Over Annealing

Sometimes the statement is made that you can never over anneal.  This statement is true only under certain circumstances. 

Annealing

The statement is also dependent on the understanding of what anneal means.  Annealing is the process of stabilising the temperature, ensuring the piece is at the same temperature throughout, and then gradually cooling the piece to avoid heat shock.  This is to point out that annealing is both the soak and the slow cool.

Long Soaks

Long soaks at the annealing stabilisation temperature can be injurious to your piece if the temperature in your kiln is not even.  This can mean that one or more parts of your piece are at different temperatures. This sets up stress within it.

Placing

You can reduce the possibility of stress by placing the piece at the centre of the kiln or avoid placing the piece in the cool spots of the kiln. 

Cool Rates

Another method of avoiding locking in the stress to the piece is to reduce the cooling rate to less than normal.  This will reduce the temperature differential within the piece.

Mass Being Cooled

In all this you need to remember that the anneal cool rate is relative to the mass of material to be cooled.  Therefore, a thick piece needs a slower annealing cool than a thin one.

But it is not just the thickness of the glass to be cooled.  You need to think about the mass of the kiln shelf or mould that supports the glass.  An example is that glass on a ceramic shelf needs slower cooling than one on a fibre board shelf, because the mass of the shelf needs to be taken into account as well as the glass. Connected to this is whether the shelf is on the floor - slower cooling - or supported on posts, allowing air to circulate under the shelf.

Peeking into the Kiln at Low Temperatures - Kiln Forming Myths 13

Do not peek between 100C and 540C – it will break the glass

Not necessarily.

To think about this systematically, you need to remember that the temperature readout is of the heat in the air, not of the glass.  On the way up, the temperature of the glass will be less than that of the air. On the way down the air temperature readout will be lower than the glass.

You can see from the readout how quickly the air temperature recovers to the original temperature.  This is generally, more quickly than the glass can lose its temperature.

The glass will be increasingly brittle as the temperature falls below the annealing point. 

The risk of thermal shock increases as the difference in air and glass temperature increases.  So shock is likely to be less at a readout of 100°C than at 400°C. 

The risk of thermal shock also increases with the thickness of the piece.  A piece of 25mm is more likely to be shocked at any given temperature than one of 6mm.

Whether you can peek depends on several things:

·        Temperature – e.g., just above the annealing soak a quick peek is less likely to cause problems than one at a lower temperature.  The shape of the glass will not change significantly below 600°C, so a peek while the kiln is cooling to the annealing point will not affect any but very thick pieces.

·        Length of peek – The key element is peeking is to affect the temperature as little as possible. So the opening should be as brief as possible.  The essential element in peeking is to take a mental snapshot of the glass, close the peep hole or lid, and think about what you saw.  Do not look or stare while the kiln is open. 

·        Size of opening – The smaller the opening you can manage during the peeking, the less risk of shock.  This is because less relatively cold air can enter the kiln.

·        Use of peep holes – set up your piece in the kiln in such a way that you can see your work through them.  At lower temperatures you will need the assistance of small intense light to illuminate the work.

·        Thickness – remember that thicker work is more likely to thermal shock because of the slow transfer of heat from the internal parts of the piece.  Peeking needs to be more cautious as the thickness increases.  Again, peeking above the annealing point should tell you everything you need to know about the final shape of the piece, making peeking in the brittle range unnecessary.

It is a good idea to minimise the viewing of your piece below annealing, but it is not impossible, if you follow the principle of avoiding drastic temperature falls during your peeking.

All myths have an element of truth in them otherwise they would not persist.

They also persist because people listen to the “rules” rather than thinking about the principles and applying them.  It is when you understand the principles that you can successfully break the “rules”.

Short Holds in Schedules - Kiln Forming Myths 5

Frequent short soaks on the way up will make a schedule safer

Safer in this context usually means less subject to thermal shock.  To determine the validity of this requires a bit of understanding on how the glass takes up heat, and as effected the lay-up.

Glass is a good insulator, both of heat and electricity, although we are only concerned about heat here.  This means that glass transmits heat poorly or, as it may be thought of, slowly.  A steady input of heat at an appropriate rate is less likely to shock the glass than quick rises with (catch up) soaks. 

In general, there is not much change in the rate required when you go over to a single rate without soaks.  For example, a ramp rate of 200°C from 20°C to 400°C with a 20 min soak, then 300°C to 540°C with another 20 minute soak could also be written as 193°C/hr to 540°C - both take 2.8 hours to achieve the same temperature. So the rate is not very different, but the way the heat is put into the glass is.

The glass is subject to heat shock below its softening point, and so rapid increases in temperature at the start of the schedule increase the risk of thermal shock below the 540C region.

When you have uneven coverage of the base glass, as most of us do, more care is required than when we have evenly thick glass.  This relates to the poor heat conductivity of glass.  The need is to have all the glass heat up at the same rate.  This is relatively simple when there are no partial layers on top as when doing a decorative tack fusing.  The pieces on top insulate the heat from the glass immediately below.  This gives a cool spot under the top glass, in relation the uncovered glass. To avoid this difference in temperature, which causes stress, becoming too great you need to slow the rate of advance as well as keeping it a steady increase.  This indicates you should be scheduling the rate of increase as though there were two more layers over the base glass.

The steady input of heat also becomes more important with thicker glass or more than two layers of glass.  The rate of heat input needs to decrease rapidly with increasing thickness – there is not a linear relationship.  For example, doubling the thickness from 6 to 12mm requires a reduction of 2.3 times the rate of advance.  Increasing the thickness by 4 times to 25mm requires a reduction of 10 times the 6mm rate of advance.

Other factors that require slower and steady increases in temperature are where you have dark and light glasses next to one another.  The same applies where you have a viscous and a less viscous glass together.  The classic is black, the least viscous of the glasses, and white, the most viscous.

However there is at least one circumstance where soaks are useful.  When draping over steel or ceramic, the free hanging glass heats up more than the centre where it is resting on the mould.  In this case, the mould forms a heat sink, drawing the heat away from the glass into itself.  You need to go very slowly or insert a few soaks to allow the supporting mould to heat up. More information here.

All myths have an element of truth in them otherwise they would not persist.

They also persist because people listen to the “rules” rather than thinking about the principles and applying them.  It is when you understand the principles that you can successfully break the “rules”.

Frit Making with Shock Treatment

Among the many ways of making frit, using thermal shock can be a simple way of producing significant quantities of frit.

The process:

Clean the cullet well.  It is not important to dry it as that will happen in the kiln.  Place the cleaned glass in a stainless steel container.  Take the temperature up to at least 300C as fast as you like – the glass is going to be fractured anyway.

The cullet in stainless steel bowl

360fusionglass.blogspot.co.uk

While the temperature is rising get a bucket or basin of cold water to place very near the opening of the kiln.  When the kiln has reached the temperature, switch off and open the kiln.   Reach in with heat resistant gloves and pull out the container.  Tip all the glass into the water.  It will steam and crackle, but no damage will occur, even to a plastic container.

The heated frit in water

360fusionglass.blogspot.co.uk

When the glass is cool (a few minutes) drain the water off and dry it, either in the kiln or on top or spread out on newspaper.  After the glass has dried you can break it up further with your hands, or any of the other ways of smashing glass into frit.

The fractured glass after drying and before breaking

360fusionglass.blogspot.co.uk

Cleaning:

My practice is to discard all of the very fine frit and powder resulting from this smashing process, as it is likely to be contaminated with other things, which can give a grey appearance to the work.  

However, you can use strong magnets to remove steel particles from the glass frit and powder.  Some have recommended the use of the magnetic trays used by car mechanics to help remove the steel contaminants. In both cases, the magnets should be covered in plastic to make cleaning of the magnets easier.  You simply take the plastic off the magnet or tray and shake the residue off the plastic, leaving an uncontaminated magnetic surface.

The magnets will not remove non magnetic materials such as a range of stainless steels, and other non ferrous metals. This requires you to use metals that can be magnetised as your breaking implements.  Also, magnets will not remove other non metal contaminants. This means it is important to clean the glass well at the start of the process and keep it clean throughout the breaking process.

Equalisation temperature

In my view a schedule has the following stages.

• Initial rate of advance to bubble squeeze,
• 
  
• Rapid increase in temperature to target, or working temperature,
• 
  
• Quick fall to temperature equalisation (often called the annealing point),
• 
  
• Slow decrease in temperature - to keep internal stresses at a minimum - to 110C below that temperature equalisation point,
• 
  
• Faster cool to 100C or less.

Of course, some of my firings have up to 10 segments, so don't mistake the stages as equivalent to schedule segments. The following graph is a generalised version of these stages.  The times and temperatures are for illustration only.

http://glassmuseum.moc.gov.tw/web-en/unit03/modepage/3-5-1-20.html

The equalisation temperature is what is most often called the annealing point. This is a mathematically determined temperature at which the glass most quickly anneals - has stress relieved. However, the way kiln formers work, annealing does not occur at one temperature point on the controller output, because of the inherent inaccuracy of our kilns and controllers. The soak at the annealing point has the purpose to equalise the temperature throughout the glass before proceeding to the anneal cool.

There is little point in soaking above this temperature, only to have another, lower temperature soak at the published annealing point. The soak at the annealing temperature will negate any effect of a soak at a higher temperature. So, a soak above the annealing temperature will simply slow the whole cooling process.

Of course, the soak at the equalisation temperature must be long enough to get the whole substance of the glass to the same temperature. The thickness of the glass will determine the length of this equalisation soak. Fortunately Bullseye have published a table to help determine the time required.

The slow decrease in temperature is to keep all the substance of the glass to within 5C difference on the cooling. Thus, the rate of cooling is related to the thickness of the glass. It will be increasingly slower with increasing thickness. The cooling to around 110C below the equalisation temperature is all part of the annealing process. The more rapid cooling after that is to control the rate of temperature fall to avoid thermal shock.

Thermal Shock

Thermal shock is a term for a break caused by a too rapid change of temperature within a piece of glass.

"Glass tends to be 

1) very brittle, 

2) expand and contract quickly when subjected to temperature changes, and 

3) is an insulator (when solid) and therefore does not readily conduct heat. 
That is why glass is highly susceptible to thermal shock"

http://www.glassfacts.info/indexf286.html?fid=210

This can occur on both an increase or decrease in temperature. Glass conducts heat poorly.  The ideal is to keep the temperature differentials within the glass to 5C or less.  This is the purpose of the anneal cool.  The risk of thermal shock can be increased by different thicknesses across the piece. Greater care is required in cooling these pieces than those of uniform thickness.

A piece showing large differences in thickness and so at greater risk of shock

Identification

The break normally is straight through the glass without following the edges of the various pieces of glass.

This shows the break crossing multiple colours of glass

The line of the break will be rounded if it parted on the heat up. In some cases, the glass will have stuck back together if it was dammed or the break was gentle enough to avoid pushing the glass apart.

If the shock occurred on the cool down, the edges will be sharp.  

The edges will also be sharp in a slump whether the break occurred on the advance or the reduction in temperature.  If the pieces fit together perfectly the break is likely to be in the down phase.  If the pieces are slightly different shapes the break likely occurred in the rise in temperature phase.

Other kinds of breaks are possible and are described elsewhere.

Diagnosis of Breaks in Kiln Formed Glass

Often more can be learned from failures than a number of successes. A common failure in kiln forming is broken glass. The appearance of the break will tell you a lot about the problem so that you know where to look for the solution.

Cracks and breaks can occur at various times in the kiln. These will have occurred by the time you open the kiln:

• Curved cracks and breaks are usually caused by inadequate annealing. Often the break will have a hook or sharp curve near the edge of the glass. The edges will be sharp.
• Cracks and breaks occurring where two pieces of glass meet is usually an indication of incompatibility between the two glasses. This means that you need to perform a compatibility test with the two glasses. Sometimes it is caused by a large difference in the thickness of the glass, especially when light and dark glasses are side by side. This is normally an annealing problem.
• Breaks in the piece (often more than one) with rounded edges indicate a thermal shock break caused by raising the temperature too quickly for the size or thickness of the piece.
• Breaks that cross the piece in a reasonably straight line, going across and through pieces of glass are an indication of thermal shock.  The line will be rounded or the pieces even formed together again if it was shocked on the rise in temperature.  If the piece was cooled too quickly, the edges will be sharp.
• Multiple breaks into small pieces - normally sharp - are an indication that the glass has stuck to the shelf or kiln furniture. This is caused by inadequate batt wash on the shelf and kiln furniture. It tends to happen with high temperature firings more than lower temperature firings.

Other cracks and breaks occur after the piece has cooled.
Breakage occurring long after a piece has been completed are an indication that the stress within the glass has overcome the strength of the piece. There are several possible individual and combined factors:
· improper annealing,
· thermal shock,
· incompatible glass,
· wear and tear.

But the most likely problem is inadequate annealing. Unless you have access to your firing records and can determine how the piece was fired and the materials used, you will need to accept it as experience and extend future annealing times.

The best cure for these is prevention.

First is to do a compatibility test to determine if the glasses fit together in the combination you plan for your piece.
Second, if you check the stresses of the flat piece between polarizing filters, you will be able to see if there are stresses within the piece before you do any further kiln forming with this glass or setup. If the stress is from incompatibility - where you see the stress halos around specific pieces of glass - you will need to destroy the piece. If the stress is more generalized, you can put the piece back in the kiln, reheat slowly and soak at the annealing point for a longer time and use a slower annealing cool.

Effects of Multiple Layers

Stacking layers of glass fully or partially over the base layer has significant effects on the firing of the whole piece.

Glass is a poor conductor of heat, so you need to be careful to allow the heat to penetrate to the base layer to avoid thermal shock. There also is the effect of the (very small) insulating space between each sheet. The effects of multiple, even layers can be seen from this table based on Graham Stone's* work:

3mm layers

1 sheet – Initial Rate of Advance =1000ºC to 475ºC (less than half an hour)

2 to 3 layers – IRA = 240ºC to 475ºC (ca. 2 hours)

4 layers – IRA = 100ºC to 475ºC (4.75 hours)

6 layers – IRA = 25ºC to 125ºC, then 30ºC to 250ºC, then 40ºC to 375, then 50ºC to 475 before 150C to top temperature (ca. 15.5 hours)

This shows the dramatic effect increasing the number of layers has on the firing schedule to make sure the heat gets to the bottom sheet evenly. If you compare the initial rates of advance (IRA) with the same thickness, but fewer sheets you can see the space between layers is important.

6mm layers

1 sheet – IRA = 320ºC to 475ºC (ca. 1.5 hrs)

2 layers – IRA = 240ºC to 475ºC (ca. 2 hrs compared to 4.75 hrs for 4 layers of 3mm)

3 layers – IRA = 200ºC to 475ºC (ca.2.5 hrs compared to 15.5 hrs for 6 layers of 3mm)

These are the fastest safe firing speeds for evenly covered sheets. 

This difference in firing times for stacks of thicker glass, shows how important it is to fire sections of the stack before the final firing of all the layers together.  It also reduces the risk of bubbles developing within the stack. 

If you are thinking of tack fusing with thicker and thinner areas, you need to take account of the differences in thickness in the various areas of the piece when preparing your schedule. You will need to decrease your IRA by quite a bit. So you might want to be thinking of firing some of your pieces to be added to the base layers before tacking them in an additional firing to reduce the risk of thermal shock to the base layer.

* Firing Schedules for Glass; the Kiln Companion, by Graham Stone, ISBN 0646 39733 8

Thick Uneven Pieces

Occasionally fused pieces come out of the kiln with one side thicker than the other. There are several things that need to be done for the present piece and for the future.


Level

First check how level your kiln is. The best for this is to begin with a check of the bed of the kiln. Check the level in four directions – left-right, front-back and the diagonals. If it is practical, wedge up the legs of the kiln to make the bed of the kiln as level as practical.

Then check how level your shelf is. Put in your shelf supports and then place the shelf on them. Again check with a spirit level the four directions. Place pieces of fibre paper under or on top of the supports to level the shelf. It is only after these checks have been made that you can consider firing your piece to help it return to an even thickness.  As part of your kiln maintenance you should check the level of your shelf at least monthly, if not every time you prepare to fire.

Variation in Thickness
Now that you know the shelf is level, you need to consider what the variation in thickness across the piece may be. The firing schedule needs to be more conservative than just for the thickest part. As the thinner parts will heat through more quickly than the thickest parts, you need to fire less quickly than you normally would for the thickest area. A rule of thumb – not always correct of course – is to add the difference of the thick and thin areas to the thicker and fire for that calculated thicknesses. This will make the firing schedule slower and so allow the thicker part to be the same temperature as the thinner. For example, a piece 6 mm at one side and 10 mm the other would have a difference of 4 mm. Add this 4 mm to the thicker 10 mm and then fire for 14 mm.

Temperature and Soak
You also need to consider the top temperature to use and the length of soak required. Glass flows relatively slowly at kiln forming temperatures. The conservative approach – one that allows further work if necessary – is to use the previous fusing temperature and extend the soak by at least twice the length of time on the previous firing, even perhaps to a couple of hours.

Bubbles
One thing that will happen is that the bubbles that previously were near the surface will rise and burst giving pin holes on this extended soak. So you should consider cleaning the bottom and putting the top face down on a separator between the shelf and the glass.   This will reverse the direction of flow for the bubbles. Few if any will break through the new top and there should be no pin holes when flipped.

Further Firings
When the piece is cool, check it for the even-ness of the piece all around. If it is not even enough, you will need to consider re-firing again. If you decide to do so, you should go no faster than the rate of advance as previously – probably even slower - but consider raising the temperature or extending the soak. Remember that achieving the heat work required at the lowest temperature is the guide line for kiln forming. So an extended soak should be preferred over a higher temperature, unless there are strong indications that a higher temperature is required.

Fire Polishing
Of course, you will now need to throughly clean the face down side and re-fire to fire polish the original top. The rate of advance should be the same or slower than the firing to even the thickness. Once you have achieved about 600C, a soak of about 30 minutes will ensure that the glass is thoroughly heated through. Then you can advance at a quick rate to the fire polish temperature with a soak of no more than a minute. This allows the surface to change without giving the rest of the glass time to begin to move.  Of course, a thorough annealing is required.

This procedure for re-firing  can be used when re-firing pieces for any reason. You only need eliminate the considerations on the uneven thicknesses.

Fibre Blanket Moulds

Fibre blanket moulds are good for free form moulds. The blanket can be cut into shapes or crumpled. It does not have binders as the papers do, so kiln wash is not necessary. Still, I have always sprinkled alumina hydrate powder over the mould. You can then support the high spots with kiln furniture – existing or custom made.

Pre-wetted fibre blanket is available - Moist Pack is one brand name.

Or you can make the mould yourself from fibre blanket and hardener. You need:

• ceramic fibre blanket. It should be 3 mm or thicker, but 25 mm needs to be compressed when wet. It is possible to use two layers of 3 mm fibre blanket, but they do not stick together well unless thoroughly wetted.
• colloidal silica - often is called mould hardener or rigidiser. Paint this onto the fibre blanket liberally, both sides if possible.


The rigidiser can be brushed on or sprayed on. Some people soak the blanket in the rigidiser and then squeeze out the excess.

You must protect the master with cling film, Vaseline, or other waterproof separator. Be sure about whether you want a draping or slumping mould, as the inside needs to be smoothest for a slumping mould and the outside smoothest for a draping mould.

Press the wet fibre blanket to the master. Then let it dry for a couple of days to become stiff enough to remove from the master. Let the negative dry for another period when out of the mould.

The drying method for rigidised fibre mould depends a bit on the structure from which you are taking the shape. If the shape is a piece of glass you can heat slowly to about 300C, but you have to be careful not to go much above that temperature to avoid the mould sticking to the glass. When cool you can carefully remove the mould from the glass and fire it to about 720C to cure it.

Other materials should be able to withstand at least 400C if you are drying in the kiln.

Materials that cannot be subjected to heat should be air dried. This will take a long time, possibly a week or more. The master should be coated with petroleum jelly or cling film to ensure the drying of the mould does not also cause it to stick to the master.

When the mould is dry, put into the kiln and fire to around 760C to cure the mould. You can fire fast, and after 10 minutes at 760C, you can just turn the kiln off, as there is no possibility of thermal shocking the mould. The point is to get the glass which has been in suspension to soften and stick together. Upon cooling the mould will be hard, as it is held together by the glass structure within the fibre blanket.

Once rigidised, you can sand the mould to refine the shapes. But you must use dust mask as the dust and fibres are dangerous to your health.  Do it out doors if possible. Otherwise a well ventilated room is necessary. You can sand down the high spots and generally smooth the mould to obtain a finer texture. Usually 100 grit sandpaper does the job quickly and leaves a relatively fine surface texture. If unhardened blanket is exposed during the sanding process, You can add a mixture of the rigidiser and the "dust" from the sanding to any holes, dimples or exposed unhardened fiber in the mould Then re-apply rigidiser to the sanded areas, and cure the mould at 760C again.

If you are rigidising, you need a separator – kiln wash – either powder or in a solution brushed on. A rigidiser does not burn off; it fuses to itself within the mould material and makes it harder. The resulting mould material will also be more brittle and should be handled with some care. I.e., never pick up the mould by the edges or with a piece of glass on top.

When you are satisfied with the shape and texture, you apply the kiln wash and fire.

The rate of heating the kiln and the soak will depend on the complexity of the shape of the mould and the thickness of the glass but there are no concerns about the mould as it is not subject to thermal shock.

With delicate treatment, the mould can be reused many times.

Two examples are shown here:

Lamp shade panel form

This is a "free form" mould made to give the glass sheet the appearance of crumpled paper

How Annealing Affects Slumping

It is often claimed that inadequate annealing of the fused blank can cause breakage during a slump firing.

If annealing is the cause, it is likely to break on the rise in temperature.  Once the piece has reached the annealing temperature, any breaks will be due to thermal shock on the way down.  An annealing break usually has a hook at both ends of the break, although this is more difficult to determine in a shaped piece.

Thermal shock tends to be along straight(ish) lines, often between thick and thin, or strongly contracting colours.  It tends to happen on the cool down. 

Breaks on the rise or fall in temperature are difficult to distinguish on slumps.  The temperature is low enough that there is little to distinguish the sharpness of the edges.  The real method of determining, is to try to fit the pieces together.  If they fit exactly, the break was during the cooling.  If they have even a little variation in fit, the break occurred on the rise in temperature.

If the annealing of the slump is marginally inadequate, it may break hours, days, weeks after cool.  The less stress the longer it will survive.  This will not be the result of any inadequate annealing of the fused blank. Only the last annealing is relevant to the soundness of the piece.

How can you ensure the annealing on a slumped piece is adequate?

You need to check the fused blank for stress before slumping to ensure it has no or very little stress.  The anneal for unstressed items needs to be at least equivalent to, or longer, as for the fused blank.

Fire more slowly than usual for blanks with moderate stress and anneal slumped piece more slowly than you did for the blank.  This will help ensure the formed piece is more adequately annealed than the mildly stressed blank.

Pieces with significant stress need to be returned to the kiln to be annealed.  Fire them significantly more slowly than you normally would for a piece that thick.  This may be one half or less the speed used on the un-fused pieces.