Crawling

Ask yourself the right questions to figure out the real cause of a glaze crawling issue. Deal with the problem, not the symptoms.

Details

Crawling is where the molten glaze withdraws into 'islands' leaving bare clay patches. The edges of the islands are thickened and smoothly rounded. This is among the most costly problems in ceramic production, ware requires refiring and is often lost anyway. In moderate cases there are only a few bare patches, in severe cases the glaze forms beads and even drips off onto the shelf. The problem is prevalent where dipping glazes are applied to bisque, shrink on drying and form cracks. This is usually because the glaze has excessive plastic clay content or is applied too thickly.

Is the glaze shrinking too much during drying?

If the dried glaze forms a network of cracks it is a sign that the glaze is shrinking too much or its bond with the body is poor. The fault lines provide places for the crawling to start (especially where the island cracks delineate raised edges that are no longer in contact with the body). There are a number of possible contributors:

It is normal to see 20% clays (ball clay, kaolin). If significantly more is present try using a less plastic clay (i.e. kaolin instead of ball clay, low plasticity kaolin instead of high plasticity kaolin, or a mix of calcined and raw kaolin). Ultimately you must tune the glaze recipe's clay content to achieve a compromise of good hardness and minimal shrinkage while maintaining the chemistry (Want to learn how? Please email).

If a slurry has flocculated (due to changes in water, dry material additions like iron oxide, or addition of an acid, epsom salts, calcium chloride, etc) it will require more water to achieve the same flow and will therefore shrink more during drying and require a longer period to dry. Try using distilled water. Always measure the specific gravity to maintain solids content and use deflocculants/flocculants if necessary to thin/thicken the slurry (remove water from an existing glaze slurry by pouring some on a plaster batt, then mixing the water-reduced mass back in).

Gerstley Borate is plastic and therefore contributes to glaze shrinkage, especially if the recipe also contains kaolin or ball clay. It also tends to gel glazes so they need excessive water. Use boron frits instead. Try a boron frit (do calculations to make these adjustments, there is no frit that has the same chemistry).

Many potters add bentonite to every glaze they make, thinking that it will help suspend. But when glazes already have sufficient, or even excessive clay content, the extra bentonite may increase the shrinkage enough to cause drying cracks.

If very fine-particled materials are present (i.e. zinc, bone ash, light magnesium carbonate) these will contribute to higher shrinkage during drying. Try using calcined zinc, synthetic bone ash or another source of calcia, talc or dolomite to source magnesia instead of magnesium carbonate. If a glaze has been ball milled for too long it may shrink excessively (for example, zircon opacified glazes can be ground more finely than tin ones). Highly ground glazes may produce a fluffy lay down.

Again, be "recipe flexible", tune the raw clay content. Use enough clay in the glaze mix to both suspend the slurry and toughen the dried layer, more than that risks excessive shrinkage. Less and the glaze does not harden and forms a powdery surface. A fool-proof way to reduce shrinkage is to calcine (see the link below) part of the clay. If there is 35% kaolin in the mix, then try using 15% calcined kaolin and 20% raw kaolin (calcine the clay yourself if needed, it is easy). If there is 70% Alberta Slip, try using 35% raw and 35% calcined. Adjust the proportion as you get experience in working with the glaze. One detail: If there is significant clay content, adjust for the LOI (weight lost during firing) when calcining. Kaolin, for example, loses 12% weight on firing, so use 12% less calcined). Alberta Slip and Ravenscrag Slip have 9% LOI.

It is possible to create glaze slurries that gel and flow extremely well, dry hard and do not crack by using the right kaolin (i.e. EPK) or ball clay (i.e. Old Hickory #5, No. 5 Glaze) in adequate amounts. It is very important to realize that many ball clays and kaolins to not produce glazes of good slurry properties (without additives), a simple substitution can make a remarkable difference. It may not seem that glaze chemistry is related to this subject of getting the right type and right amount of clay in a recipe. But it very much is. Why? Because the ability to juggle a recipe to source Al₂O₃ from a clay or a feldspar/frit, as needed, enables controlling the amount of clay in the recipe while maintaining its fired character. Chemistry also gives you the ability to switch between different types of clay (that have different oxide makeups).

Is the glaze's dry-bond with the ware surface inadequate?

The mechanism of the bond between dry glaze and body is simply one of physical contact, the roughness of the ware surface copmbined with the ability of the liquid glaze to flow into all the tiny surface pores and irregularities and the degree to which it is able to dry hard without shrinking too much, these determine its ability to 'hang on'.

Some surfaces can seem very smooth (e.g. slip cast surfaces), but microscopically they are not. Still, the glaze needs enough fluidity to flow onto and wet the surface well.
If the dried glaze surface is excessively powdery incorporate more plastic clay (chemistry will be needed to juggle the recipe). Or add a little bentonite. If there is 10% kaolin, for example, consider changing that to 7% kaolin and 3% bentonite (there is a chemistry impact here, but not significant).
Adding gum to a glaze will harden it and bond it better to bisque, this often solves crawling problems. Beware that excessive gum can increase drying time and dripping considerably, so do test to achieve a compromise.
If a glaze is flocculated it may lack the necessary fluidity to run into tiny surface irregularities in the bisque and establish a firm foothold. However, some degree of flocculation is valuable, it enables the glaze to set and gel quickly after dipping, this can really accelerate production.
Wetting agents are available and can be added to the slurry to improve bond. But again, do not substitute these for the basics, the right amount of clay and degree of fluidity.

Does application technique or handling compromise the fragile glaze-body bond?

If ware is heavily dusted consider blowing it off. But a little dust is not an issue (remember glaze itself is 100% dust).
If glaze is applied too thickly the forces imposed by its shrinkage will overcome its ability to maintain a bond with the ware surface (especially inside corners or at sudden discontinuities). If a glaze can be applied more thinly, do so.
Use a fountain glazing machine to do the insides of bowls and containers to achieve a thinner layer.
If glaze needs to be applied in a thick layer, you can achieve a lower water content by deflocculating the glaze (i.e. with some sodium silicate or Darvan). Use caution with this, it may then tend to dry very slowly or form drips that crack and peel and instigate crawling. It is better to just have the right clay content.
When applying the glaze in the normal layer thickness be careful to prevent drips that form thicker sections that can crack away during drying. It is practical to 'gel' the glaze slightly (i.e. with vinegar, Epsom salts) so that it 'stays put' after dipping or pouring.
If a double-layer of glaze needs to be applied be careful that the second does not shrink excessively and pull at the first, compromising its bond with the body. If possible, the upper layer should have less clay, lower shrinkage, should be applied thinner and dry quickly. It may be necessary to bisque each layer on before applying the next. Double-layering typical raw art and pottery glazes is difficult, special consideration must be given. Have you successfully done this in the past without any special attention? You may have simply been very lucky.
When doing double-layer glazing be careful that the second layer is not flocculated (with an associated high water content). This will rewet the first layer and loosen it from the body. Adding iron oxide, for example, to a glaze will often flocculate it and require the addition of much more water to restore the same fluidity.
Spraying glaze on in such a way that the glaze-body bond is repeatedly dried and rewetted could a produce shrinkage-expansion cycle that compromises a glaze-bisque bond that could otherwise withstand one drying-shrink cycle.
Force-drying of the ware can make the glaze visibly crack when it otherwise would not (slower shrinkage associated with slower drying gives it the glaze time to ease body interface tension by micro cracking). Preheating the bisque too much may cause escaping steam to rupture the bond with the ware.
Rough handling of ware can compromise sections of the glaze body bond.
For the inside of slip cast ware, consider pouring a thin glaze slurry into the mold of a just-drained piece (perhaps a minute or two after the mold has been drained) and immediately pouring it out again. This base layer can be fired on in the bisque.

Is the glaze drying too slow?

This often occurs where ware is very thin (e.g. slipcasting or fine porcelain), the bisque is fired to higher density, glaze has a high water content or ware is already wet from a previous glazing (or a combination of these factors). If the glaze dries too slowly the most fragile stages of the adhesion mechanism are extended and cracks or bubbles develop. These low-bond areas instigate crawling during melting. To fix this problem speed up drying. You can do this by preheating the bisque (in a kiln to 150C or more if necessary) before dipping, doing separate interior and exterior glazing (with a drying period in between), applying glaze in a thinner layer, reducing glaze water content, bisquing lower to increase porosity or increasing wall thickness in the ware.

Is the ware once-fire?

Once-fired ware is prone to crawling because the mechanical glaze-body bond is more difficult to achieve and maintain. If glaze is applied to leather hard ware it must shrink with the body. During the early stages of firing the ware also goes through volume changes and chemical changes that generate gases, these affect the ability of the glaze to hang on.
When glaze is applied to leather hard ware you must be able to tune its shrinkage by adjusting the amounts and nature of the clays in the recipe while maintaining the overall chemistry (calculations may be needed).
In damp conditions the powdery layer may reabsorb water from the air causing slight expansion of the glaze layer, thereby affecting the adhesion.

Is the problem happening during firing?

If glaze is applied over stains or oxides that lack flux (e.g. chrome pinks, manganese types, greens, cobalt aluminate) it can be prevented from bonding well with the underlying body. Mix under-glaze stains with a flux medium so that over lying glazes can 'wet' them and form a glassy bond.
If the glazed ware is put into the kiln wet and therefore dried quickly during the early stages of firing, the glaze layer will tend to crack and curl and crawling can occur.
If glazed ware is put into a kiln containing heavy damp ware such that early stages of firing occur in very high humidity conditions the glaze could be rewetted and forced through an expansion-shrinkage cycle (that could affect its bond with the body).
If a glaze contains significant organic materials (i.e. gums, binders) that gas off excessively during firing the bond may be affected. Decomposition of materials like whiting can also generate significant amounts of gas within the glaze layer (try switching to wollastonite, it has no LOI and supplies SiO₂ and will permit reducing the silica content accordingly).
Raw zinc oxide is very fine and tends to pull a glaze together during firing, use calcined zinc instead (most zincs sold to potters are calcined).
If the glaze contains significant zircon opacifier, alumina, some stains, magnesium carbonate, the melt may be much 'stiffer' and flow less. This can affect its ability to resist crawling.
Watch out for glazes with slightly soluble materials like Gerstley Borate or wood ash. With the former the soluble portion tends to be the borate, which will be absorbed into the bisque during application. Then, during firing, it creates a highly fluid layer between the body and the less developed over glaze, it thereby prevents adhesion of the glaze to the body. Use frit to source boron instead. In addition soluble materials tend to flocculate (thicken) the slurry and attempts to thin them result in higher water content and therefore increased shrinkage.
If the bisque firing is reduced or not adequately oxidized and excessive gases are generated during certain stages of the glaze firing, these can affect the glaze-body bond.
The chemistry of glaze may be such that the surface tension of the melt encourages crawling (e.g. high alumina, high tin, significant chrome/manganese colorants, lack of fluxes of low surface tension).

Is there a problem with the body?

If the clay body contains soluble salts that come to the surface during drying, these can affect the fired melt's ability to form a glassy bond with the body. Precipitate these salts with a small body addition of barium carbonate (for information on how this works search for Barium Carbonate in the materials section).
As noted above, if the body surface is too smooth, the glaze may not be able to adhere properly.

Is this problem inherent in the type of glaze being used?

Matte glazes are more prone to crawling. Why? Because they usually have high Al₂O₃. The major contributor of that oxide is clay, especially kaolin. Matte glazes commonly have 35% kaolin in the recipe. Use part calcined material, part raw kaolin to deal with this problem.

Many pottery glazes have high feldspar and low clay contents, simply because they were improperly formulated. Using chemistry, you can shift the recipe to supply part of the Al₂O₃ from kaolin instead of the feldspar, reducing the feldspar percentage (this involves corrections in the amount of silica and other materials also).

Slip glazes can have 70, 80 or even 90% of a slip clay in them. Alberta Slip and Ravenscrag slip are examples. These materials melt by themselves to make glazes. But they are clays, they shrink. Follow the instructions from the manufacturer on how to use them properly.