•The secret to cool bodies and glazes is a lot of testing.
•The secret to know what to test is material and chemistry knowledge.
•The secret to learning from testing is documentation.
•The place to test, do the chemistry and document is an account at https://insight-live.com
•The place to get the knowledge is https://digitalfire.com

Sign-up at https://insight-live.com today.

Crazing in Stoneware Glazes: Treating the Causes, Not the Symptoms

Section: Glazes, Subsection: Thermal Expansion


Band-aid solutions to crazing are often recommended by authors, but these do not get at the root cause of the problem, a thermal expansion mismatch between glaze and body.

Article Text

While potters can be very creative and inventive, they can definitely learn something from industry when it comes to dealing with crazing. First, you can't ignore the hygienic and strength issues associated with the problem. The second is that although a piece may not exhibit crazing out of the kiln, glaze fit may still be very poor, it may only be a matter of time until it shows up.

Crazing is due to a thermal expansion mismatch between body and glaze. As a piece of ware is heated and cooled during normal use, it expands and contracts. An incompatible clay and glaze usually means the glaze either immediately or eventually fails by crazing or shivering (the former being more common).

Thus since glaze and clay are inseparably joined they must be expansion-compatible, no amount of 'covering over' this incompatibility is going to make it disappear. Band-aid approaches at best just lengthen the time before crazing starts. The only real solution is to adjust the expansion of body or glaze or both. If a suggested solution doesn't accommodate this, then you are treating the symptoms and not the cause.

In middle-fire and high-fire where glaze and interface development is good, the appearance of crazing suggests a huge incompatibility, not a small one. Strength tests have demonstrated up to a threefold difference in the strength of glaze test bars between high and low expansion versions of a glaze without visible crazing to suggest a problem. This means that by the time crazing shows up, there are already unreleased stresses that result in a big strength problem. To fix it, a significant change is needed, one that not only addresses fixing the crazing phenomenon but that restores the fired strength of the clay/glaze 'marriage'.

The bottom line is that crazing is best solved on the oxide level for the glaze and the material level for the body. That means body testing and glaze chemistry are involved in any real solution.

Often articles on how to deal with glaze crazing appear in ceramics journals. Unfortunately many continue to perpetuate a treat-the-symptoms approach rather than getting to the root of the problem. Consider some of the suggestions often given:

  1. Apply a thinner glaze coat

    Of course, it is not advisable to apply any functional glaze super thick, but if it does not fit with some variations of normal thicknesses, it does not fit. Every time it gets applied too thick it is going to craze. While a thinner application results in a greater percentage of the total glaze thickness being part of the clay-glaze interface and it might appear that crazing has stopped, either it fits or it does not, time will be the true indicator. And during the time while you wait for it to appear the differential stresses will be weakening your pot.

  2. Add increasing amounts of silica

    Yes, this may work in some cases of slight crazing where the glaze can tolerate more silica. However adding enough to simply hide out-of-the-kiln crazing may not deal with the deeper fit problems, the internal stresses between body and a glaze of higher expansion will weaken the ware and eventually craze it anyway. Also, it is often surprising how much must be added to reduce expansion to any degree. Why? Yes, silica is a low expansion oxide, but if the glaze is full of high-expansion fluxes like sodium and potassium, adding it is a little like adding white paint to dilute the color of black. The amount of silica present is usually what is required, so adding more can introduce unwanted gloss, higher melting temperatures, and change in surface character. Since mattes, which usually maintain a critical ratio of silica:alumina, are the ones that often craze, adding silica is not a solution. In addition, highly melted reactive glazes typically depend on low silica for their unique character. What about functional glossy glazes? Many high temperature glossy glazes will tolerate silica additions, but again, to put enough in to solve the problem will normally detrimentally affect melting. Adding silica to middle or low fire glazes will almost always mean less melting.

  3. Remove some feldspar and line blend additions of silica

    This is going in the right direction, sort of. The feldspar does contain the offending high expansion oxides of K2O and Na2O, but if you remove these the glaze is not going to melt enough (unless of course it was over melted before). Then when you add more silica it will melt even less. When you remove feldspar you are removing flux, silica and alumina. That means you have to add a different flux and silica and alumina to restore the balance. This approach cannot possibly work and still maintain the appearance of the glaze.

  4. Firing higher or over a longer time.

    One contention here is that the clay will tighten and have a better developed interface with the glaze, this better interface will produce a better the fit. But a better interface for a crazing glaze does not mean a better fit, it means a forced fit. If the expansion is wrong, then this is what should be changed. If the interface is conceptually the 'glue' that holds the glaze on, then gluing them together more strongly does not make them compatible. And firing the body higher could cause other issues like warping, color darkening not to mention inconvenience, impracticality and energy waste. Another contention is that more heat-work applied will take all particles into the melt and even out phase separations and thus produce the optimal glass that hopefully fits better. But the phase separations are probably a key to the appearance. And if melting it better fixes the problem then the problem was under firing, not crazing. And firing higher will certainly gloss up any matte effects you liked.

  5. Add increments of 5% silica to the clay body

    It is not usually practical to adjust the body because most people use prepared clays. But if you can do this it is true that adding coarser sizes of high-expansion silica have the end-effect of compressing the glaze (because quartz particles contract near 1% when going down through quartz inversion around 570C, thereby putting a squeeze on the glaze). Supposedly the added silica can impose its higher expansion on the surrounding matrix. However, vitrified ceramics are brittle, if individual silica particles are securely cemented into it and playing their own expansion game, this sounds like it could mean trouble if there are too many. Another issue is cristobalite. The term "cristobalite inversion" gets thrown around alot when crazing is discussed. It is a crystal phase of silica and quartz converts to it above 1100C when conditions are right (feldspar is lacking and there is plenty of fine quartz available). Like quartz, cristobalite suddenly expands/contracts on heat/cooling, but does so at a much lower temperature (around 225°C), over a narrower range and to a greater extent. This puts a late squeeze on glazes, preventing crazing. But is this desirable? Google the thermal expansion curve of cristobalite and you will see it looks alot like the final section of the Jurasic park water ride at Universal studios. Do you want that kind of low temperature thermal expansion behavior in vessels made from your porcelain or stoneware clay body?

  6. Slow cool the glaze kiln, don't open it until it is below 200°C (390°F)

    If the glaze can barely withstand a normal cool cycle after firing, then how will it take the thermal punishment of the dish washer or the normal hot-cold shocks of everyday use? It is possible that a glaze under tension can be eased down and may well appear to be OK when the kiln is opened. But the first thermal stresses it undergoes in use will reveal what the slow cool temporarily covered up.

  7. Bisque higher if low fire glaze is not fitting

    Making a glaze fit on a low temperature body where there is a lack of clay-glaze interface is much more difficult. Band-aid approaches just won't work, the thermal expansion of the clay and glaze must match closely. It is true that crazing will be evident at low-fire with a much smaller difference between clay and glaze expansions than with stoneware. Thus smaller changes have a bigger effect. It is true that firing higher can change a clay body's expansion. However, the above approach will only work if the increased temperature lowers the expansion and if the glaze firing is lower than the bisque. However, adjusting low fire glazes with boron sourcing materials (i.e. frits) is much more practical. Low expansion frits, specifically targeted as additions to reduce crazing, are available.

  8. Change to a frit of lower expansion

    This suggestion goes in the right direction as it proposes to change the expansion. However, few frits are used in stoneware glazes (the few that are are normally there for a very specific purpose). Thus simply changing to a frit of different chemistry and lower expansion is likely to mean altered glaze properties unless you know the implications. In lower fire glazes, this approach is much more common since many boron frits are quite balanced, almost stand-alone glazes by themselves.

Rather than thinking of crazing as something that is "caused" by something else, it is better to think of crazing as evidence that a clay and glaze are not expansion-compatible. Viewing crazing as a material level problem might be OK for certain highly fritted low temperature glazes, but it does not work for stoneware glazes. It is oxides of high expansion that cause crazing, not materials. That is why crazing is a problem that is much more effectively solved on the oxide or formula level using simple calculations (e.g. substituting fluxing oxides of lower expansion for those with higher expansion).

In the end, while it may be feasible to increase the body expansion, most often it is more practical to control crazing by adjusting the chemistry of the fired glaze in such a way that fired properties are maintained while expansion is decreased. Just remember, if a glaze crazes, it is a sign of big trouble; not the type of thing that small adjustments, material additions, or firing changes are going to easily solve.

Adding silica will fix crazing, right? Not here.

Adding silica will fix crazing, right? Not here.

G2926B (center and right) is a clear cone 6 glaze created by simply adding 10% silica to Perkins Studio clear, a glaze that had a slight tendency delay-craze on common porcelains we use. Amazingly it tolerated that silica addition very well and continued to fire to an ultra gloss crystal clear. That change eliminated the crazing issues. The cup on the right is a typical porcelain that fits most glazes (because it has 24% silica and near-zero porosity). The center one only has 17% silica and zero porosity (that is why it is crazing this glaze). I added 5% more silica to the glaze, it took that in stride, continuing to produce an ultra smooth glossy. It is on cup on the left. But it is still crazing just as much! That silica addition only reduces the calculated expansion from 6.0 to 5.9, clearly not enough for this more severe thermal expansion mismatch. Substituting low expansion MgO for other fluxes will compromise the gloss, so clearly the solution is to use the porcelain on the right.

Out Bound Links

In Bound Links

By Tony Hansen

Feedback, Suggestions

Your email address


Your Name


Copyright 2003, 2008, 2015 https://digitalfire.com, All Rights Reserved