Steps in the firing process

The firing pattern

This is the pattern entered into a controller to regulate the progress of the firing. It follows the 5 stage profile shown in Fig 1 below.

Fig 1

However each stage may consist of one or more steps, depending on:

• the sophistication of the kiln instrumentation;
• the complexity of the firing; and
• the wishes and preferences of the operator.

Each of these stages will be discussed at length, and also the various conditions which may arise.

Please note the distinction between stages (as shown in Fig 1) and steps (the divisions made in the stages as shown in diagrams below) in the following explanations.

Firing Pattern: Stage 1

Heating the glass, moving from solid to transition

Is the initial heating stage, where the glass is heated through its ‘solid’ state and to the top of the transition zone, so that we know that it isn’t solid any more and can’t be damaged by thermal shock. It is heated through this stage at a rate slow enough to prevent it cracking. There are multiple ways to do this with various combinations of steps. 

A simple, one step stage is shown in Fig 2

Fig  2

Glass is a relatively poor conductor of heat and it takes time for the heat to flow within or through it.

Excessive or erratic rates of heating can cause too great a difference in temperature within the glass, and this in turn can cause excessive localised expansion and stress, often sufficient to crack it.This is most likely to occur during the first 200°C (400 °F) or so of a firing, so it is common practice for some glassies to use two steps for the stage, as in Fig 3.

Fig 3

 This entails a slower rate of rise up to about 200°C  (400°F) and a faster rate of rise to the top of the transition zone. For most situations this can be set at about 540°C to 550°C (1005°F to 1020°F). This is especially done when firing multi-coloured glass or when using expensive pieces of glass. Another reason that two steps are used in this stage is that, even with the most sophisticated controller, the heating curve of a kiln can be somewhat erratic at very low temperatures. This is especially the case when power switching is done by contactor or when the cycle time is long. (See 55. Power switching and 68. Programmable controllers for more on this.)

Sometimes a hold will be seen in a firing pattern at a low temperature, as in Fig 4.

Fig 4

It will usually be better to add the hold time to the time for the first step, making the time for that first step equal to what was previously steps 1 and 2. The rate of rise for the first step will be lower, it will be kinder to the glass, and the firing will take no longer.

Sometimes a ‘hold’ is introduced at the end of stage 1, as in Fig 5.

Fig. 5

The need for this depends greatly on the thickness of the glass and the rate of heating of the kiln. There’s little need when firing thin glass in a slow kiln, possibly more so with thicker glass. The faster the kiln heats the greater the difference between hottest and coldest points in the glass. By the time the kiln reaches the top of stage 1 the glass is at or close to the top of the transition zone. A slower rate of heating lower down would be of more benefit.

If the hold is for only one or two minutes it will be of little practical value. If for a considerably longer period, the hold time could be included in the heating step to lower the heating rate for the entire step, as shown by the red line.

When firing a stack of pieces, heat will flow unevenly through to the bottom layer. Whilst the kiln and top layer may be at transition, parts of the bottom piece could be still at a temperature where it will be thermally shocked. A hold as in Fig  5 or a slower overall rate of rise can both meet this situation.

A single step to transition as in Fig 2 is usually adequate when slumping float glass, or for the single colour rolled ‘leadlighting’ glasses. If in doubt as to where the transition zone lies, err on the side of caution and increase the top temperature for this stage. It can only add a short time to the firing.

In the case of paint decoration on glass, this is often fired at around 600°C (1110°F). If painting onto float it is normal practice to extend Stage 1 to this temperature, with a short soak of only a few minutes, as in Fig. 5.
In this case the ‘work has already been done’ and one can delete Stage 2 and go directly to Stage 3.

When fusing multiple layers of glass on a flat surface the heat can enter the lower layers only by passing through the top layer; except for that which enters from the sides of the stack. Each underneath layer will be subject to a greater temperature gradient between edges and centre  than the one above. This is due to the lower ones receiving less heat from those above but similar amounts through the edges. The heating rate for a stack should be slower than for one or two layers of a similar thickness glass.

Firing Pattern: Stage 2

‘Getting the work done’ on the glass

It can involve glass bending, slumping, fusing, casting, painting, or a combination of two or more of these techniques at once. In Stage 2 the glass is fluid: Restriction on the rate of heating is imposed more by the heating capacity of the kiln than by the glass. At its simplest, this can be a one step process where the glass is heated to a temperature and then immediately cooled, because the ‘work has been done’.

Fig  6

(Note: Steps are identified from here on in alphabetical references, as who knows how many steps have been used in stage 1)

This is often the way pottery is fired, albeit at considerably higher temperatures and usually at slower heating rates. Many older pottery kilns will provide only this level of control.

Fig  7

More frequently, the work is done on the glass in two steps; a heat step A and a soak step B, as in fig 7. The soak time can vary widely, from minutes when slumping thin glass to an hour or more for complex slumps and fuses.

It is in this stage that one becomes acutely aware of the effect of kiln performance on theoretical firing patterns. One author has calculated a theoretical Stage 2 heating rate for 3mm float glass of 30°C (54°F) per minute. It is a well designed and high powered kiln which can heat at better than about 5°C (9°F) per minute at 800°C (1480°F). Most are slower. Kiln performance does get in the way of theory.

Fig  8

Sometimes, an additional holding step is introduced in the heating part of the stage, as in Fig 8. Among the many reasons why this may be done are:

• allowing the temperature within individual layers of a complex piece to equalise;
• to allow the air trapped between layers of glass to expand and escape before the ‘sticking together of the edges’ traps the air and causes bubbles to develop.  

Firing Pattern: Stage 3 

Cools the kiln down from top soak to anneal soak temperature. 

The glass is fluid throughout this stage and can withstand cooling at fast rates. It can be as simple a process as allowing the heat to leak out through the kiln shell.

The rate at which natural cooling takes place can vary widely between kiln types, depending mainly on the efficiency of the kiln insulation and the amount of heat stored in the kiln.There can also be a lesser variation between firings in a particular kiln, depending on differences in the mass of the load in the kiln: glass, shelves, mould, or whatever.

The red lines show but two of many curves which a cooling kiln can follow when influenced by the above mentioned factors.

Fig  9

On kilns fitted with a modern programmable controller it is common practice to enter only a few minutes for this stage of the process. Of course, leaving the kiln closed will mean that the cooling rate will be much slower than programmed: but it can be speeded up if required.

Speeding up the cooling from soak (called venting, crash cooling, or Rapid Cool) can be for a number of reasons, including:

• To minimise the movement of the glass after the end of soak. With normal cooling the glass can continue to move after the power is turned off and until the temperature has dropped appreciably. Venting the kiln can minimise this effect

• To speed up the firing process, perhaps to get another load in or, dare one say it, because one is impatient.

Whatever the reason, this is one part of the firing cycle where the process can be speeded up, but the degree of risk in doing so varies with the type of kiln being used.   

See 23. Rapid cooling your kiln for more on this.

Firing Pattern: Stage 4

This takes the glass through the annealing process.

Annealing can be defined as “the process of controlled cooling of glass to prevent the building-in or retention of stress”.

Fig  10

A difference in temperature between points throughout the glass piece when it becomes solid means that the amount of shrinkage at the various points will differ as it cools to room temperature, and this builds in stress. These differences can be between the inside and the outside or between the top and bottom faces, anywhere in the piece.

In practical terms, annealing is a two step process, as in Fig 10, performed in the ‘anneal zone or range’. At the top is the ‘anneal soak’ step where the  glass is held at a constant temperature for as long as is necessary to allow it to equalise throughout the piece.

In the ‘anneal cool’ step the glass is cooled sufficiently slowly that temperature differences between various points is kept to a minimum: whether it be the top and bottom of a sheet on a hearth or the inside and outside of a slab or casting.

Obviously, a temperature gradient must exist during cooling, no matter how slowly the cooling takes place. Heat could not flow from the inside to the outside of the glass if a temperature difference didn’t exist. 

Fortunately, glass can withstand a certain amount of stress, so cooling at a rate appropriate to the glass thickness will keep the stress within safe limits.

The anneal cool takes the glass to the bottom of the transition zone so we know that the glass is solid and no more stress can be built into it.

In essence, that’s what annealing is about. The process is not complex, even though it may sometimes appear so because of the multiplicity of terms and approaches encountered in the literature.  

Anneal soak does not have to be done at one specific temperature. It can be done at temperatures higher or lower than those given by the glass maker for the reference points mentioned. The numbers stated in a glass data sheet are the makers take on the temperatures for the stated viscosity values.

The hotter the glass the more easily the particles can move to relieve the stress, so it follows that a longer soak time at a lower temperature can achieve a satisfactory result. Similarly a soak at a higher temperature for the same or a shorter time.

This is just as well, as otherwise the widely varying differences between glass and indicated kiln temperatures across kiln types and models could cause no end of problems.  There are limits on the amount one should deviate from the guidelines, but plus or minus 20°C (35°F) or so, with appropriate adjustments in soak time should work OK.

The time for the anneal soak step varies with the thickness of the glass. It can also be affected by the way in which the glass was cooled in Stage 3. The faster the rate of cooling the greater will be the temperature gradient across the glass when the kiln reaches anneal soak temperature.  

If the glass was crash cooled in Stage 3 there will be a greater temperature gradient across the glass than there would be if it cooled naturally and more slowly, so a longer anneal soak time may be required to allow the temperature to equalise. This is of little consequence when firing thin glass or fuses up to about 6mm, but may need consideration with thicker pieces.

The anneal cool step takes the glass down into the solid state at a rate sufficiently slow that stress is not re-introduced. Again, it varies with glass thickness; the thicker the glass the slower the rate of cooling.

Fig  11

Of course, the anneal cool step in reality won’t be a nice straight line. The pattern entered should describe the fastest rate of cooling acceptable to the particular piece of glass. The rate at which the temperature does drop will be determined by the rate that the heat can escape from the kiln.

The kiln is cooling by losing heat through the shell. As is shown in  Fig 12, the rate at which heat is lost is not constant, but is a gradually flattening curve.

Fig  12

When cooling in a kiln without a controller, or with the power switched off, the temperature will follow this curve as the heat is lost through the kiln insulation. The rate at which it is lost will depend on the thickness and quality of the kiln lining. This can limit the thickness of glass which can safely be annealed in a kiln without a programmable controller, such as in pottery kilns with a ‘Kiln Sitter’ or ‘Temperature switch’.

As indicated in Fig 11, a slow cooling kiln which is losing heat at a slower rate, makes it easier to manage glass annealing. In kilns with a programmable controller, should the temperature drop too fast, then the controller intervenes and applies extra heat to slow down the rate of cooling. In such a case there is no limit to the thickness of glass which can be safely annealed.

It helps greatly to know the heating and cooling curves of your kiln. Even when the kiln is fitted with a modern controller, who knows when one may need to let the glass cool naturally (such as with a prolonged power failure). Knowing your kiln will mean you can make wise decisions about which thickness of glass will be safe in these circumstances.

Firing Pattern: Stage 5

Cools the glass down to room temperature.

The glass is hot, but solid when the annealing is complete. The concern is no longer about building in stress, but only with avoiding thermal shock.

Thermal shock is immediate. It can happen only when the glass is in the solid state, and either when it is being heated or cooled. If a kilnformed piece is cracked when taken from the kiln, look at the edges of the crack. If they are sharp it cracked when cooling. If nicely rounded it happened on the way up.  

A well insulated kiln can take a long time to cool down, so rapid cooling is frequently resorted to.  Of course, the extent to which any pattern can be followed depends on a number of factors:

• The heating rate of the kiln.
• The cooling rate of the kiln.
• The level of sophistication of the kiln controls. 
  

See 23: Rapid cooling the kiln

Firing the platter

As an example, and not as any recommendation, the slumping pattern for the platter shown in Basic Kilnforming / Kilnforming a platter was:

The glass was 6mm float, the kiln a Riley GS-1 top-hat kiln with elements in the walls. The controller a Shinko model PCD-33A-S/M which can store 9 patterns each of 9 steps. Power switching is by Solid State Relay.

There are a number of ways of entering the time component into a controller. The method is chosen by the instrument manufacturer.

The Shinko company uses “time for the step”. Thus 540°C in 54 minutes equates to 10°C per minute or 600°C per hour.

Other makers use degrees change per minute whilst still others use degrees change per hour. This is discussed in 68 Programmable controllers.

No guarantee whatsoever is given that the above pattern will work in any other kiln or situation.

Compare the step times above with those recommended by the makers of fusible glass. Float glass is “tough stuff”. It can stand rough treatment; and just as well.

If the float glass makers had to follow the conservative annealing and cooling patterns of other makers they would never get any glass made.

As has been seen, float glass doesn’t require extended firing times, so devit is generally not a problem. However, multiple firings can cause problems.

There are many ‘glassies’ in Australia who use firing patterns for fusible glass much faster than those recommended by the makers. By all means follow the makers recommended patterns until one has a ‘feel’ for both the glass and the kiln. Then, experiment. Who knows what one will discover.

The total number of steps described in the stages above may well exceed the steps available with a particular controller, so the kilnformer must design the firing pattern to best take advantage of the steps of control available; having regard to the type and complexity of the firing.