I hesitate to write an article on glaze formulation when there is so much on the Internet. Still, there are a few generalities that might be useful.

Some years ago I categorized many glaze formulations according to firing temperature and surface finish. I did this for lead and non-lead glazes both glossy and matt. I was amazed how easy it was to correlate composition with firing temperature (pyrometric cone or Buller ring).

You might run the same type of exercise if you are looking at many glaze compositions trying to come up with just one. You can start your formulation activities using the averages of the ranges of composition for each oxide addition for the temperature range you are shooting for.

The properties of a glaze can be calculated from the composition. For example, you can calculate the thermal expansion coefficient. In my experience, the calculation do not match the measured properties of the final glaze. However, relative calculations are usually true. You might not calculate the exact measured value but you will know that Glaze A will have a higher expansion and lower viscosity than Glaze B.

My associates formulated hundreds of glazes over the years for earthenware and fine china. Much of this work was done in a joint effort with frit suppliers as existing frits dont always do the trick in large production operations.

Here are some factors that I found to be important:

Glaze Composition

Lead is magic in glazes. It must be used in fritted form and in the minimum amount possible to maintain the proper glaze flow in the molten state. The glazed ware must pass all FDA and other restrictive testing. If lead is allowed, your problems are minimized. Your workers must be monitored for lead in the blood if lead air levels exceed OSHA standards. Borate frits can help reduce lead content but may increase solubility.

Complexity of formula is important. Every element you add to a glaze impacts the final-glaze properties. For example, too much alkali content will increase the solubility of a glaze, raise its thermal expansion, and in general raise havoc. A minimum amount of alkali content will give the glaze fusibility. Also, individual alkalis act differently in extent. Therefore one would use more than one alkali and the minimum amount of each that yields the best balance in properties.

The same is true of the alkaline earths, the glaze modifiers, and the glass formers.

You should consider the entire periodic table of the elements when formulating a glaze.

Remember that a very small addition of a particular element might give you a property you want, but increasing the amount only slightly can ruin your progress. As an inconvenience to glaze formulators, other elements interfere.

One other thing: You should use the minimum amount of binder. Sometimes its best to use several binders in small amounts rather than just one binder. Binders can be purified clays or organic compositions like gums or resins. Like I say, a mixture is usually best. None would be better.

To learn how to formulate gazes from scratch or using frits you can find it in my book Ceramics: Industrial Processing and Testing. Use http://tinyurl.com/gbxbz to obtain a new or used copy at Amazon.com. You dont need the second edition if you are only interested in glaze formulation.

Firing Conditions

First, from my experience glazes fire better in periodic kilns than in tunnel kilns. Matching the tunnel kiln cycle to the periodic kiln's cycle may not give the same results. The spatial configurations are different as is the atmosphere.

The firing curve naturally has heat up and cool down periods.

In between these slopes is a flat or modified soaking time/temperature.

During heat up, the binders are removed from the glaze.

They must be completely removed and they must not be reduced to carbon during heat up (preheat). This does not mean that you cant approach reducing conditions. Some compounds like MnO and FeO can greatly improve melting although they are often in the glaze only in tiny amounts as impurities. These compounds do not form under oxidizing conditions. Anyway, do not reduce the binders to carbon. Its near impossible to remove for many glazes during the rest of the firing.

There are a number of instruments to determine the nature of the burnout for a particular binder. Thermal gravimetric analysis (TGA) and Differential Thermal Analysis (DTA) come to this old head.

The preheat or heat up is usually a production standard for tunnel kilns. Therefore if you are not getting what you are looking for, you must adjust the burners or heating input in this zone. You dont want bubbles from binder forming all during the soak period. You have enough bubble problems without that.

During the soak phase and at all higher temperatures the glaze is changing composition. That is, the glaze is loosing volatile elements because of the high temperatures. This can be complex as when one component of a glaze is removed it may increase the volatility of other components.

For you chemical engineers, this would be somewhat like steam distillation.

When you lose volatile elements you also increase glaze viscosity in the molten state. For this reason, and others I suppose, fast firing is often better than a long firing cycle. In fact, Ive seen glazes that required fast firing.

Cooling is important to both the glaze and ceramic ware. You can cool quickly to just above the silica phase transition and then cool slowly through the transitions. This may not be good for the glaze which is trying to get rid of the bubbles created after the glaze melted. This can put you between a rock and a hard place. You have to be able to control the whole cooling zone of a tunnel kiln from the minute the ware comes out of the hot zone.