Glass Bending Temperatures

Glass bending is the process by which glass is shaped without obtaining mould marks on the glass.  It also attempts to shape the glass without changing the thickness of the glass across its length and width. Glass bending can be done as a free drop curve or into a mould. This bending is usually done at much lower temperature than slumping.

Determining the temperature at which glass should be bent is a matter of experimentation with each new shape and thickness of glass.

If the temperature is too high you find distortions are created in the glass.  Sometimes wrinkles develop.  In general, a high temperature leaves a lack of time to compress and stretch evenly into irregular shapes.

If the temperature is too low the whole process takes an impractically long time to complete.

The just right temperature is in the region of 50C above the annealing point of the glass being used.  Experimentation with the shape and thickness of the glass is needed to establish a reasonable time for the bending; and for it to be achieved at a low enough temperature to get the shape required.

An example is this tapered cylinder.

 

Lantern frame for the glass

Mould shaped from the lantern into which the glass is to be bent

Flat template for cutting the glass

The bent glass

The curve was achieved at 590C in 20mins

A 1/8 sphere requiring bending in two directions was achieved at 570 in 45 mins to avoid ripples at edges.

The span as well as the shape affects the temperatures and times.  More information on bending glass is given in this blog entry.

Low temperature breaks in flat pieces

The usual advice in looking at the reasons for breaks in your pieces must be considered in relation to the process being used.  Breaks during low temperature processes need to be considered differently to those occurring during fusing.  

The advice for diagnosing breaks normally, is that if the edges are sharp, the break occurred on the way down in temperature. Therefore, the glass must have an annealing fracture or a compatibility break.  It continues to say if the edges are rounded it occurred on the heat up, as it broke while brittle and then rounded with the additional heat.

This is true, but only on rounded tack and fused pieces.

I exclude low temperature tack fuses from the general description of when breaks occur in flat pieces as it is not applicable at low temperatures.  

Low temperature flat work includes sintering, laminating, sharp profile tack fusing, etc.  There are lots of other names used for this "fuse to stick" work.  In all these cases, the finished glass edge will be barely different than when placed in the kiln.  It stands to reason therefore that you cannot know when the break occurred, as the edge will be sharp whether it broke on the way up or the way down.  

Periodic observation during the firing is the only way to be sure when the break occurred. These observations should coincide with the move from the brittle to the plastic stage of the glass.  Therefore, about 540C.  It can be at a bit lower temperature, but not a lot.  If the glass was not broken by that time, you can be fairly certain it broke on the way down.

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

Devitrification

What is it? When does it happen? Why does it happen? These are frequent questions.

Dr. Jane Cook states that devitrification is not a category (noun), but a verb that describes a process. Glass wants to go toward devitrification; a movement toward crystallisation.*

Mild devitrification is the beginning of crystallisation on the surface of the glass. It can look like a dirty film over the whole piece or dirty patches. At its worst, the corners begin to turn up or a crackling can appear on the granular surface.  This is distinct from the effects from an unstable glass or the crizzling as in a ceramic glaze. Divitrification can occur within the glass, but normally is a surface effect.

Differences in the surface of glass promotes precipitation of the crystal formation of silica molecules.  This fact means that two defences against the formation of crystals are smooth and clean surfaces. There are other factors at play also.  The composition of the glass has an effect on the probability of devitrification.  Opaque glass, lime, opalising agents, and certain colouring agents can create microcrystalline areas to "seed" the devitrification process.  One part of the composition of glass that resists devitrification is the inclusion of boron.

Devitrification generally occurs in the range of approximately 700°C – 840°C, depending to some extent on the type of glass.  This means that you need to cool the project as quickly as possible from the working (or top) temperature to the annealing point, which is, of course significantly below this range.

There is evidence to show that devitrification can occur on the heat up by spending too long in this devitrification range, and that it will be retained in the cooling. Normally this is not a problem as the practice in kilnforming is for a quick advance on the heat up through this range.  The quick advance does not (and should not for a variety of reasons) need to be as fast as possible.  A rate of 300°C per hour will be sufficient, as time is required for devitrification to occur.

The devitrification seen in typical studio practice results more often from inadequately cleaned glass than from excessive time at a particular temperature, up or down through the devitrification range.  

It is often seen as a result of grinding to fit shapes.  Even though the ground surface is cleaned, it may still be so rough as to promote devitrification.  The surface must be prepared for fusing by grinding to at least 400 grit (600 is better).  Alternatively, use fine frit of the same colour as the darkest glass to fill the gaps. This normally is applied in the kiln, so the pieces are not disturbed.

Dr. Cook suggests three approaches to devitrification:*
Resistance through:
 - Schedules
 - Flux
Dealing with it:
 - Cold work
 - Acids
Embrace it:
 - Allow it
 - Use it

Temperature range for devitrification
Homemade devitrification solution
Frit to fill gaps


* From a lecture given by Dr. Jane Cook at the 2017 BECON

[entry revised 25.9.19]

Fibre Dams

Fibre dams are a good and relatively inexpensive refractory material to form dams around regular and especially irregular shapes.  You need only cut the shape you want from the fibre board, if it is not a shape with straight lines.  

You can fire without any kiln wash or hardening if it is a one-off use.  For shapes you want to keep, you can harden the fibre board. 

Once hardened with colloidal silica, you need to paint the board with a separator – kiln wash, boron nitride or similar.

There are some precautions in the use of fibre paper and board.  The main physical one is that refractory fibre is lighter than glass and so will float on top of “molten” glass – that is fusing compatible glass higher than about 800°C.

Fibre board dams can be weighted with kiln furniture on the surface of the board.  If the board is flat this can be on the surface.  If the board is vertical, weights can be placed at the corners.

In the absence of fibre board, you can use layers of fibre paper.  If you have 6mm fibre paper, you need only one layer for two-layer glass, but remember that to get a bullnosed edge to the glass without needling, the fibre paper should be 3mm less than the final height of the fired piece. Thicker glass will require more than one layer of fibre paper.  Place as many layers of fibre paper as required to be at least equal in height to the finished piece on top of one another.  Push “U” shaped pins into the layers of paper to fasten the layers together.  Then cut the required shape out of all the layers all at one time. 

When finished cutting the shape out, you may want to line the edge with 1mm fibre paper to keep any of the layers of fibre paper showing through.  This dam will not need any kiln wash to prevent the glass sticking to it, unless you want multiple uses and so need to rigidise it with colloidal silica.

You can weight this fibre paper dam down by placing kiln furniture near the edge, all around the shape just as for the fibre board.

Safety in use of refractory fibre is described in Gregorie Glass.

Scroll down to Dusts/Particulates for safety recommendations.

How Close to the Edge

“How close to the edge of my shelf can I place a large piece?”

It depends in one sense how thick the piece is.  A 6mm piece that maintains the same footprint after firing as before, does expand beyond that footprint by about half a centimetre during the firing, so it would be safe to have a full centimetre space to the edge.  Thicker pieces will need more space – 9mm will need about two centimetres to accommodate the expansion at the top temperature. 

But

The real answer to this question is: When you know the heat characteristics across your shelf, you will know how close you can go to the edge for a relatively large piece. 

This Bullseye Tech Note number 1 tells you how to test the variations of temperature across your kiln. - http://www.bullseyeglass.com/methods-ideas/technotes-1-knowing-your-kiln.html

The objective in cooling glass is to have less than a 5C difference in temperature over the whole of the glass piece – top to bottom, and side to side.

If you have greater differences in temperature than that at the edges of your kiln shelf, you need to avoid placing large pieces in the danger area. Small pieces will not suffer by being close to the edges as their temperature differentials will be small.

I have found that the temperature differential in one of my kilns is great enough at the edges that I cannot have the edge of a relatively large piece of glass nearer than 50mm (2") from the edge.

Controlled cooling

It is sometimes stated that you can simply turn the kiln off below 370C and let the kiln’s natural rate of cooling take over the cool down.

This works for most flat 6mm pieces in most kilns, but as you work thicker or with greater contrasts in thickness, lots of tack fused elements or in a small rapidly cooling kiln, you do need to control the cooling toward room temperature.

The first thing you need to know is the natural cooling rate of your kiln.  

The rate of cool is not just about the annealing soak. The soak at annealing temperature is to equalise the temperature throughout the blass to have a differential of not more than 5C. 

The rate of cool is about avoiding thermal shock, too. The glass needs to maintain the temperature variation to less than 5 degrees Celsius difference throughout the glass as it cools.  This requires a slow controlled cool.  

You may program a cool of 100C to 370C thinking that the kiln will maintain that rate or less.  If the natural cooling rate of your kiln at 370C is 200C/hour, you risk thermal shock due to the rapid increase in the cooling rate.

You really do need to know the natural cooling rate of the kiln from the point you turn the programmer off to room temperature to be safe from thermal shock.

The alternative to turning off at 370C is to program the schedule all the way to room temperature.  The kiln will use no energy unless the kiln cools too quickly on its own.  At which point the program will kick in to slow the cooling of the kiln.

Finding Your Kiln’s Natural Cooling Rate

You need to observe how your kiln behaves while cooling without any power to be sure when you can safely &turn it off and let it cool without power.

Assuming you have programmed your kiln for a shut off at 370C, you need to observe every quarter hour or so to record both time and temperature.  From those observations you can calculate the cooling rate at the various temperatures.

Say at 6:00 your kiln was at 370C;

At 6:15 it was at 310C;

At 6:30 it was at 265C;

At 6:45 it was at 230C;

At 7:00 it was at 200C;

At 7:30 (you missed the quarter hour) it was at 160C;

At 8:00 it was at 140C;

At 9:00 it was at 125C;

At 10:30 it was at 110C.

To calculate the rate, you divide the temperature difference by the proportion of an hour between observations, as demonstrated in the following table.

Kiln Name/Description
Size
Shelf composition
Amount of glass
Observations
	Time	Temperature	minutes	Proportion	temperature	Rate of
1st	06:00:00	370	difference	of an hr	difference	cooling
2nd	06:15:00	310	15	0.25	60	240
3rd	06:30:00	265	15	0.25	45	180
4th	06:45:00	230	15	0.25	35	140
5th	07:00:00	200	15	0.25	30	120
6th	07:30:00	160	30	0.50	40	80
7th	08:00:00	140	30	0.50	20	40
8th	09:00:00	125	60	1.00	15	15
9th	10:30:00	110	90	1.50	15	10

Although this is an example, it shows how the cooling rate slows down as the kiln cools. 

If you were cooling a flat piece 12mm thick, you might get away with turning the kiln off at 370C, as a flat piece can cool as quickly as 300C/hr.

If you were cooling a piece 19mm thick, the natural cooling rate of the above kiln is too fast. 19 mm thick pieces need a cooling rate of 150C/hr, so according to the figures above you need to programme this kiln down to 230C to get the appropriate final cooling rate.

If it is a tack fused piece with a 6mm base and areas of two layers of tack fusing, you should fire as though it is 24mm thick.  In this case, the final cooling rate needs to be 90C/hr.  For the kiln in the example above, that rate is not achieved until below 160C, so that is the minimum temperature for switch off.

This method can be used for any temperature range.  For example, you may want to know the rate of cooling from the top temperature to the annealing temperature.  This method will work there too. You may want to record the temperatures more frequently than every quarter of an hour though.

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

You really need to know your kiln’s natural cooling rate before you can be confident of switching the kiln off at 370C.  This blog shows a method of determining the natural rate of cooling. 

Bubble Mystery

A question was asked about a collapsed bubble. There were two pieces in the kiln and one (strips) was fine and the other (flat plate) had the collapsed bubble.  Both on the same dried shelf.  The question also asked if the collapsed bubble piece could be flattened by fusing again.

Collapsed Bubble

The bubble collapsed because it had not burst by the time the cool toward annealing had begun.  As the air pressure under the bubble dropped, and the weight of the thinned glass bubble sank down as there was not enough air pressure to hold it up.

The glass is now thinner at the centre of the bubble than the main part of the piece, and thicker at the edge of the bubble. I don't think it is possible to successfully flatten it to become an even thickness across the whole piece. To get the same thickness across the whole piece would require high temperatures and long soaks there. 

Another possibility is to use a pressing solution. 

My suggestion is to add elements or repurpose it. I don’t think any repairs would present a good-looking piece.

Diagnosis

The on-line diagnosis of the possibilities for the cause of the bubble was extensive and sometimes inventive.  It was finally determined the bubble was from under the glass, that is, between the glass and the shelf. A slight depression in the shelf is the usual explanation.  The user tested the shelf for smoothness and found no depressions.

It was clear the bubble came from under the glass.  All the suggestions about how bubbles can form under glass were given, but none seemed to apply.

How can you get a bubble on a dry shelf that is perfectly flat and that has not been subjected to too rapid or too high a temperature?

Solution

The answer is that a little spot of grit or tiny ball of fibre paper can keep the glass raised up enough for air to be trapped.

Prevention

It is not enough to test the shelf is flat.  You need to use clean kiln wash with a clean brush to avoid any grit being brought to the shelf. It is also a good reason to vacuum the shelf before each use in case any dust or grit has fallen onto the shelf. Covering the shelf or putting it into a cupboard will also reduce the possibility of small bits of grit falling onto the shelf.

Of course, if you smooth the kiln wash with a nylon or similar fine cloth, you will remove any specks of grit.  A vacuum of the shelf after smoothing is still a good idea.

Conclusion

It is as important to keep tools and materials clean as it is to clean the glass you are going to kilnform.

Simultaneous Firing of Different Moulds

Often you have moulds of different sizes or depths that you would like to fire at the same time to use the space or save time.  If the moulds are of distinctly different sizes or shapes, you will not save time, as the likely outcome is that some will be over-done and un-shapely or, conversely, that some will not have completed their slump.

The main things that act against firing moulds with distinctly different firing requirements are:

·        Moulds with different spans require different temperatures or different soak lengths.

·       Moulds of different depths, even if they have the same span, require different soak lengths.  

·        Moulds of different shapes, even if they are the same depth, require different soaks or different temperatures. 

As an example, if you have two moulds that require less time or lower temperature than three smaller ones. If you get the smaller, relatively deeper ones fully slumped, the larger, shallower ones will be more marked by the mould than necessary.

The best thing you can do if you want to make full use of the kiln space each time you fire, is to save up the glass until you have enough to put in a full kiln load.  This may require more moulds of the same size than you currently have.

Usually trying to fit in a lot of slumping into one firing relates to a concern on how much electricity will be used in multiple firings. However, the kiln does not use huge amounts of electricity.  A 50cm square kiln will normally use less than 10Kwh for a slump with a long soak.  This will cost much less than a glass of beer or wine.

Specific Gravity

This is an important concept in calculating the amount of glass needed to fill a pot melt, and in glass casting.  This will also help in the calculation of the amount of glass required to fill a given area to a defined thickness.

Specific gravity is the relative weight of a substance compared to water. For example, a cubic centimetre of water weighs 1 gram. A cubic centimetre of soda lime glass (includes most window and art glass) weighs approximately 2.5 grams. Therefore, the specific gravity of these types of glass is 2.5.  

If you use the imperial system of measurement the calculations are more difficult, so converting to cubic centimetres and grams makes the calculations easier. You can convert the results back to imperial weights at the end of the process if that is easier for you to deal with.

Irregular shapes

Water fill method
Specific gravity is a very useful concept for glass casting to determine how much glass is needed to fill an irregularly shaped mould. If the mould holds 100 grams of water then it will require 100 grams times the specific gravity of glass which equals 250 grams of glass to fill the mould.

Dry fill method
If filling the mould with water isn't practical (many moulds will absorb the water) then any material for which the specific gravity is known can be used. It should not contain a lot of air, meaning fine grains are required. You weigh the result and divide that by the difference of the specific gravity of the material divided by 2.5 (the specific gravity of soda lime glass). 

This means that if the s.g. of the mould filling material is 3.5, you divide that by 2.5 resulting in a relation of 1.4   Use this number to divide the weight of the fill to get the amount of glass required to fill the mould.   If the specific gravity of the filler is less than water, then the same process is applied.  if the specific gravity of the filler is 2, divide that by 2.5 and use the resulting 0.8 to divide the weight of the filler.  This only works in metric measurements.

Alternatively, when using the dry fill method, you can carefully measure the volume of the material.  Be careful to avoid compacting the dry material as that will reduce the volume.  Measure the volume in cubic centimetres.  Multiply the cc by the specific gravity of 2.5 for fusing glasses.  This will give the weight in grams required to fill the mould.  If you compact the measured material, you will underfill the mould. The smaller volume gives a calculation for less weight.

Regular shapes

If you want to determine how much glass is required for a circle or rectangle, use measurements in centimetres.  

Rectangles
An example is a square of 20cm.  Find the area (20*20 =) 400 square cm. If you want the final piece to be 6mm thick, multiply 400 by 0.6cm to get 240 cubic centimetres, which is the same as 240 grams. Multiply this weight by 2.5 to get 600gms required to fill the area to a depth of 6mm.

Circles
For circles you find the area by multiplying the radius times itself, giving you the radius squared.  You multiply this by the constant 3.14 to give you the area.  The depth in centimetres times the area times the specific gravity gives you the weight of glass needed.

The formula is radius squared times 3.14 times depth times specific gravity.   R*R*3.14*Depth*2.5
E.g. 25cm diameter circle:
Radius: 12.5, radius squared = 156.25 
Area: 156.25 * 3.14 = 490.625 square cm.
Volume: 490.625 * 0.6 cm deep =294.375 cubic cm.
Weight: 294.375* 2.5 (s.g.) = 735.9375 gms of glass required.  
You can round this up to 740 gms for ease of weighing the glass.

Firing uneven layers

Firing uneven layers requires more care than a piece equally thick all over.

My rule of thumb is to add the difference between the thick and thin to the thick and fire for that. So, a piece with a 6mm base and a total height of 12mm gives a difference of 6mm added to 12mm gives a firing thickness of 18mm. If you look up the bullseye site annealing for thick slabs, follow the schedule for 19mm. The initial heating rate can usually be half the final cooling rate shown in the table.

The Bullseye recommendations are more conservative.  They recommend that the firing rate should be for something twice the thickest part of the piece.  In this case, the firing would be for a piece as though it were 24mm. Again, the initial rate of advance would be equal to half the final cool segment in the Bullseye table Annealing Thick Slabs.  

If you are slumping a thick piece, you can use the initial rate of advance all the way to the slumping temperature and then anneal according to the thick slabs table.

Terminology for degrees of fusing

Can anyone describe what a contour fuse is?

No one can satisfactorily describe, to a high level of acceptance, what a contour fuse is. For me it is just before a full fuse. That will not be acceptable for many, just as describing something as a rounded tack fuse is not a contour fuse for me.  A sharp-edged tack fuse is sintered glass. This will be important to observe as you move to other glass processes such as pate de verre.

There is not yet an accepted terminology and will not be as long as people choose to invent new descriptions for what are essentially the same things.

The closest you can get to a sensible range of descriptors is in the Bullseye document "heat and glass" where the temperature ranges are the important constants.

  

The fourth column of this document gives names for the process. It would be a good idea to adopt these terms, as Bullseye is the company doing the research in the area of kilnforming.

Bullseye terminology gives the following:

A slump or bend occurs in the 540C – 670C range

Fire polishing and sintering occur in the 670C – 730C range

Tack fusing (a rounding of edges) occurs in the 730C – 760C range

A rounded tack fusing that begins to sink into the base glass occurs in the lower end of the 760C – 816C range.

Contour fusing occurs in the middle of the 760C – 816C range.

Full fusing (flat) occurs at the upper portion of the 760C – 816C range.

Getting the Right Firing Temperature

“what temperature should I use to get a tack fuse that is just less than a contour fuse?”

This is the kind of question that appears on the internet often.  Unfortunately, no one can answer the question accurately, because it depends on some interrelated variables.

Kiln characteristics

Top or side elements, size of kiln, relative size of piece, all have an effect. Also no two kilns even of the same model have exactly the same characteristics.

Ramp Rate 

How quickly or slowly you fire has a big effect on the temperature and soak needed to achieve the desired result. This is the effect of heat work.

Temperature

There are no absolute temperatures for a given effect, given the above two variables.

Soaks

The length of time and the number of soaks will affect the temperature required to achieve your effect.

OK. So, what can I do?

Observation

The only certain way to get the effect you want is to observe.

Set a schedule, guessing the top temperature and length of soak.  Know your controller well enough that you can extend the soak or end the segment by advancing to the next.  Your manual will tell you how to do this.

Peek at intervals from 10-15C below the selected target temperature. Peek at 5min intervals until the effect is achieved.  Advance to the next (cooling) segment.  Record the temperature and length of soak at which the effect was achieved.  On subsequent firings you can experiment with reducing the temperature by 5C – 10C with a 10-minute soak.  Observe and record the temperature and effect as before.

The reason for going for a 10-minute soak rather than longer is to avoid holding at the target temperature for a long time, as that can help induce devitrification.  The reason for a soak at all is to achieve the minimum of marking on the reverse or picking up kiln wash or kiln paper on the back.

If effect is not achieved by the end of the soak, extend it by using the appropriate key or combination of keys.  Keep observing at five-minute intervals until the effect is achieved.  Advance to the next segment and record both the temperature and time.  The objective is to get the heat work done with a 10-minute soak, so you will need to increase the temperature on the next firing.  The amount of increase will depend on the length of soak required to get the desired surface on the previous firing.  The longer the soak, the more temperature you need to add.  You will need to repeat the observations and recording until you find a temperature that will achieve the effect with a 10-minute soak.

Use the lessons from the observations to lower temperature, extend soak, raise temperature, reduce ramp speed, or reduce soak as required.  It will also help you judge on other pieces the approximate temperature and time required for the new layups or new moulds.