How to correct crazing
by Linda Bloomfield

Crazing or crackle is a problem in glazes where the glaze is too small for the clay body. On cooling in the kiln, the glaze shrinks more than the clay, and a network of fine cracks develops, accompanied by pinging sounds. Cracks can continue to appear over weeks and months, particularly if there are dramatic temperature changes such as those that occur when the pot is put in the oven or dishwasher. Crackle is often used as a decorative effect in Oriental ceramics, but the cracks can greatly weaken the pot, causing it to crack and break more readily than a pot with an uncrazed glaze. Crazing on earthenware pots can cause them to leak, as the fired clay body remains porous and water can seep through. The cracks can also harbour dirt and bacteria, so are not ideal on functional pots.

Linda Bloomfield, porcelain bowl with runny turquoise glaze, fired to cone 8. The crazing was caused by a change in the feldspar composition.

Adding silica and clay
There are several ways to correct crazing. However, changing only one material may change the appearance of the glaze, making it more glossy or matt. A reliable method is to increase both the silica (flint or quartz) and clay in the ratio 1.25:1 silica to clay. This ratio comes from a series of porcelain glaze tests made by RT Stull in the USA in 1912. He measured the effect on the glaze of changing the alumina and silica content.

He found that a molecular ratio of 1:5 alumina to silica gives a matt glaze, while 1:8 gives a shiny glaze. Alumina is found in clay, which is added to glaze to stiffen it in the melt and prevent it running off the pot. Clay contains silica as well as alumina, and the ratio 1:8 alumina to silica molecules converts to a ratio by weight of 1:1.25 clay to silica.

To change your glaze recipe, add increments of 1% clay and 1.25% silica until the crazing disappears. This has the effect of moving along the dashed line in the graph, from the crazed to the glossy, uncrazed area. You will need to multiply the percentage by the total weight. For example, in a 100g dry batch of glaze, add increments of 1g china clay and 1.25g quartz, then wet sieve and test on a pot to see if the crazing is reduced. If not, add increments of 1g up to 4g clay and 5g silica and test again.

Test tiles are often too small an area to see crazing, so you may need to test on a larger bowl or plate. Crazing is also affected by the glaze thickness. A thinner application of glaze or the addition of more water is sometimes enough to reduce the crazing.

Adding fluxes
Adding too much silica and clay may make the glaze appear underfired, dry matt. Another way to correct crazing is to add a low-expansion flux material such as talc, which is magnesium silicate. Both magnesium oxide and silica have low expansion. Both will decrease the expansion and contraction of the glaze during cooling, so will prevent crazing. Talc is a better choice than dolomite, which contains both magnesium and calcium oxide, the latter of which has a relatively high expansion.

If you look at the fluxes below, listed in order of decreasing expansion, with high expansion at the top and low expansion at the bottom, you can choose the most appropriate low-expansion flux to add to your glaze. It’s simpler to increase a material already in the glaze than to add a new one. Stoneware glazes can have the addition of 5% talc or zinc oxide. However, these fluxes may affect the colour of chrome green and chrome-tin pink glazes.

Earthenware glazes can have the addition of a flux containing boron, such as calcium borate frit. Opacifiers and colouring oxides such as zirconium silicate and tin oxide also help to reduce crazing.

The effect of changing the alumina and silica in porcelain glazes fired to cone 11 with constant flux 0.3 K2O and 0.7 CaO. The ratio of 1:5 alumina to silica gives a semi-matt glaze, while 1:8 gives a shiny glaze. The straight lines on the chart represent alumina:silica ratios of 1:4 (matt) 1:5 (semi-matt) and 1:12 (shiny, crazed glaze). The dashed line is 1:8 Al2O3:SiO2 (bright, shiny glaze). The hatched area shows crazed glazes. By increasing the alumina and silica in the molecular ratio 1:8, you can move a glaze from the crazed area to the glossy, uncrazed area. Data from RT Stull, 1912. Graphic by Henry Bloomfield. Thanks to Matt Katz of Ceramic Materials Workshop.

High expansion
 ∧        Sodium oxide (in soda feldspar and high-alkaline frit)
|         Potassium oxide (in potash feldspar)
|         Calcium oxide, strontium oxide
|         Barium oxide
|         Titanium oxide, lead oxide
|         Lithium oxide (in lithium carbonate and petalite)
|         Zinc oxide
|         Magnesium oxide (in talc)
|         Tin oxide, zirconium oxide
|         Alumina (in clay)
|         Silica (quartz, flint)
∨        Boron oxide (in calcium borate frit)

Low expansion


However, adding a flux will often make the glaze more runny. Alternatively, you can remake the glaze, reducing the high-expansion materials such as feldspar, or substituting low-expansion materials for the high-expansion ones, such as a lithium feldspar for soda feldspar, or calcium borate frit for high-alkaline frit. It’s a good idea to change one material at a time, to keep a clear idea of the effect of each individual material on the glaze.

A few years ago, I had a crazing problem in my transparent coloured glazes (see recipe below) when my supply of feldspar changed. I added different materials to correct the crazing, depending on each glaze colour. The opaque glazes containing tin oxide didn’t craze. The copper turquoise had the addition of 5% talc, which corrected its crazing. The chrome green had the addition of 4% silica, as talc would have altered its colour.

Changing the clay body or firing temperature
Others ways to correct crazing include changing to a different clay body that better fits the glaze, adding silica to the existing clay body, or increasing the firing temperature. In stoneware, the addition of silica sand to the clay body can help prevent crazing. In earthenware, bisque firing to a higher temperature can eliminate crazing.

Calculating the molecular formula
Glaze recipes can be written either as a recipe by weight percent, or as a formula. To calculate the formula, you need to convert each material in the glaze into a number of molecules, depending on their relative proportions in the glaze recipe. To do this, divide the weight of each material by its molecular weight. Dividing by the total number of alkaline fluxes (sodium, potassium, calcium) gives the unity molecular formula, where the sum of the alkaline fluxes is arbitrarily set to 1. This will give you the ratio of silica to alumina, and whether the glaze is likely to be matt or glossy. You can use to calculate the molecular formula for your glaze recipe.

Biaxial grid with high alkaline glaze from Mike Bailey’s cone 6 glazes book and copper oxide. Silica increases from 0-50% left to right. Clay increases from 0-40% bottom to top. The high-alkaline fluxes (Na2O+K2O+Li2O 0.7, CaO 0.3) remain constant throughout. The only uncrazed tile is in the centre row, fourth from left. Interestingly, the copper changes from turquoise to green as more clay is added, which is why transparent turquoise glazes are often crazed. Grid made by May Luk and the London Potters glaze group.

Glaze recipes

Runny turquoise glaze cone 6-9

This glaze is in the glossy area on the graph (Na2O 0.25, CaO 0.75, B2O3 0.38, Al2O3 0.38, SiO2 3.1)

Soda feldspar 47
Calcium borate frit 16
Whiting 14
China clay 5
Quartz 18
Copper oxide 2


High-alkaline turquoise crackle glaze cone 6

This glaze is in the crazed area on the graph (Na2O 0.7, CaO 0.3, Al2O3 0.23, SiO2 2.5).

Soda feldspar 15
High alkaline frit 47
Lithium carbonate 2
Whiting 6
Quartz 18
China clay 10
Copper oxide 2


Runny turquoise glaze with no crazing. This glaze contains calcium borate frit.


M Bailey, ‘Cone 6 Glazes, A&C Black, 2001.
I Currie, ‘Stoneware glazes, a systematic approach, Bootstrap Press 1985.
CW Parmelee, ‘Ceramic Glazes, Cahners publishing company, 1973.
RT Stull, ‘Influence of silica and alumina on porcelain glazes’, Transactions of the American Ceramic Society 14, p.62, 1912.
M Katz, T Gebhart and W Carty, ‘The re-evaluation of the unity molecular formula limits for glazes’, Ceramic Engineering and Science Proceedings 24, (2), p.13. 2003.


Linda Bloomfield is a scientist turned potter. Glaze recipes can be found in her books, Colour in Glazes (A&C Black, 2012) and The Handbook of Glaze Recipes (Bloomsbury, 2014). Her latest book, Science for Potters (The American Ceramic Society, 2017) is now available in the UK from


For more ‘How to’ guides, see here