Glueing Glass Pieces

The best solution is to avoid the use of glue completely. If you cannot, use as little as possible and make sure it burns out cleanly.

The glues to which kiln workers have normal access, do not survive to tack fusing temperatures. Therefore they can only be considered as a means to get the glass assembly to the kiln. The glue will not hold the pieces in place until the glass begins to stick, so the pieces must have a stable placement. If not, the pieces will slip, roll and move once the glue has burned out.

The second requirement of glues is that they burn out without leaving a residue.

Glues that have been used with little or no residue include:

Powdered CMC that can be disolved in warm water

-CMC (carbylmethylcellulose). It is a cellulose based binder used in a wide variety of industries, including food. For our purposes, it is also used in the ceramics industry and is often called glaze binder. It is a main constituent of "glas tac" from Bullseye. This can be made up into a viscous solution to catch and hold frits and other sprinkled elements in place.

- PVA (Polyvinyl Acetate) is water-based glue. It is sometimes known as school glue. It can be diluted to about 10parts water to 1 part PVA. This is sufficient to hold the glass pieces together with only a drop for each piece of glass. It does not work so well for small sprinkled elements.

One of many brands of  Ethyl Cyanoacrylate
 glue

- Super glue burns off with no concerns about cyanide. It should be used sparingly and also works best for pieces of glass.

One of many hair laquers in pump spray bottles

- Hair lacquer is normally applied as drops at the edges of the assembled pieces and so can be used to hold pieces of glass as well as sprinkled elements.

In all uses of glue the principles to remember are:
- Use the minimum to hold pieces together while getting the work into the kiln.
- Put the glue at the edges of the glass or where its combustion gasses can escape easily.
- And in all cases, you need to test to see if a residue is left on the glass at full fuse when using a new glue.

An alternative to glue is frit as described here.

Effect of Heat on Sandblasted textures

This is based on Graham Stone’s work with float glass. The temperatures are applicable to float glass, and so need to be adjusted for other glasses, but illustrate the principle of how heating temperatures affect the glass.
Temperatures in degrees Celsius.

650 Blasted surface softened, evened, less "brutal".
690 Blasting still opaque but less "white"
700 Blasting becoming too sheeny but still okay for certain effects.
740 Blasting now subtle and glossy

Based on Firing Schedules for Glass; the Kiln Companion, by Graham Stone, Melbourne, 2000, ISBN 0-646-39733-8, p24

Approximate Temperature Characteristics of Various Glasses

Various glasses have different temperature characteristics. This listing is an attempt to indicate the differences between a variety of popular glasses used in kiln forming. They are not necessarily exact, but do give an indication of differences.

Bullseye Transparents
Full fusing 832C
Tack fusing 777C
Softening 677C
Annealing 532C
Strain point 493C

Bullseye Opalescents
Full fusing 843C
Tack fusing 788C
Softening 688C
Annealing 502C
Strain point 463C

Bullseye Gold Bearing Glasses
Full fusing 788C
Tack fusing 732C
Softening 632C
Annealing 472C
Strain point 438C

Desag GNA
Full fusing 857C
Tack fusing 802C
Softening 718C
Annealing 530C
Strain point 454C

Float Glass
Full fusing 835C
Tack fusing 760C
Softening 720C
Annealing 530C
Strain point 454C

Oceanside
Full fusing 788C
Tack fusing 718C
Softening 677C
Annealing 510C
Strain point 371C

Wasser
Full fusing 816C
Tack fusing 760C
Softening 670C
Annealing 510C
Strain point 343C

Wissmach 90
full fusing  777C
Tack fusing
Softening  688C
Annealing  510C
Strain point

Wissmach 96
Full fusing  777C
Tack fusing
Softening  688C
Annealing  510C
Strain point

Youghiogheny 96
Full fusing  773C
Tack fusing  725C
Softening  662C
Annealing  510C
Strain point

Polishing with Cerium Oxide

If you want to go beyond cork in your polishing, cerium oxide will give an optical polish.

You need to grind your glass at 800 or higher grit, or use the cork belt after 400grit. Any rougher surface will not give a smooth polished surface. It will only polish the high spots.

Many do not like to use cerium oxide as it is messy. Especially so on a wet belt sander as the speed is really too fast for the use of polishing pastes. The speed sprays the slurry all over the place.

You need a felt wheel or belt to which you apply the cerium oxide. First you mix the cerium with water to a yoghurt consistency and apply that to the wheel or belt. Begin polishing and add more water and cerium paste as the polishing surface dries. You will notice this as the glass will begin to drag. Do not delay, add more of the paste before continuing. Otherwise you will heat up the glass and risk breakage.

His Glassworks has good descriptions and videos on use of cerium oxide.

It is helpful to mark the glass with a paint or china marker before starting the polishing process to show the areas that are to be polished. This enables you to see what work has been done without completely drying the piece.

For large surfaces you will need to use a horizontal grinder with a polishing pad attached, or a hand held polisher.

An alternative is to use “trizact” belts that are about 4000 grit. These achieve a polish that is very good, if not as optical as with cerium oxide.

Viscosity Changes with Temperature

This image is taken from Pate de Verre and Kiln Casting of Glass, by Jim Kervin and Dan Fenton, Glass Wear Studios, 2002, p.27.

It shows in graphic form how the viscosity of glass decreases with increases in temperature. The temperatures are given in Fahrenheit.  

The coefficient of expansion also changes with temperature. 

This graph is also from Kervin and Fenton

 It is these two forces of viscosity and expansion that must be balanced around the annealing point to give a stable and compatible range of fusing glass.

Drop Rings

Mould

It is possible to purchase drop rings of various sizes. It is also easy to construct one from vermiculite board or ceramic fibre board. Merely cut a circle of the desired radius from the board. Leave at least 50mm of board outside the circle, and more for thinner boards.

Kiln wash the top and inner sides of the drop ring

Glass

The glass should be larger than the hole in the ring. This will vary by radius of the hole. The glass will need to be from 50mm larger diameter than the hole for smaller holes to 100mm larger diameter for holes over 300mm.

Glass should be at least 6mm thick for the first 100mm of drop and an additional 3mm for each 50mm more. So, a drop of 200mm would require glass of 12mm thick


Temperatures

The temperature rise should be no more than 150C per hour to about 675C for 6mm glass and less for thicker glass. Remember the glass is much closer to the elements than normal and it is easy to thermal shock the glass.

With close inspection you can see that the edge of the glass rises from the mould as it sinks in the middle.

The outside edges of the glass rise from the mould as the centre begins to drop in the centre.  As the glass gets hotter, this raised edge settles back on to the mould.  If the glass is really near the elements, there is a small risk the glass will touch the elements.  No harm will be done to the kiln, but the glass edge may have some needles.

The rate and amount of slumping is controlled by temperature, span (the width of unsupported glass on the mould) and time. The higher the temperature the faster a piece will slump and the thinner the walls will be. However you can slump at lower temperatures by holding the temperature for a longer time to reduce the thinning of the sides.

Also note that the wider the span, the faster the glass slumps.

If you slump at high temperatures with a drop ring the sides of the bowl tend to be straight and steep. The strain is limited to the region immediately inside the rim. Therefore the glass tends to thin next to the rim and the colours are diluted. If you slump at a lower temperature for a longer period of time the strain is distributed over the entire unsupported area. This results in a more rounded shape for the bowl and even thickness of the glass across the bottom of the bowl.


Experiment

Finding the right combination of time and temperature requires a bit of experience and guess work. If you want a rounded bottom, heat the glass to the point that it starts to bend on the mould and wait for 30 minutes. If it has slumped about 1 inch in that time wait another 30 minutes. You are looking for a slumping rate that is acceptable. If it hasn't moved very much then increase the temperature 15C and check again in 15 minutes. Keep moving temp up and waiting for 15 minutes until the piece has completely slumped. This might take several hours.

If you want straight sides keep heating the piece rapidly.

Stopping
When the piece has slumped to the desired shape, flash cool the kiln to about 30C above the annealing point to stop movement in the glass. Extend the annealing soak and increase the length of the annealing cool time (reduce the rate of temperature fall) over normal slump firings of the same thickness.

Glass falls through drop rings in relation to the size of the glass on the drop ring, the size of the opening, the temperature rise rate and to some extent the colours and amount of opalescent glass used. 

Mould Cleaning

There are a variety of moulds available to kilnformers – slumping/draping, texture and casting are currently popular ones.  Each has a slightly different maintenance regime.

Slumping and Draping

Slumping and Draping moulds are the easiest to maintain, as they are not taken to high temperatures. Normally one application of kiln wash will last very many firings.  The kiln wash needs to be renewed when bare spots appear on the mould.  Some people immerse their moulds in water to wash off the old kiln wash.  This is excessive and requires a long slow drying time for ceramic before you can re-apply the separator.  I’m not even sure the practice is good for ceramic moulds.

Normally, you only need light abrasion such as with a green washing up scrubby to clean off the old kiln wash. You can also use a small nylon brush to take off the old kiln wash and prepare it for a new application.

If you are using boron nitride on your slumping or draping mould, you need to brush off the old separator each time you fire the mould. And then renew the boron nitride surface to ensure there is no sticking.

Texture Moulds

Texture moulds require cleaning before applying any additional separator to avoid blurring or obscuring the textures of the mould.  It is best to use a kiln wash that does not have a lot of china clay in it, such as Primo, to allow easy brushing of the separator off.  If you use a kiln wash with little or no china clay, you will need to clean and re-apply each time you prepare to fire the mould.  Boron nitride works well for texture moulds, but also needs to be carefully brushed off the mould before re-applying the separator in preparation for the next firing.  This is both to avoid blurring the texture and to ensure there is sufficient separator to avoid sticking.

Casting

Casting moulds that are intended to be re-used multiple times are best coated with boron nitride.  The boron nitride should be lightly brushed off after each use to ensure the detail is retained, and then re-coated.  A nylon brush is good for this.

The materials and purposes of moulds have an effect on the separators used and the methods of maintaining them clean.

Wire for Fusing

Although there are other ways to combine wire with glass, one popular method involves fusing wire inside the glass. This technique generally fuses and seals the wire between two layers of glass, so it is important to select a wire with the right characteristics. The main characteristics are:

1. The wire must be capable of withstanding the heat of the kiln.

2. The wire must emerge from the kiln in a relatively pristine condition, or at least can be easily cleaned.

3. The wire must also retain the desired flexibility and pliability. If it's too soft or brittle it may not support the piece.

4. The wire must not react with or contaminate the glass. In most cases colour changes and metal flakes are not desirable.

5. The wire must be of a small enough diameter to avoid causing excessive stress within the glass.

6. It is a bonus if the wire is reasonably priced or even inexpensive.

This post gives the characteristics of some types of wire for fusing. 

Paint Markers to Identify Scratches in Grinding

As you pass from one grit to another, it is often difficult to tell if you have really removed the grinding marks from the previous grit.

By using a paint marker, you can mark at random over the ground surface. Let it dry before you began the next grit.

As you grind with the finer grit you will be able to check that you have removed the glass down to the depth of the deepest scratch by the absence of the paint. Of course drying the glass and making a visual check of the surface will provide insurance. If you are satisfied with this stage, paint again before changing to a finer grit. This allows the paint to dry before you begin grinding again.

Types of Wire for Fusing

Having mentioned the characteristics needed of the wires for inclusion, this is a description of the good and bad points of some common wires used as inclusions within glass.

Nichrome (nickel chromium) is a generally favoured wire, due to it easy workability, ability to hold up in the kiln and maintain its strength afterwards. It does turn dull after firing, but can be cleaned up with a brass wire brush.

Copper is a softer wire to use, and exposed parts tend to be weakened. It may tarnish or change colour. Some twisted/braided copper can work better than single strand copper, but test first.

Sterling silver will work, but tends to scale and needs to be cleaned after firing. It can react with the glass and change colour. It tends to be soft after firing.

Fine (pure) silver works better than sterling, but even more prone to react with the glass - turning yellow. Some glasses (French vanilla and certain reds) will also change colour when exposed to silver.

Stainless steel is very stiff and hard to work with, but can be fused if desired. It retains its strength and if of the appropriate grade requires only treatment with a brass wire brush.

Gold or platinum wires will work, but are very expensive.

Damming Ovals

There are various ways of damming oval shapes in kiln forming. Some of these are outlined here.

One set of methods depends on having a soft surface such as ceramic Fibre board or vermiculite.

Photo from Clearwater Studio

You can wrap your shape with fibre paper. For this you need to cut a strip or strips 3mm narrower than the height of the piece you are wrapping. You then stick sewing pins down through the fibre paper and into the shelf of fibre board or vermiculite. This will be easiest if you use 1 to 3mm thick fibre paper, as the pins must not contact the glass – the pins will stick to the glass if they do.

You can cut a form out of ceramic fibre board and use that as a dam. You can pin this to the base fibre board or allow it to merely rest on the board. It is possible to cut arcs from fibre board and place them around in sections. In this case they will need to be pinned together so they do not move apart. Staples can form the attachments. You can make your own – larger – ones from copper wire.

You can buy stainless steel banding which needs to be lined with any separator – batt wash or fibre paper.

Bonny Doon stainless steel dams

You also can layer fibre paper up to the height required – remember 3mm less than the thickness of the piece. You then need to fasten the layers together to avoid movement between the layers.


If you are firing on ceramic kiln shelves the same materials can be used but need to be supported a little differently.

If you are wrapping the piece on mullite shelves, use some pieces of kiln furniture to block the strips up against the glass. The thicker the glass, the more weight will be pushing out against the dams and the sturdier the dams will need to be. Make sure the strips contact the shelf evenly- if you have gaps, you'll have leaks.

The disadvantage to this method is that the glass can take up the irregularities of the kiln furniture.

You can use fibre board with a void cut out to the shape required and place it on the shelf.

You can also use layers of fiber paper around the shape and pin the layers to each other. This is the same method as used on ceramic fibre board.

Again stainless steel can be used to form the dam. Remember to line the steel with fibre paper that is 3mm narrower than the height of the piece.



In all these cases of dammed forms, the edges will be of varying degrees of roughness and some cold working will be required.

Slow and Low

Low and Slow Approach to Kilnforming

We are often impatient in firing our pieces and fire much more quickly than we need. After all, our computerised controllers will look after the firing overnight. So there is no need to hurry more than that.

The concept of heat work is essential to understanding why the slow and low method of firing works. Glass is a poor conductor of heat which leads to many of our problems with quick firings. The main one is stressing the glass so much by the temperature differential between the top and the bottom that the glass breaks. We need to get heat into the whole mass of the glass as evenly and with as smooth a temperature gradient as possible. If we can do that, the kiln forming processes work much better. If you add the heat to the glass quickly, you need to go to a higher temperature to achieve the desired result than if you add the heat more slowly to allow the heat to permeate the whole thickness of the piece.

Graphs of the difference (blue line) between upper and lower surfaces of glass of different thicknesses against cooling time

However, this slower heating means that the glass at the bottom has absorbed the required heat at a lower temperature than in a fast heat. This in turn means that you do not need to go to such a high heat. This has a significant advantage in forming the glass, as the lower temperature required to achieve the shape means that the bottom of the glass is less marked. The glass will have less chance of stress at the annealing stage of the kiln forming process as it will be of a more equal temperature even before the temperature equalisation process begins at the annealing soak temperature.

Applying the principles of low and slow means:

• heat is added evenly to the whole thickness of the piece
• there is a reduction in risk of thermal shock
• the glass will achieve the desired effect at a reduced temperature

The alternative - quick ramps with soaks – leads to a range of difficulties:

• The introduction of heat differentials within the glass. Bullseye research shows that on cooling, a heat difference of greater than 5ºC between the internal and external parts of glass lead to stresses that cannot be resolved without re-heating to above the annealing point with a significant soak to once again equalise the heat throughout the piece.
• It does not save much if any time, As the glass reacts better to a steady introduction of heat. Merely slowing the rate to occupy the same amount of time as the ramp and soak together occupy, will lead to fewer problems.
• It can soften some parts more quickly than others, e.g., edges soften and stick trapping air.
• Quick heating, with “catch up” soaks, of a piece with different types and colours of glass is more likely to cause problems of shock, bubbles, and uneven forming.
• Pieces with uneven thicknesses, such as those intended for tack fusing, will have significant differences in temperature at the bottom.
• Rapid heating with soaks during slumping and draping processes can cause uneven slumps through colour or thickness differences, or even a tear in the bottom because the top is so much more plastic than the bottom.

However there are occasions where soaks during the initial advance in heat are useful:

• for really thick glass,
• For multiple - 3 or more - layers of glass,
• for glass on difficult moulds,
• for glass supported at a single internal point with other glass free from contact with mould as on many drapes.

Of course, if you are doing small or jewellery scale work, then you can ignore these principles as the heat is gained relatively easily. It is only when you increase the scale that these principles will have an obvious effect.

Slow, gradual input of heat to glass leads to the ability to fire at lower temperatures to achieve the desired results, with less marking and less risk of breaking.

Further information is available in the ebook Low Temperature Kiln Forming.

Are Holes Needed in Stainless Steel Moulds?

 “Do you drill holes in the bottom of the stainless steel moulds the same as with the ceramic ones? I imagine so, as the air issue is the same?”

When draping over stainless, holes are not required unless there is a depression at the highest point.  It is debatable whether required even then.  The steel is expanding more than the glass during the heat up and contracting more on cool down.  This effect means there is sufficient space for any air to escape.

In slumping moulds, stainless needs to have a significant draft to avoid the steel trapping the glass during its greater contraction during the cooling.  The combination of the draft and the greater expansion during heating allows air to flow from under the glass, unlike ceramic where the glass is the faster expanding material.  The greater expansion of steel leads to less chance of the glass sealing to the mould and creating bubbles. 

However, there is no harm in being cautious by drilling small holes at the last places the glass will touch down.  These usually are at the join of the curve and the flat bottom.  The glass will touch down first in the middle of the bottom, so no hole is required there.

Freeze and Fuse

"Freeze and fuse" is a term devised to describe a technique to obtain complex edge shapes and some bas relief.

The basic method is as follows, although there are a number of variations that can be successfully adopted.

Mix enough water with fine frit to make a damp slurry.

Then place about 3mm into your mould and tap on a hard surface. Tap quite vigorously to bring any air bubbles to the top and compact your powder.

Use a paper towel at this point and blot off any water that has risen to the surface.

Continue to layer, tap and blot until you're level with the top of the mould.

The more you tap and blot out any water (and every time you tap, more will rise to the top) the better your results will be.

When your paper towel won't absorb any more water, you're ready to put your mould into the freezer. One to three hours should be enough, but it must be frozen throughout.

Take the frozen glass from the mould. Letting it sit while you programme the kiln will allow it to come from the mould more easily. Place the glass form on the kiln shelf. Raise the temperature as fast as you like to 90C. Soak there for at least half an hour to remove any water in the piece. Then raise the temperature at about 100C/hour (depending on the thickness and size of the glass form) to a low fire polishing or sintering temperature (about 720C to 740C). Higher temperatures will flatten the form and change its shape. Soak at this sintering temperature for an hour or so.  Check on the progress of the firing by peeking at 10 minute intervals and advance to the next segment of the schedule when  the surface begins to shine.  

If you are planning another firing, you should not fire beyond the first hint of a shine appearing.

Some experimentation is required to get the best combination of rate, time and temperature.

Experience will show you variations on this basic method.

Further information is available in the ebook Low Temperature Kiln Forming.

Air Brushing onto Glass

Air brushing paint and enamels onto glass can give extremely subtle graduations of colour and tone.

The consistency of the mix of the paint or enamel with the medium will need to be many times thinner than that used for painting with a brush. Also the air brush will need to be able to cope with the relatively large (in relation to inks and other paint) particles that make up the glass paint. This may require a little experimentation to find nozzle sizes that can cope with the glass paint particles.

There are two main media that you can use – there are others of course. Water and alcohol or methylated spirits are common and easily available. The advantage of spirits is that it evaporates from the surface more quickly. You do not need to use any gum arabic in the mix to help it flow. You could however add a touch of washing up liquid to overcome any surface tension within the mix.

The paint should be applied in steady sweeps across the area to be covered about 300mm from the surface. Start moving your brush before you switch on the paint and keep moving after you stop applying the paint. This avoids heavy applications at the start and finish. As there is no absorption of the moisture by the glass (as there would be on paper or board), you must pause after a very few passes. This is where the spirits show their value, as they evaporate more quickly than water, allowing you to apply the next layer sooner.

You can assist the drying by using a hair drier to gently blow warm air over the surface. This will also reduce the waiting time between applications.

One thing you will notice is that the paint will settle within the medium unless you agitate it frequently. So you should make sure the paint is evenly dispersed within the container by agitating it before starting each layer. The movement of the air brush during application will be enough to keep the paint suspended in the medium while you are applying the glass paint or enamels.

Heat Work is Cumulative

“…. the first fuse (contour) I brought it up to 1385°F and held for 5 minutes - it did not contour as much as I would like - do I re-fire at same temp and hold longer or go up in temp and hold same amount of time or something else?”

Observe

Of course, the smart answer is “Observe to get it right first time”.   Observation will enable you to determine when the piece is fully fired.  To observe you need only peek at 5-minute intervals to determine if the piece is as wanted. 

Know your Controller

In combination with this you will need to know your controller well enough to be able to advance to the next segment if the piece is done before the segment finishes; or how to stay on the same segment until it is finished and then advance to the next segment.

Of course, there are circumstance when you cannot or do not want to be present at the top temperature of the firing.  Then consider using the delay function to enable you to be present. This gives a countdown until the kiln starts.  The practice is fully described in this blog entry.

Time or Temperature

If you are experiencing an under-fired piece and want to re-fire it to get a better finish, the usual question is whether to fire for longer or at a higher temperature.

The response is – “Neither”.

Re-fire to the same temperature and time as before, unless you are looking for a radically different appearance.  Heat work is cumulative.  You have put heat into the glass to get the (under fired) result.  By firing it again, the heat will begin to work on the glass as it rises in temperature.  The piece, in this instance, is already a slight contour.  The additional heat of this second firing will begin to work just where the first firing did, and will additionally change the existing surface just as the first firing did.  The degree of contour achieved by the first firing will be added to equally in the second firing.  It is of course, a good idea to peek in near the top temperature to be sure you are getting what you want. More information on heat work is available here with its links. 

Rate of Advance

It is important to remember that on the second firing the glass is thicker, and you need to schedule a slower rate of advance until you get past the strain point – about 540°C for fusing glasses, higher for float and bottle glasses.

Future firings

At the finish of the second firing you will have soaked at the top temperature for twice the scheduled time.  You can use this extra time for the next similar firing, or increase the temperature slightly and keep the original firing’s length of soak. 

As pointed out earlier, observation for new layups, sizes, thicknesses, etc., is important to getting the effect you want the first time.

Equalising Effects on Both Sides of the Glass in the Same Firing

The desire is to have the same degree of fusing on both sides of the glass.  An example is where a person wants to have their colourline paints equally matured on both sides of the glass in one firing.  This is difficult and requires a different strategy than normal fusing.

Background

A bit of background first. Glass is a very good insulator. This means that heat travels slowly through the glass. Its practical effect is that we have wavy lines on the top and very crisp lines on the bottom.  This results from the temperature differential between the two surfaces.  This can be many degrees different during the plastic phase of the glass.  It is dependent on how fast the temperature rise is.  The faster the rise in temperature, the greater the difference as the glass transmits the heat from top to bottom so slowly.  The problem is how to keep the temperature differential as small as possible.

Heat Work

The concept of heat work relates to the way heat is put into the glass.  It can be done quickly to a high temperature, or slowly to a low temperature and still get the same effect.  This shows glass reacts to the combination of temperature and time. Putting heat into the glass slowly allows lower temperatures to be used to achieve the desired effect, than fast rises in temperature.

The insulating properties of glass means that the heat work needs to be applied slowly to achieve similar temperatures on both sides of the glass.  The thicker the glass the longer it will take to temperature equalisation.

The mass of materials also needs to be considered.  The glass will normally be on a ceramic shelf of 15mm to 19mm.  This mass also needs to heat up to the temperature of the top of the glass.  Until it does, it will draw heat from the glass.  This also points to the need for slow heat input.

The question that prompted this note was how to get glass strainers paints to have the same degree of maturation on both sides at the same time.  The maturation temperature of Reusche tracing paints is around 650°C.  If you use a normal rate of advance – say, 200°C – the bottom of the glass will be considerably cooler than the top.  This is both because of the insulating properties of the glass and the mass of the shelf.

Methods to achieve the effect.

Some methods are worthy of consideration separately or in combination.

Use refractory fibre board as shelf.  This dramatically reduces the mass of the shelf to be heated up.  This kind of shelf requires more care to avoid damage than a ceramic shelf.  It would be possible to place smaller fibre shelves on top of the standard ceramic shelf rather than having one large fibre board shelf.  This will not be so efficient an insulating mass as fibre board on its own.  Also, it will not be sufficient on its own to obtain equal temperatures on both sides of the glass.

Use 3-6mm refractory fibre paper between shelf and glass.  This again reduces the heat sink effect of the ceramic shelf, but not as much as a fibre shelf on its own.  Again, the fibre paper on its own is not enough. The scheduling is important.

Use very slow rates of advance.  A slow rate of advance in temperature is important to achieving equal temperatures throughout the glass.  Even using 3mm glass, the rate of advance might need to be as slow as 50°C per hour.  The corollary of this is that you will not need to use as high a temperature to achieve the effect.  Heat work means that it is not an absolute temperature that will achieve the effect.  The slower you put the heat into the glass the lower temperature required.  The understanding of this relationship will require experimentation to establish the relationship to the rate of advance and the top temperature required.  For example, a satin polish of a sandblasted surface can occur at 650°C, if held there for 90 minutes.

In this case, a 50°C rate of advance will probably not require more than 600°C – and probably less - to achieve the shiny surface normally achieved at 660°C with a 200°C rate of advance.  At 50°C per hour, it will take 12 hours to reach 600°C, although a little more than 3.25 hours at an advance of 200°C to reach 660°C.  The input of heat acts upon the glass throughout the process, making lower working temperatures possible.  The reduction in temperature required is not directly related to the reduction in the rate of advance.  You will have to observe during the experimental phase of this kind of process.

If it was desired to fire enamels that mature at 520°C to 550°C you could put the sheets in vertical racks to allow the heat to get to both sides equally as Jeff Zimmer does.  But this will only work for very low temperatures and for quick firings, otherwise the glass will begin to bend.

There are limits to this strategy of getting upper and lower surfaces to the same temperature, both in terms of physics and practicality.  There are temperatures below which no amount of slow heat input will have a practical effect, for example,  due to the brittle nature of the glass.  Even where it is possible, it can take too long to be practical.  For example, I can bend float glass at 590°C in 20 minutes into a 1/3 cylinder.  I could also bend it at 550°C (just 10°C above the annealing point), but it would take more than 10 hours – not practical.