Longer Soak or Higher Temperature?

 ‘Is it better to extend the soak or add more firing time when the firing program isn’t quite enough? What are the meanings of “soak,” “hold,” “ramp,” “working temperature” and “top temperature”?’  

Let’s start with some of the terms.

“Soak” and “hold” have the same meaning in scheduling.  Schedules are made up of a series of linked segments.  Each segment contains a rate, temperature, and time.  The time is often called a “hold” in the schedules.  That time can have several effects.  It can allow enough time for a process, such as slumping, to be accomplished.

Although “soak” is entered into the schedules in the same way as a hold, it has a different concept behind it.  The hold when used as a soak allows the set temperature to permeate the whole thickness of the glass.  An example is in annealing. An annealing hold/soak is set.  This is to allow the glass to become the same temperature throughout. 

The ramp is the rate at which the controller is set to increase/decrease the temperature.  This is normally the first element in the segment.

“Top” and “Working” temperature are the same thing.  It is the temperature at which the desired effect is achieved.  They have slightly different nuances.  Top temperature is normally considered as a point where the desired profile will be achieved in a few minutes.  The working temperature is also that, but includes the idea that it will take time for the effect to be achieved.

Which should you alter first – soak time or temperature?

Most important is that you alter only one at a time.  If you alter the two elements at the same time, you do not know which was the cause of the result.

In general, you lengthen the soak if the effect is not achieved at the temperature and in the time set.  There are two reasons for this.  Glass has fewer problems at lower temperatures.  Secondly, the controllers are set up in such a way that it is easy to extend the time. Check your manual for the key sequence to extend the time.  It is more difficult to alter the temperature during a firing. 

To determine if you need more time, you peek into the kiln as the kiln approaches the top temperature.  If the profile has not been achieved by the time set at your working temperature, you enter the combination of keys to keep the kiln at the top temperature until you see the effect you want.  Then enter the combination of keys to skip to the next segment.

Whether you alter time or temperature, depends on what you are doing.  Soak plus temperature equal heat work.  With heat work you can accomplish the same effect at lower temperatures.  It may be that taking more time (usually slower ramp rates) to get to the same or lower temperature, will give the results desired.

For slumping, draping and other low temperature processes extending the hold/soak is appropriate. It reduces the amount of marking that is created by the mould or surface supporting the glass.

When tack, contour, or full fusing, you should be aiming to finish the work in about 10 minutes. Soaking/holding significantly longer increases the risk of devitrification.

For high temperature processes such as pot and screen melts and some flows, increasing the temperature is probably the right thing to do, to avoid the devitrification possibilities of long holds of open face high temperature work.

These can only be guidelines.  Your instincts and experience will help you determine which is the right thing to do in the circumstances.

 

Draping Different Thicknesses and Sizes

Scheduling for different sizes and thicknesses of drapes requires schedules specific to these factors in addition to observing the progress of the drape.

Bob Leatherbarrow's research shows 6mm drapes more slowly than 3mm. It seems the thicker glass takes longer to begin the slump.  Glass behaves in a similar way for a drape.  My experience of draping 6mm/0.25” and 3mm/0.125” in the same firing confirms that 6mm takes longer.  Or, it needs a higher temperature.  I know this goes against common sense, but tests and experience show it to be true.

If you try to drape 3mm/0.125” and 6mm/0.25” pieces at the same time, the 3mm will reach the desired shape before the 6mm. You then have the choice of an under draped 6mm piece or an over draped 3mm piece.  This indicates that draping different thicknesses in the same firing will be unsuccessful.  To a lesser extent, the size of the drape will influence the speed of the drop.  So, you are unlikely to achieve completely desirable results with significantly different sizes of drape in the same firing either.

Observation is essential in all draping operations. You cannot know how long it will take for a piece to drape or drop to your requirements.  To be sure of your result you need to observe the progress of the drape.  There is rarely a safety net of a form to drape onto as in slumping.  To observe, set your top temperature with a long soak/hold.  Start peeking at frequent intervals from the time top temperature is reached.  

When the glass has reached the desired shape, advance to the next segment.  Your controller manual will give you instructions on how to do that.

Different thickness and sizes of glass require different firing conditions.

Frit Additions to Shaped Pieces

It is possible to stick frit to slumped pieces. But soaking for a long time – several hours - at 650°C/1200°F is required to stick the frit.  The added pieces will remain relatively sharp. You need to observe frequently from 600°C/1111°F to make sure that the form of the glass is not distorting. 

Credit: Pyramid Gallery, Smyth and Zebrak

Although it is possible, adding pieces to already shaped objects is not best practice, nor will it frequently give satisfactory results.  If the slump is shallow, it is more possible to do this successfully than steep or highly shaped forms.  But the most suitable practice is to flatten the piece, then tack fuse the pieces onto it. Follow this fusing with the new slump or drape.  This flattening process will not be possible with all shapes. 

The best results will be achieved by accepting what you have and make a new piece with the planned additions from the start.

This process will not be suitable for draped glass as the glass will drape further during this low temperature soak.

 

I've a book that gives more detail. Low Temperature Kilnforming, an evidence based approach to scheduling  or at Bullseye

Square Drapes

Two pyramidical moulds. One stepped and the other smooth.

 This kind of draping mould with flat sides will never work very well as a draping mould.  The draping sides have to compress. This takes a long time and is likely to cause folds in the glass.

 The common experience is that two opposite sides drape first and conform to the mould. This displaces the compression necessity to the other two sides. This "taco" style initial drape is common in all drapes. It is usually observed in handkerchief drapes.  In the early stage of draping two sides of the glass fall, creating a taco shape. With continued heating, those long sides fall and spread the initial draped sides to become almost equal. 

 This taco formation also occurs on the pyramid style mould, giving two flat sides.  The glass on the other sides then fall. As the glass area is now larger on these sides than the mould area, a drape or fold is formed.  Imagine the drapes a square piece of cloth place on a pyramid would create. The cloth has more area than the sides of the pyramid.  The excess cloth creates folds at each corner.  The same happens with the glass.

 This draping fold can be minimised by using low temperatures and long (multiples of hours) soaks.  This allows all the sides of the glass to begin forming at more or less the same time.  I am not sure the folds can ever be completely eliminated.  With extremely long soaks, the drapes will flatten to the rest of the glass. 

 Annealing difficulties are caused by this folding.  It will create thick overlaps.  This in turn will cause the annealing difficulties. There are areas that are much thicker than others.  If you started with 6mm glass, the folds will create areas that are 18mm thick. 

 Making sure this glass - with such large differences - is all of the same temperature will require long annealing soaks.  It will also require very slow cooling segments.

 Square drape moulds are rarely successful. Folds are created at the corners, rather than fully conforming to the mould.

Ceramic Drape Moulds

Characteristics of materials

One of elements you need to consider in selecting a mould for draping is the characteristics of the ceramic material in relation to the glass being draped.  

Ceramic

Ceramic materials have what are called inversions. These are  points at which the ceramic has a quick change in expansion both on the heat up and cool down.  The two major ones are cristobalite inversion temperature at around 225°C and the quartz inversion at about 570°C.  The Crystobalite inversion is a sudden change of about 2.5% and the quartz is a sudden change of 1%.  These are very sudden and dramatic changes in comparison to the average of around 0.1% over the temperature range of 570°C to 800°C.  The crystobalite inversion does not occur until ca. 225°C.  This means that the whole structure of the ceramic is contracting less than the glass – exhibiting a CoE of ca. 66 rather 90 to 96.

Ceramic drape mould from Creative Glass Guild

Glass

We are used to saying glass expands and contracts at a standard rate, depending on the glass this may be a CoE of 83 to one of 104.  This is not the case.  The coefficient is an average calculated between 20°C and 300°C.  If you change the temperature range, the coefficient will also change.  And if you look at the range 570°C to 580°C you find the CoE is around 500.  This means that as the glass cools into the annealing range, it is contracting about 7 times faster than the ceramic. 

This dramatic difference in contraction means that the glass is attempting to crush the ceramic by enclosing it tightly.  Sometimes it does it so strongly that the strength of the glass is exceeded, and it breaks.

Possibilities

It is possible to drape over ceramic in certain conditions.

Influence of draft

The term “draft” indicates the slope of the sides of the form.  The steeper the sides, the more likely the glass is to trap the ceramic mould.  To be useful, the draft of the mould needs to be sufficient for the glass to slide upwards on the mould as it cools. This means the mould needs smooth sides and be well covered with a separator.

Compensations

You can compensate for steep drafts by wrapping the ceramic form in 3mm refractory fibre paper.  You can bind this with high temperature wire to ensure it stays throughout the firing. The fibre paper can be compressed and so provides a cushion between the rapidly contracting glass and the slowly contracting ceramic.

These need a circle of 3mm fibre paper over the open top of the kiln posts that have no draft at all before use. Of course, they need to have a circular piece of fibre paper over the hole in the post.

The use of ceramic forms to drape over requires care about the draft of the ceramic or addition of a cushion to avoid the greater contraction of the glass than the ceramic grabbing the mould so tightly it cannot be removed.

Frosting on slumped glass

 [We’re] Having a few challenges with a stainless-steel S-curve mould (15cm x 10cm [prepared with boron nitride]. … When we slump a piece of glass, we get a frosted effect … in places where the glass was touching the mould at the beginning. I don't think it's devitrification, because the glass itself isn't cloudy, it's just hundreds of little bumps and dimples.

Photo credit: Adrian Cresswell

 This is a mould that combines draping and slumping in the same firing. The glass must drape over the hump and slump into the valley at the same time. This is effectively two processes in the same firing. It does require some compromise in scheduling as a result.

 The evidence presented shows boron nitride – a slippery surface – was used to prepare the mould. In another firing Thinfire – a powdered surface – was used as a separator. Both created this marking on the back. The schedule was not presented.

 This indicates something other than the separator is creating the problem. Note that the marking also occurs at the extreme right end where the glass would be resting on the lip of the mould. The marking does not occur where the glass is slumping down into the curve. It only occurs where the glass is draping.

 As suggested, this is not devitrification. That occurs on the surface rather than on the bottom. This further indicates the difficulty is between the glass and the mould.

 I suggest the marks are from the glass sliding along the mould. These are frequently called stretch marks. The glass is sliding and stretching along the mould. This blog post contains much more information.

 It is of course possible that insufficient boron nitride was placed on the steel and the glass grabbed the steel. It is worth checking, although I don’t think it likely.

 You might think the Thinfire covering of the steel would make everything smooth. However, Thinfire turns to powder and fibreglass particles after about 400°C/753°F. These particles are drawn along as the glass moves against the mould. The particles can bunch and remain as bumps on the surface of the mould. This may account for the rougher surface with Thinfire than boron nitride.

A summary

These stretch marks occur when the glass moves excessively against the mould. This is usually a combination of high temperature and fast ramp rates. Slumping should be done at the lowest practical temperature. The soak should be long, rather than brief.

The Remedy

 Fire more slowly and to a lower temperature. The Bullseye suggestion from their quick tips is for a double curve (or wave) mould of 250 x 210 x 40 mm (9.85" x 8.25" x 1.6"). They suggest a ramp rate of 167°C to 660°C /1221°F with a soak of 10 minutes for a 6mm/0.25” thick piece.

 In my experience this is too fast. Slumping into this mould can be done at 630°C/1167°F with a 30-minute soak in my kilns using the same ramp rate.

 This is a simple mould to slump and drape into. It is essentially two partial cylinders pushed together in opposite directions. The curves are gentle and progressive. There are no sharp changes of direction. These factors mean that the slow and low approach will work well.

 However, in this case the curves will be even tighter as the full length is 150mm x 100mm/ 6” x 4”. The height is not given, but for a self-standing piece, it likely to be a minimum of 40mm/1.6". This makes for a tight curve on the mould. It is likely the glass will slide more than on a gentle curve. My thought is that the steel mould has been produced without knowledge - or testing - of the practicalities of getting the glass to bend to such a small radius. This means that I would be trying a schedule of about 125°C/225°F per hour to a top temperature of about 630°C/1167°F with a 1.0 to a 1.5-hour soak. The glass is going slump much more slowly with this smaller span.

 With gentle heating -slow ramp rate, long soak - the glass gradually conforms to the shape of the mould without stretching over the hump/crest of the mould. Instead, what happens is that the glass slips slightly from the opposite end of the mould. To counteract this, I place the glass 6mm/0.25” over the upraised end of the ceramic mould. This then finishes just inside the mould’s edge.

 With a steel mould, this is not possible without the glass hanging up on the hot and sharp edge. The glass will need to be at the edge or just inside to prevent hanging up on the end of the mould. The glass will slip down the mould a little, but not so much as to cause problems. It is possible to prop a piece of fibre board at the upturned end of the wave mould to support the overhanging glass if the full curve is required.

 The glass on the high part of the mould will not stretch at the low temperature, but gradually conform to the shape of the mould.

 I have much more information on this and other things in the eBook:

Low Temperature Kilnforming; an Evidence-Based Approach to Scheduling

 

Candle shades

These kinds of drapes are generically known as handkerchief drops, as they form the kind of shape that is formed by holding the cloth in the middle and letting it drape.  They can be done as small drapes over kiln posts, cocktail shakers, and much larger forms.

Two heights of new cocktail shakers

A well used cocktail shaker with kiln wash

A kiln post wrapped in preparation for firing

Two short kiln posts after firing

When preparing several drapes to be fired at one time you need to consider several factors.

Higher in the kiln is hotter.

The heat in a kiln, as in an oven, is greater the higher in the kiln is supported.  This means that taller supports will drape quicker than shorter ones. The consequence is that all the drapes should be of the same height.

A single layer that has begun to stretch at the shoulder of the former 

Larger spans fall quicker than smaller.

The more of the glass that is unsupported, the quicker it will fall, even at the same height. This is because the larger amount of unsupported glass has more mass than a smaller one and so falls quicker.  Plan for all the glass to be of similar sizes.

These two were fired at the same time. The back one is larger than the front 

Different shapes fall in different ways.

Squares and circles are the most common shapes used in a  drape. The corners of squares are points that are further away from the centre of the support than the sides.  These points begin to fall first, drawing the sides in later in the firing.  Circles form a taco shape before the ends of the “taco” begin to fall.  This deformation of the circular “taco” takes longer than a square takes.

Care needs to be taken that the glass does not thin excessively at the shoulder of the support.  There is less difficulty, if the same shapes are fired together as different heat work is required for each shape.

Observation by peeking is required to stop free drops at the right time.

As in all drapes, it is important to observe the progress of the drape at intervals.  This is best done by quick peeks to note the development of the shape and to move to the cooling segment when the drape is complete. This also requires a scheduling of a long soak and knowledge of how to advance the kiln controller to the next segment of the schedule.

Using Ceramic to Drape

Characteristics

Before choosing a ceramic shape to use in draping of glass, you need to consider the characteristics of the two materials.  This is one circumstance where CoE is actually useful. 

The expansion of the two materials is different. Soda lime glass typically has an expansion rate - in the 0°C to 300°C range - of 81 to 104.  Ceramic has an expansion rate - in the 0°C to 400°C range - of 30 to 64.  This is important in the final cooling of the project.  As the glass expands more than the ceramic on the heat-up, so it also contracts more during the cool.  This means that the glass will shrink enough to trap the ceramic or even break if the stress on the glass is too much. 

Shape

The shape of the ceramic form will have a big effect on the usability of it as a mould.  Ceramics with right angles between the flat surface and the sides will not be suitable for draping without modifications or cushioning.  The forms suitable for draping need to have a significant draft to work well.

Ceramic forms such as rectangles, cubes, and cylinders do not have any draft in their form.  

A cube shape unsuitable for draping

Ceramic cylinders with straight sides

Although rounded at the base, the sides are too straight to be a draping mould

The glass will contract around these forms until they are stuck to the ceramic or break from the force of the contraction around the ceramic.

You can experience this trapping effect in a stack of drinking glasses.  Sometimes one glass sticks inside another even though there is a slope (i.e., a draft) on the sides of the glasses. This happens mostly when you put a cold glass inside a warm one.  On cooling the warm glass contracts to trap the cooler one. You can separate these by running hot water on the bottom glass, so that it expands and releases the inner, now cool, one. 

Effect of Shape

The ceramic contracts at about half the rate the glass contracts (on average), unlike steel which contracts faster than the glass. This means steel contracts away from the glass, while the glass contracts against the ceramic, on the cooling.

Because the glass is in its brittle or solid phase during the last 300°C to 400°C, this contraction tightens the glass against the ceramic, causing stress in the glass, even to the point of breaking.

However, if you choose ceramic forms with significant draft, you can drape over ceramic.  This is possible when the slope is great enough and the form is coated with enough separator, to allow the glass to slip upwards as it contracts more than the form. Experience with different draft forms will give you a feel for the degree of slope required. 

 

These pyramid shapes have sufficient draft to allow the glass to move up the mould during cooling.

Compensation for Lack of Draft

You can compensate for the insufficient draft of ceramic forms by increasing the thickness of the separators for the form.  The hot glass will conform to the hot ceramic, so there needs to be a means of keeping the glass from compressing the form while cooling.  This can most easily be done by wrapping the form that has little or no draft with 3mm ceramic fibre paper.  It is possible to get by with as little as 1mm fibre paper, but I like the assurance of the thicker material.

Kiln posts wrapped in 3mm fibre paper and secured with copper wire

The fibre paper can be held to the form by thin wire wrapped around the outside of the fibre paper. The advantage of the 3mm fibre paper is that the wire will sink below the surface of the paper.  You can tie off the wire with a couple of twists.  Cut off the ends and push the twist flat to the fibre paper to keep the glass from catching onto the wire.  If you want further assurance, you can put a bit of kiln wash onto the wire.

Conclusion

The choice of ceramic shapes to drape glass over is very important.  It needs to have sufficient draft and separator to allow the glass to slip upwards as it contracts more than the ceramic during the cooling.  You often can use items with no draft if you wrap fibre paper around the sides of the form.

Draping over steep moulds

Draping over a narrow or small supporting ridge with large areas of glass is difficult.

One solution might be just to invert the whole piece and let the glass slide down into the mould. However, there rarely is enough height in a glass kiln for deep slumps, especially with a “V” shaped mould. It has to be high enough for the edges of the glass to be supported at its edges. You could also approach this by having a first mould with a shallower angle or broader support at its centre. Drape over this first, then use the steeper mould as the second draping mould. This makes the balance less critical.

The idea of supporting the glass is the key to doing this kind of slump that seems to require an impossible balancing act, if it is to be done in one go. Place kiln washed kiln furniture at the edges of the otherwise unsupported glass. Fire the kiln, but watch until the glass begins to slump. Then reach in with a wet stick and knock the kiln furniture aside to allow the glass to continue its slump and conform to the mould shape.

The lower temperature you use to do the draping and the slower your rate of increase is, the less the glass will be less marked by the mould. Frequent brief visual inspection during the drape is vital.

Also have a look at a suggestion for the kind of firing required for this here.

Schedules for Steep Drapes

I have been asked for a schedule for draping in the context of a tip on steep straight sided drapes.

What you are trying to do with a steep drape is two things. One is to compensate for the heat sink that the glass is supported by, and the second is to compensate for the relative lack of weight at the outer edge of the glass.


The supported glass transmits its heat to the support, leaving it colder than the unsupported glass. This often leads to breakage due to heat shock at much lower temperatures and slower rates of increase than glass supported at its edges. My experience has shown that - contrary to what I recommend for other kinds of firings - a slow rise with short soaks at intervals up to the working temperature works best. The reason for these slow rises and soaks is to try to get the support and the glass to be as nearly as possible at the same temperature throughout the rise in temperature. The soaks help ensure the mould is gaining heat without taking it from the glass.

The other problem with steep drapes is that the edges of the glass begin to drop more quickly than the area between the support and the edge. This leads to the development of an arc that touches the mould side near the bottom before the glass between the edge and the and the support. Extended soak times are required to allow the glass to stretch out and flatten. If this is done at high temperatures, the glass will thin - possibly to the extent of separating.

So the requirements for a firing schedule on this kind of drape are slow increases in temperature with soaks to avoid thermal shock, and an extended soak at the (low) forming temperature.

Whether using steel or ceramic moulds, I use a slow rise in temperature to 100C with a soak of 15 minutes. I then increase the rate of rise by 50% for the next 100C and give a 15 minute soak there. For the next 200C I raise the temperature at twice the original temperature rise, again with a 15 minute soak. The glass and mould should now be at 400C. This is still at the point where the glass could be heat shocked, so I only increase to 2.5 times the original rise rate but use this rate all the way to forming temperature.

Each kiln has its own characteristics, so giving schedules is problematic. 

•  A side fired kiln will need slower heat rises than a top fired one. 
• The closer the glass is to the elements, the slower the rate of increase needs to be. 
• The kind of energy input - electric or gas - has an effect. 
• The thickness of the glass is also a factor in considering what rate to use. 
•  The size of the glass in relation to the size of the support is important - the greater the differential, the slower the heat rise should be. 

So in making a suggestion on heat rises, it is only a starting point to think about what you are doing and why you are doing in this way.

I have usually done this kind of draping in top fired electric kilns where the elements are about 250mm above the shelf, and about 120mm apart. In the case of a 6mm thick piece about three times the size of the support area, I use 50C/hr as my starting point. This is one third of my usual rate of temperature rise. However you must watch to see what is happening, so that you can make adjustments. You should observe at each of the soaks, so you know how the glass is behaving. It will also help you to pinpoint the temperature range or rate of advance that may be leading to any breakages.

On steep slumps, the temptation is to use a high temperature to complete the drape. This is a mistake as the glass will be more heavily marked and tends toward excessive stretching and thinning. What you really need is a slow rate of advance to a relatively low temperature. If you normally slump at about 677C, then you want to do this steep, straight sided drape at 630C or less. It will need a long soak - maybe up to an hour. It will also need frequent observation to determine how the drape is progressing. So plan the time to make yourself available during this forming soak.

Annealing is done as normal, since the mould and glass are more closely together and will cool at the same rate.

The original tip on the set up of a steep straight sided slump is here.

The relative order of kiln forming events

When preparing for multiple firings of elements onto a prepared piece, you need to consider the order and temperatures of events so that you do not harm an earlier stage of the project.  This blog entry will not give definitive temperatures, as that varies by glass and by kiln.  Instead, it indicates what happens in progression from highest to lowest temperatures in approximate Celsius degrees.  

ca. 1300C  -  Approximate liquid temperature 

ca. 850 – 1000C  -  Glass blowing working temperature

ca. 950C  -  Raking and combing

ca. 850C  -  Casting

ca. 810C  -  Full fuse

ca. 790C  -  Large bubble formation

ca. 770C  -  High tack, low contour fuse

ca. 760C  -  Tack fuse

ca. 750C  -  Fire polish

ca. 700C – 760C  -  Devitrification range

ca. 700C  -  Lamination tack

ca. 600C – 680C  -  Slump and drape

ca. 650C  -  Vitreous paint curing temperature

ca. 600C  -  No risk of thermal shock above this temperature 

ca. 540 – 580C  -  Glass stainers enamel curing temperature

ca. 520 – 550C  -  Silver stain firing temperature

ca. 550C  -  Glass surface beginning to soften

Slow rates of advance needed from room temperature to ca. 500C

These temperatures are of course, affected by the soak times. The longer the soak time, the lower temperature required. The rate at which you achieve the temperature also affects the effective temperature.  Slower rates of advance require lower temperatures, than fast rises in temperature.  These illustrate the effect of heat work.

The table shows for example you need to do all the flat operations and firings before slumping or draping.  It also shows you can use vitreous glass paints at the same time as slumping and draping.  This emphasises that the standard practice is to plan the kind of firings you will need for the piece and do them in the order of highest temperature first, lowest last.

In general, you do need to do the highest temperature operation first and lowest last.  But there are some things you can do with heat work.  For example, if you needed to sandblast a tack fused piece, but did not want to risk reducing the differences in height there things you can do.  From the list above, you can see the glass surface begins to soften around 500C.  It is possible to soak the glass for a long time around 500C to give it a fire polish, instead of going to a much higher temperature.  You will need to experiment to find the right combination of temperature and soak length, but it can be done.

This article is to show that knowledge of what is happening to the glass at different temperatures, can help in “fooling” the glass into giving you the results you want without always following the “rules”.  This may also be what it is to be a maverick glass worker.  Use the behaviour of glass to your advantage.

Tin Bloom

Using float glass sometimes produces partial clouding as though devitrification were present. Although float glass is prone to devitrification, not all the cloudy film on the surface is due to devitrification.

Float glass, which these days, is almost all clear smooth glass, gets its name from the process of floating the glass on molten tin. The tin in compression gives an apparent devitrification effect which is called tin bloom.

it is different from devitrification, to which float glass is particularly subject. Devitrification sprays and solutions will not have an effect on this surface defect called tin bloom. 

When the tin layer is stretched, it does not create a tin bloom on the surface.  Therefore, it is important to have a means to detect which is the tin surface.  Always fire the glass with the tin in the same relative location to each other.  I.e., on several layers of glass have all the tin side down or all up, but not mixed. 

This example of a test by Glass Art by Margot shows the tin bloom on the outer portions of the platter where the tin side was up, causing the tin too be compressed and show.  The flatter portion of the piece did not show this tin bloom as there was not the same extent of compression. You can visit the description of the experiment here.


When forming the glass (slumping, draping, kiln carving) make sure the tin sides will be stretched rather than compressed.  Of course, you can take advantage of the tin bloom by controlling the compression of the tin layers.