they’re lifesavers.)

Hold the glass envelope upright, with the short side facing you, and get a scoopful of your first frit color. Tilt the envelope about 30 degrees backward, and rest the scoop on the short edge of the envelope, almost touching the back glass.

(This is why one edge of the envelope is shorter than the other–it forms a natural ledge for running your scoop back and forth to deliver controlled amounts of frit. It also makes it very easy to redirect any stray frit that falls down the front back up and into the envelope.)

Tap your filled scoop lightly on the glass to start the flow, sliding it along the short edge to distribute frit down into the envelope. Sliding produces enough vibration in the scoop to dump the frit.

bargellic-3Once you’ve dumped in as much frit as you need, stand the envelope up on its base and tap it gently against the work surface.

The frit should slide easily to the bottom and settle into place. Now scoop up your second frit color, and repeat the process.

You can fill up the layers in straight, even lines all the way across or lay them down with hills and valleys. Each new layer settles into the crevices left by previous layers.

You can also disrupt the layers and make new patterns using your skewer. Insert it into the envelope and slide it down into the frit layers. Push straight down and you’ll move the top layers of glass into the bottom, creating fine outlines.

bargellic-4Move the skewer from side to side, and you’ll open “canyons” in your layers that will be filled in as you add more frit (above, right). Rock the skewer around a center point, and the powder will move into swirls (left).

Just remember than anytime you push into the powder, or pull the skewer up out of the powder, it will trail color behind. If you are working with reactive colors (i.e., a copper glass with a sulfur glass), your powder trails can start to look like handwriting.

Insert a long piece of 3mm glass into the envelope to tamp down the frit, but don’t tamp so hard that you break open the glue seals on the envelope. Tamping helps (slightly) reduce bubbles, and you can also use it to make shapes. I’ve cut shapes into the ends of 3mm glass scraps–points, half-circles, zigzags–so that they’re essentially layer shapers.

Occasionally tap the assembly against the worktable, which also helps settle the frit and minimize bubbles.

Achieving different color effects

The type and transparency of the frit strongly influences your result, as does your choice of adjacent glasses:

  1. Strong, sharp graphic lines: Layer high-contrast dark powders (opaque or transparent) with pale opal powders.
  2. Soft, glowing pastel bands: Use mediumbargellicstrip-4– to light-colored transparent powders.
  3. Sparkling, effervescent areas: Build up larger layers of fine transparent frit in pale/clear shades. They tend to trap lots of small bubbles, which reflect light and look glittery .
  4. Blended, very organic and watercolory lines: Use darker, fine transparent frit against a paler color, such as the aventurine green against the spring green. The outlying chunks tend to blend in with the surrounding frit, giving a feathery effect. It’s enhanced if you put this layer between medium-to-pale opaque powders.
  5. Scattered small, glowing windows. Sift in a layer of opaque powder, then a layer of transparent frit in a medium size, followed by another layer of the same opaque powder. Now tap the mix sharply on your worktable. The powder will sift into the medium frit, outlining its grains. It works best if you use a darker opaque powder with a light transparent frit.
  6. “Pebbles.” Medium frit in an opaque color gives a pebbly, irregular edge to the color band. If you add a layer of powder in a strongly contrasting color on other side, the powder will sift in to outline the “pebbles.”
  7. Dark outlines. Place any size frit next to its reactive color (sulfur glasses such as yellows against a copper glass like turquoise, for example) and you’ll get a fine dark line between them, almost like a cartoon.

Prepping for the kiln

Keep filling the envelope until the frit is very slightly below the shortest side. Tap it a few times against the table to settle the frit, then slip the cap (that 3mm piece you cut to fit inside the top of the sheets) into the envelope. Push it down a bit (not too hard, or you might break through).

bargellic10When the cap is seated, drip some superglue into either side and let it harden. Remove the tape, and clean the outside of the panel (both sides) thoroughly with denatured alcohol and a green scrubby or similar. (There’s frequently a bit of powder on the outside; you can’t see it, but it WILL show up in the firing.)

bargellicSet the panel in the kiln, on a clean, smooth kilnshelf that’s either freshly kilnwashed or has a piece of smooth fiber paper on it. There should be at least a couple of inches between the envelope and the edge of the shelf.

Now center your piece of well-kilnwashed kilnshelf over the panel (it also should overhang the piece by a couple inches on both sides). Set it gently down on top of the assembly.

Now add 6-10 pounds of weight to that kilnshelf, to press down your frit assembly, primarily in the center. The weight helps squeeze trapped air out to the edges of the assembly, and it will also compress the glass to about 4-5mm, making the panel easier to cut and fit onto a piece of thin base glass.

Firing schedule
I modify a standard fusing schedule to include a big bubble squeeze. By holding at 1200-1300F for extra time, I’m giving all that frit-trapped air time to move to panel edges.

  • 300dph to 1150F/593C, no hold (heat to the start of the bubble squeeze)
  • 50dph to 1240/671C, hold 60 minutes (bubble squeeze and hold)
  • 200dph to 1485F/807C, hold 45 minutes (process time)
  • AFAP to 900F/482C, hold 2 hours (the extra time is mostly because there’s a LOT of extra insulation on this glass, and I want to give it time to relax evenly)
  • 40dph to 700F/371C, turn off the kiln (let the kiln freefall to room temperature)

It’ll take awhile for the panel to completely cool (those kilnbricks are holding in heat), so I generally wait an hour or two after the controller tells me the inside of the kiln is at room temperature. Then I remove the panel from the kiln–the reveal is kinda like Christmas morning.

bargellic-7How to cut these panels
Clean the panel well, then cut it up as needed. You need a bit more pressure on the score line than with sheet glass–I keep an older glass cutter for this kind of stuff–and some patience when you’re applying the breaking pliers.

Center the pliers over the score exactly and squeeze very firmly, waiting a few secs (sometimes) for the score to start running. If it doesn’t break right away, switch to the opposite end and start squeezing there. The glass is flat and fairly solid, so I can make clean, square cuts down to about 3/8 inch maybe 95 percent of the time.


“Sand panels” are fairly easy to cut with standard glasscutters. Here you can see how little actual frit is trapped between the layers of clear glass.

The cut is clean, and you can see how little actual frit you use in this method–it’s thinner than the thin clear sheet glass that forms the envelope.

The surface of the panel retains the texture of the kilnshelf on both sides, so you’ll want to fire this a second time to smooth out the texture, gloss it up, and brighten the colors. If you’re planning to incorporate the panel (or pieces of it) into other work using a second firing, that’ll do it. If you want to make cabochons or slump the panel as-is, you’ll need a second full fuse firing, or a lot of coldworking.

bargelliccabs-2The panel surface can scum up slightly, so while I’m not terribly fond of the “sprinkle with clear powder” method suggested by Bullseye and others for shining up a piece, it’s probably the fastest way to get a good gloss on the second firing.

(I don’t like covering scum/devit with clear powder because using too little powder will leave dots of scum behind that look worse than the original scum. Using too much introduces bubbles into the surface that cause their own haze and can also leave pinholes if you coldwork the surface. If you’ve got time, I think you’re much better off sandblasting and/or grinding the scummed surfaces to about 400-grit, then carefully firepolishing.)

If I’m making cabochons, I’ll do a final shaping with the coldworking equipment, up to 220 or 400-grit, and firepolish. That leaves me with a flawless surface and a nice, finished shape.

Postscript: Remember above where I said I wasn’t claiming to have invented anything? Good thing. My friend Bob Heath did some checking, and there is a guy named David Alcala who does indeed make full-blown pictures using this method. In fact, he appears to have started doing this with real sand, sandwiched between two pieces of window glass, then graduated to doing it with frit and tossing the result in the kiln.

So nope, I didn’t invent the idea of sandwiching frit between two sheets of clear glass, although there are significant differences between my method and Mr. Alcala’s. Now that I think of it, it’s so obvious I’m surprised more people haven’t “invented” this. 😉 Thanks for checking, Bob.