Devitrification on Repeated Firing

 Devitrification is defined as the crystallisation of the glass, making it a non-vitreous substance.

Molecular level difference between vitreous and devitrified silica
from Digitalfire.com

You can see that there is not much difference between the the two states of the glass in structure, but mainly the arrangement of molecules.

The appearance of devitrification has a range of appearances from a mild smeary look through a dull surface to a crazed, crumbly aspect in severe cases. 

Mild devitrification

Medium level devitrification requiring abrasive cleaning

Causes of devitrification are related to slow changes of temperature (up or down) and most importantly nucleation points such as dust, oils, or cleaning residues. So, thorough cleaning is most important. 

Causes in repeated firings of the same piece relate to:

        Cleaning. 

It is important to thoroughly clean the piece before each subsequent firing.  Many times abrasive cleaning such as sandblasting is important to clean out impurities from the previous firing.  The resulting surface from any abrasive cleaning requires further cleaning with lots of clean water and a thorough drying with clean cloths or paper.

        Slow cooling or heating. 

Devitrification normally occurs in the range of 670⁰C to 750⁰C. This is the reason for the rapid rates of advance in this temperature range rather than other factors.  It can form both on the rise and on the fall in temperature. Slower rates in the devitrification range allow enough time for the crystallisation to begin.

        High temperatures.

Both high temperatures and long soaks can promote devitrification.  It is not just the slow rise or fall in temperature, but long periods at high temperature can lead to devitrification even though other precautions have been taken.

Changes in the composition

High temperatures and many repeated firings of the piece can lead to changes in the glass.  Some metals and fluxes are more likely than others to change composition or oxidise at extended soaks at high temperatures.  This can reduce the ability of the glass to resist devitrification.

Prevention/Correction

Prevention relates to thorough a) cleaning and b) firing rates.

All correction of devitrification relates to the modification of the surface.  If the problem is only at the surface, you can use either abrasive cleaning or the addition of fluxes to the surface, or a combination of the two. 

Where you have a mild dulling of the surface due to devitrification you can apply a flux.  This softens the surface by reducing the melting temperature of the glass and so reverses the crystallisation at the surface. The devitrification solution can be a proprietary spray such as Super Spray. Be aware that some sprays use lead particles as the flux, so are inappropriate for pieces intended to be food bearing. You can make your own devitrification solution by dissolving borax in distilled water.  When the devitrification is wide spread or deep, abrasive cleaning is required.

Abrasive cleaning can be by hand with sandpapers or diamond pads.  Be sure to keep them damp.  This keeps dust from rising, and the sanding surfaces clean for better working.  Sandblasting can be quicker, especially on uneven surfaces or where there are deep imperfections.  The surfaces resulting from abrasive cleaning need to be scrubbed clean with sufficient water, and then polished dry as for a finished piece.

It is possible to combine both these methods to be more certain of a shiny finish.  When combining, you need to do the abrasive cleaning first, then the wet cleaning and finally add the devitrification solution.

A fourth possibility is to sprinkle a fine but consistently thick layer of clear fine frit or powder over the piece.  This, when fused, provides the new surface concealing the devitrification below.  Again, this must be done at a full fuse, so it is not applicable to items you wish to remain tack fused.

However, if the devitrification has progressed to a crazed appearance, it is so deep as to be almost impossible to reverse.  The piece will also probably have developed incompatibilities. So the only real option in crazed pieces is to dispose of them.  They will not be useable in combination with any other glass. They will make any glass with which they are combined subject to devitrification and possible breakage.  These are pieces which truly cannot be cut up and re-used.

Annealing Multiple Levels of Tack Fusing

A question was asked of me about schedules for tack fusing multiple pieces – three layers thick in places – as a single unit, then placing on a 6mm fused base and tack fusing.  Special interest was in how the different thicknesses and the tack fusing would affect the scheduling of the annealing.

My response – edited – was as follows.

This is going to be a long reply.  I have written a general guide to tack fusing that will be useful, but this response will try to be more specific to your project.

First, tack fusing of pointed things is more sensitive to annealing than rounded things.  For up to 3 layers of triangles, I would be thinking of annealing for at least 12mm (four layers). This means a 2hour soak at 482⁰C, followed by a cooling rate of 55⁰C for the first 55⁰C degrees and then 99⁰C for the next 55⁰C. After this 110⁰C degrees of cooling the rate can be as fast as 330⁰C/hr.  This will apply whether Bullseye or System 96 is involved.

Second, from the description I take it that a 6mm clear under a 3mm layer of two colours side by side is being fused as a base.  [This was confirmed], so you could fire at 200⁰C/hr to a bubble squeeze of 30mins and then 300⁰C/hr to top temperature.  Anneal at 482⁰C for 60-90mins and cool for first 55⁰C at 65⁰C/hr and the next 55⁰C at 150⁰C/hr, followed by 300⁰C/hr to room temperature.

The third stage is to combine them.  Think about how thick this is physically – ca.18mm.  Then think about the differences in thickness – 9mm.  My rule of thumb is to add the difference between thicknesses to the thickest part – in this case to 18 plus 9 equals 27mm.  This is the “scheduling thickness” for this variation with rounded elements.  As your piece has lots of triangles, you need more care.  It is an additional level of difficulty.  So I add another 3mm to my “scheduling thickness” to accommodate the angular aspect of the piece, making a total of 30mm for putting the two fused pieces together. 

This thickness leads me to propose a relatively complicated schedule.  I suggest 70⁰C/hr to 250⁰C, 100⁰C/hr to 540, 120⁰C/hr to 620 and then 150⁰C/hr to top temperature.  The top temperature will be lower than your normal tack fuse temperature because this is a much slower rate of advance than normal.  This in turn, means that you will want to be checking at intervals on the tack fuse progress from at least 720⁰C.

The annealing will be long and slow. About 5 hours at 482⁰C, 11⁰C/hr to 427⁰C, 20⁰C /hr to 370⁰C and 65⁰C /hr to 30⁰C. This will be a schedule of about 35+ hours.

The two sources mentioned earlier give the rationale for this kind of schedule.  Think about the considerations I have listed, and then decide whether I am being too cautious or not.  The principle remains - as you increase the risk factors, you

·         slow down rates of advance and cooling rates, and

·         extend soak times.

You should note that I have used Graham Stone’s Firing Schedules for Glass, the Kiln Companion and the Bullseye chart for Annealing Thick Slabs in preparing the proposed schedule, although you will not find this exact schedule in either of them.

Carved Fibre Moulds

The question of whether you can use carved moulds more than once will arise.

This refers to moulds made from refractory boards or materials.  Once fired, refractory boards and materials become more fragile as they have lost their binders.  If the carving is simple with lots of support, and the mould is kept supported in a container of some sort, rigidising is not essential.  The life of the mould may be short though.

To make a longer lasting mould, you can rigidise the refractory material using this method.  This can apply to board as well as blanket.  The process will make a much longer lasting mould that is light weight, and is not affected by rapid changes in temperature.

Do the fibre moulds need kiln wash?

This depends on both the nature of the material and whether hardened or not.  Refractory fibre boards – often called ceramic fibre – do not need kiln wash to separate the glass from the mould.  However, putting powdered kiln wash and smoothing it with a piece of glass or plaster’s float can give a less grainy finish.  If applied wet, the dried kiln wash can be gently sanded to give a very smooth surface.

Other refractory boards such as calcium silicate or vermiculite do need kiln wash to separate the glass from the mould.

Any refractory mould which has been hardened with colloidal silica will need to be coated with kiln wash to keep the glass from sticking.  The kiln wash needs to be re-applied each time the mould is used above tack fusing temperatures.  Otherwise it does not need renewal until or unless the kiln wash is chipped, scratched or in other ways damaged. 

Another popular separator is boron nitride.  It is sold under various brand names.  This must be applied each time the mould is used.

Rounded Bottom on Drapes

Sometimes drapes, such as the handkerchief drape over a cocktail shaker, finish with a rounded base.

The base is rounded because not enough time or heat was allowed to get it flat. The glass will benefit from a moderate, but steady advance in temperature all the way to the top temperature.  This rate will be around 100°C to 150°C per hour.  There is no need to speed the rate of advance at any time during the process of the drape.  Too rapid an increase in temperature may even give uneven drapes if there are differences in thickness or colour.  There is no need for a soak at the strain point on the way to the top temperature. Any thermal stress from the rate of advance - that some suggest may occur - will already have taken place by this temperature.

This slower rate of advance will mean that the glass will not dome so much on the drape.  It will have time both to conform to the top (which will become the bottom of the piece) of the mould support during the drape stage. 

You need to visualise what the glass is doing during the forming process. As the glass begins to drape, the glass on the support rises because it is not yet soft enough to stay flat on the supporting mould. It is only later at higher temperatures, that the glass on top of the support can conform to it.

If you watch the process – a really good practice - you will be able to tell when you have a good drape. And with this reduced rate of advance, you should have a flat bottom. And all of this may happen at a lower temperature than you expected.