A furnace looks like a box that gets hot. In practice it is a system with three parts that fail in a predictable order: the crucible, the thermocouples and the elements. Understanding which one is about to fail saves both money and a ruined week.
What the furnace is actually doing
A studio melting furnace holds glass between 1100 and 1200 °C so that it stays fluid enough to gather but viscous enough to stay on the pipe. It does this with a controller reading a thermocouple and switching elements, usually split across three zones: crown, sidewall and base.
Three zones matter because heat rises. Without independent control of the base, the bottom of the crucible runs cold and glass freezes there, which is how a furnace ends up with a permanent heel of unmeltable glass.
The crucible is the consumable
Everything else in the furnace can last years. The crucible will not. It lives inside a bath of molten glass and metal oxides, and it is exposed to room-temperature air every time the gathering port opens. Thermal shock is what kills it, and daily thermal cycling is what causes thermal shock.
- A furnace run continuously at temperature will get more life out of a crucible than one cycled up and down every day.
- Never add cold cullet to a hot crucible. Pre-warm it on the crown.
- Keep the glass level above roughly one third of capacity. A nearly empty crucible overheats at the walls.
- Skim the surface before gathering. Oxide crust eats refractory when it is dragged down the wall.
A hairline crack in a crucible weeps glass slowly before it fails. The drip tray is where you will see it first, and it is the only warning you get.
Temperature, and why the readout lies
The controller shows the temperature at the tip of the thermocouple, not the temperature of your glass. Those are different numbers. A thermocouple in the crown reads the air above the melt; a thermocouple in the wall reads refractory. Only a probe immersed in the glass reads the glass.
This is why a furnace can read a stable 1150 °C while the glass is too stiff to gather. Before you chase the controller, check the thermocouple: a drifting reading is far more common than a failing element.
| Symptom | Likely cause | First check |
|---|---|---|
| Glass stiff at the usual setpoint | Thermocouple reading high | Compare against a handheld probe |
| Temperature drifts slowly upward | Thermocouple junction degrading | Replace the thermocouple |
| Cannot reach setpoint at all | Element or relay failure | Measure element resistance |
| Uneven melt, cold base | Base zone controller or element | Check zone output independently |
A daily routine that pays for itself
Morning: check the standing temperature
Compare it with yesterday. A one or two degree drift is normal. Ten degrees is a thermocouple on its way out.
Before the first gather: skim the surface
Remove unmelted batch and oxide crust. Thirty seconds here prevents a contaminated gather all day.
Between sessions: look at the drip tray
Any glass where glass should not be is a crack. Note where it came from.
Weekly: test the extraction
Smoke pencil at the port, with the furnace open. Capture should still pull the smoke into the hood.
Monthly: log the ramp performance
Time how long the furnace takes to climb 100 °C. A slowing ramp is the first sign of a weakening element.
A welded relay contact will let a furnace run away with no controller input at all. An independent mechanical limit is the only thing that stops it.
A furnace is a long-term relationship. The studios that get ten years out of one are not the ones with the most expensive equipment, they are the ones that kept records.
This article is educational reference material. Hot glass work involves severe burn, fume and fire hazards. Wear eye protection, heat-resistant gloves and natural-fibre clothing, ventilate every heat source to open air, fit a carbon monoxide alarm, and take hands-on instruction before operating any hot glass equipment.
No guarantee of results. Techniques, temperatures and schedules described here are reference guidance based on common studio practice. Glass behaviour depends on your materials, equipment and technique. Test everything on your own setup and follow the manufacturer's instructions.