One Pass Only or the Break Wanders

Stained Glass 🎮 Play: Prism Path

6:47 PM — Studio on 7th Avenue, second floor above a thrift shop. Drop-in Thursday session, $25 materials included. Eight workbenches with overhead task lamps. Smell: flux fumes, kerosene from the cutter reservoirs, solder smoke. Three people already working when I walked in—someone grinding edges on what looked like a hummingbird pattern, someone else brushing copper patina onto finished seams.

Instructor showed me the glass rack first. Sheets organized by colour, maybe forty different types. Held up cobalt blue—stamped “96 COE” on the edge. Coefficient of expansion marking, same as the glass blowing session back in May. Different glasses shrink at different rates when cooling. Mix incompatible COE values in a panel and thermal stress will crack it apart within days. Here it’s a purchasing constraint, not a furnace disaster.

Picked a simple geometric pattern from the binder: three rectangles, blue and amber, framed in clear. Beginner project. Copper foil technique—Tiffany method, tape the edges and solder them together. Lead came is traditional for large church windows but requires different tools.

7:02 PM — Scoring glass doesn’t cut it, just weakens it along a line. Carbide wheel cutter, looks like a fountain pen with a tiny pizza wheel on the end. Reservoir in the handle filled with kerosene to lubricate the cut and dissipate heat—friction at the scoring point generates enough thermal energy to dull an unlubricated wheel in minutes.

Drew the wheel across blue glass in one continuous motion, firm downward pressure. Makes a crunching sound, quieter than I expected. Left a faint white line. The instructor leaned over and said: “Good. One pass only. Don’t go back over a score—you’ll chip the edges and the break will wander.”

Running pliers to snap it. Curved jaws, center line marked in red. Align that line over the score, squeeze gently. Glass fractured cleanly along the line with a sharp click. Pressure differential—the curved top jaw compresses while the bottom supports, forcing the fissure to propagate exactly where you weakened it. Physics constraint: if the score isn’t continuous, the break will veer off course toward the nearest stress point.

7:18 PM — Four pieces cut, edges rough. Grinder next: diamond-coated bit spinning in a water bath. Submerge the glass edge, press lightly against the bit, watch the corner radius smooth out. Water keeps the glass cool and washes away the slurry. Without water the dust goes airborne—silica particulates, bad for lungs. Also the friction heat can crack thin sections before you notice.

Grinding removes maybe 0.5 mm per pass. Enough to eliminate the sharp snapped edge, not enough to reshape the piece. If your cut was off by more than a millimeter, you recut. Grinding isn’t a correction tool for bad scoring—it’s finish work.

7:29 PM — Copper foil time. Adhesive-backed tape, 7/32“ wide, wound on a spool. Peel the backing, center the glass edge on the tape, press it down while rotating the piece. Foil wraps the perimeter with equal overhang on both sides. Then burnish it flat with a wooden tool—fid, same term bookbinders use for creasing.

Foil thickness matters for joint strength. This was 1.5 mil copper. Thicker foil (2 mil) gives more solder grab, thinner looks cleaner but can peel during soldering if you overheat it. The tape’s already oxidized to dull brown. Flux will clean that oxide layer off so solder wets properly.

7:41 PM — Laid the pieces on the pattern, edges touching. Brushed liquid flux over every seam—zinc chloride solution, acrid fume smell. Flux removes copper oxide and prevents new oxidation during heating. Without it the solder beads up and won’t bond.

60/40 tin-lead solder, the same stuff electronics people phased out twenty years ago for RoHS compliance. Stained glass work still uses leaded solder because it melts at 190°C and flows smoothly. Lead-free alternatives melt higher (220°C+) and don’t wet as cleanly. Temperature margin between “solder melts” and “copper foil delaminates” is only 30°C. That margin matters.

Iron set to 345°C. Drag the tinned tip along the foil seam with solder wire feeding continuously. The solder melts, wets the copper, and forms a raised bead about 3mm wide. Speed controls bead height—too slow and you get a lumpy buildup, too fast and you get a shallow film that doesn’t cover the foil completely.

8:04 PM — First side done, flipped the panel. Repeat on the back. Solder joints have to penetrate through the seam to bond both sides, otherwise you get a mechanical joint that can separate under stress. The front bead melts slightly when you solder the back, fusing them into a continuous joint.

Tinned the border edge last—pure cosmetic, no structural purpose. Just makes the copper foil transition into the solder smoothly instead of showing raw tape.

8:16 PM — Copper patina. Black liquid in a bottle, copper sulfate solution. Brush it over the silver-grey solder seams and they darken to charcoal within seconds. Chemical reaction, copper plating onto the tin-lead surface. The patina is fragile until it oxidizes fully—takes a day. You can also patina copper-coloured (no treatment) or leave it silver (just wax).

Held the finished panel up to the overhead light. Amber rectangle passed warm orange, blue passed cyan shifted slightly violet, clear passed white with minimal tint. Not the discrete emission lines I was measuring last week with the prism—this is absorption spectroscopy. The glass removes specific wavelengths from the transmitted white light, leaving what your eye perceives as colour.

Cobalt oxide in the blue glass absorbs yellow-orange-red (580–700 nm) and transmits the blue-violet end. Iron and sulfur compounds in the amber glass absorb blue-violet (400–500 nm) and transmit yellow-red. Both are subtractive filters. A sodium street lamp behind this panel would show two 589 nm yellow lines in the clear section, nothing in the blue section (absorbed), and the same two lines in the amber section (transmitted). Filtering is wavelength-selective. The glass doesn’t care whether the incoming light is continuous or discrete—it just removes photons below its absorption edge.

Red glass is allegedly much darker. Copper or gold dopants at high concentration absorb so much that 3 mm thickness looks nearly black. The old technique: flash a thin red layer onto clear substrate. Gives you red transmission without opacity, and you can engrave through the red layer to expose clear glass underneath. Pattern without additional lead lines. Didn’t see any flashed glass in the rack tonight—probably not stocked for beginners because it’s expensive and easy to ruin.

8:34 PM — Panel’s wrapped in newspaper. Flux residue hardens into white crust if you don’t clean it, but I’m walking home and it’s cold out. I’ll wash it tomorrow with dish soap and an old toothbrush.

Right hand smells like solder flux. Left thumb has a small burn from touching the iron barrel while adjusting the angle. Not the steam-barrier burns from handling molten glass, just careless contact with a hot metal surface. Different thermal regime entirely—370°C iron versus 900°C glass—but both disciplines require the same spatial awareness of where the heat is and how fast it moves through materials you’re holding.

Three rectangles, six solder joints, ninety minutes start to finish. No structural purpose, no functional use. Just coloured glass filtering afternoon light in a way I can now mount in a window frame and look through whenever the sun hits it at the right angle.