A glaze is a thin layer of melted glass, built from silica and fluxes, that bonds to a clay body in the kiln and freezes onto it on cooling — and it cracks or crackles when the glass and the clay beneath it shrink at different rates as the kiln falls in temperature. The phenomenon, called crazing when fine and shivering when the glaze pops off in flakes, is one of the most documented subjects in ceramics literature; the standard technical references, including the long-published accounts in Ceramics Monthly's glaze chemistry columns, trace nearly all of it to a mismatch in thermal expansion. Understanding the chemistry turns a frustrating fault into a controllable variable.
What is a glaze made of, chemically?
Every glaze balances three documented roles, per standard glaze-chemistry references such as those collected by ceramic materials databases including Glazy's published material pages: glass-formers, chiefly silica, which melt into the glass itself; fluxes, such as sodium, potassium, and calcium oxides, which lower the melting temperature so silica can fuse at achievable kiln temperatures; and stabilisers, chiefly alumina, which stiffen the melt and keep the glass from running off the pot. Colourants and opacifiers ride along in small quantities. The fired result is a glass whose composition the potter has, in effect, designed from minerals.
Why does the fit between glaze and clay matter?
Because both materials contract as they cool, and glass does not forgive a disagreement. If the glaze contracts more than the clay, it finishes cooling under tension and splits into the fine network of cracks called crazing; if the clay contracts more, the glaze finishes under compression and can shear off in flakes at the rim — shivering, the less common and more dangerous fault. The compression case is deliberately exploited in some traditions: saturated iron glazes and raku crackle both organise tension into pattern, and centuries of published ceramic history document craze lines filled with ink or stain as deliberate decoration. The fault and the feature are the same physics, differently welcomed.
How do potters actually adjust the fit?
Through documented recipe moves, per glaze-chemistry literature:
- Raise silica in the glaze slightly — more glass-former generally lowers the glaze's contraction.
- Substitute fluxes: sodium and potassium (from feldspars and frits) expand strongly; calcium and magnesium contract less, so recipes shift between them to tune fit.
- Change the clay body's silica content in the other direction, widening or narrowing the gap from beneath.
- Fire to the temperature the recipe expects — underfired fluxes leave unmelted material that behaves unpredictably on cooling.
What do cones and temperatures have to do with it?
Kiln temperature is measured by pyrometric cones — small slender pyramids of calibrated clay that bend when the kiln has delivered a specific combination of heat and time, a system documented across a century of ceramics practice and standardised by bodies such as Orton (Ohio, US), whose published cone charts potters use worldwide. Cones exist because heat work is not just temperature: an hour longer at a slightly lower heat can mature the same glaze. The glaze's glass forms properly only when the cone bends properly, which is why firing schedules, not just peak temperatures, fill the technical literature.
When is a crack not about fit at all?
When it is the clay's own problem — dunting, the cracking of the ceramic itself during cooling, happens when quartz inversions in the body pass too quickly, a documented physical transition at around 573°C where silica crystals rearrange and the piece momentarily changes size. Large pieces and thick bases crack at dunting for reasons no glaze adjustment can cure; slower cooling through the inversion range is the standard remedy, per kiln manufacturers' published guidance. The materials remember everything; the potter's job is to give them the time they ask for.