Observations on Some Suggestions about Annealing

There are writings from a teacher attempting to make glass fusing simple.  Unfortunately, glass physics and chemistry are very complicated.  Attempting to avoid these complications leads to failures and other difficulties as the practitioner progresses. 

Proper annealing is one of the fundamentals to achieving sound kilnforming results.  Some suggestions have been made by a widely followed person to “simplify” the understanding of the annealing process.  Discussion of the meaning and importance of annealing can be found in many places, including here.  

Annealing temperatures

It has been suggested that the annealing temperatures can be inferred from the CoE of the glass that is being used. Discussion of what CoE is and is not can be found here and here.

Annealing temperatures are not directly related to the expansion coefficient (CoE) of the glass.  This can be shown from the published annealing temperatures for different glasses organised by presumed CoE:

·        “CoE96”: Wisssmach 96 - anneal at 482°C;  Oceanside - anneal at 515°C

·        “COE94”: Artista - anneal at 535°C

·        “CoE 93”: Kokomo - anneal between 507°C and 477°C – average 492°C

·        “CoE 90”: Bullseye - anneal at 482°C; Wissmach90 - anneal at 482°C; Uroboros FX90 - anneal at 525°C

·        “CoE 83”:

o   Pilkington (UK) float - anneal at 540°C;

o   typical USA float - anneal at 548°C;

o   Typical Australian float - anneal between 505°C and 525°C, average 515°C

This shows there is no direct relationship between CoE and annealing temperature.  Do not be tempted to use a CoE number to indicate an annealing temperature.  Go to the manufacturer’s web site to get the correct information.

Temperature equalisation soak

Annealing for any glass can occur over a range of temperatures.  The annealing point is the temperature at which the glass can most quickly be annealed.  However, the glass cannot be annealed if it is not all at the same temperature throughout the substance of the glass.  It has been shown through research done at the Bullseye Glass Company that a temperature difference of more than 5°C will leave stress within the glass piece. To ensure good annealing, adequate time must be given to the temperature equalisation process (annealing). 

From the Bullseye research the following times are required for an adequate anneal soak:

6mm /   1/4"            60 minutes

[9mm /  3/8"           90 minutes]

12mm  / 1/2"          120 minutes

[15mm  /   5/8"       150 minutes]

19mm   / 3/4"         180 minutes

[ ] = interpolated from the Bullseye chart for annealing thick slabs

Anneal Cooling

There are suggestions that a “second anneal” can be used on important pieces.  Other than observing that all pieces are important to the maker, the suggestion should be investigated.  On looking into the idea, it is essentially a second soak at 425°C, which is slightly below the strain point, rather than controlled cool from the anneal soak temperature.

It is reported that the Corning Museum of Glass considers 450°C as the lower strain point – the temperature below which no further relief of strain is possible.  This means that any secondary soak must occur above 450°C rather than the suggested 425°C. Such a soak is unnecessary if the appropriate cooling rates are used. 

Cooling Rate

Except in special circumstances, the cooling rate needs to be controlled as part of the annealing process.  Soaking the glass at the anneal is not the completion of the annealing.  Most practitioners follow the practice of choosing a slow rate of cooling from the annealing soak to some point below the strain point rather than a rapid one with a soak at the strain point temperature.

Annealing is not just the soak time (which is there to equalise the temperature), it is about the rate of the annealing cool too. The rate at which you cool is dependent on the thickness of the glass piece and whether it is all of one thickness or of variable thicknesses.

Even thickness

                                         Cooling rate

Dimension      time (mins)     1^st 55°C   2^nd 55°C

6mm              60                 83°C       150°C

9mm              90                 69°C       125°C

12mm            120                55°C       99°C

15mm            150                37°C       63°C

19mm            180                25°C       27°C

The “first 55°C” and the “second 55°C” refer to the temperature range below the chosen annealing temperature.  So, if you choose to anneal at 515°C, the “first 55°C” is from 515°C to 460°C and the “second 55°C is from 460°C to 405°C.  If you choose 482°C as the annealing temperature, the “first 55°C” is from 482°C to 427°C and the “second 55°C from 427°C to 372°C.

Tack fused/ uneven thickness

If your piece is tack fused, you need to treat the annealing rate and soak as though it were twice the actual total thickness. This gives the following times and rates:

Tack fused

Dimension (mm)                                Cooling rate

Actual     Calculated       time (mins)     1^st 55°C   2^nd 55°C

6            12                 120                55°C       99°C

9            18                 150                37°C       63°C

12          25                 180                25°C       27°C

15          30                 300                37°C       63°C

18          38                 360                7°C         12°C

Contour fusing required firing as though the piece were 1.5 times thicker.  Sharp tack or laminating requires 2.5 times the the actual thickness.

Fusing on the floor of the kiln

There is a further possible complication if you are doing your fusing on the kiln floor, or a shelf resting on the floor of the kiln.  In this case you need to use the times and rates for glass that is at least 3mm thicker than the piece actually is. 

Thus, a flat 6mm piece on a shelf on the floor would use the times and rates for 9mm: anneal soak for 90 minutes, anneal cool at 69°C to 427°C and then at 124°C to 371°C.  It would be safest if you continued to control the cooling to room temperature at no more than 400°C per hour.

But if it were a tack fused piece of a total of 6mm you would use the times and rates for 18mm.  This is using the rates for twice the total thickness plus the additional 3mm for being on the base of the kiln.  This gives the times and rates as being an anneal soak of 360 minutes and cooling rates of 7°C to 427°C and 12°C to 370, followed by 40°C per hour to room temperature.  Any quicker rates should be tested for residual stress before use.

Source for the annealing and cooling of fused glass

These times and rates are based on the table derived from Bullseye research, which is published and available on the Bullseye site.   It is applicable to all fusing glass with adjustments for differing annealing soak temperatures.

Annealing over multiple firings

It has been recommended by this widely followed person that the annealing soak should be extended each time subsequent to the first firing.  I am uncertain about the reasoning behind this suggestion. But the reasons for discounting it are related to adequate annealing and what is done between firings.

If the annealing is adequate for the first firing, it will be adequate for subsequent firings unless you have made significant alterations to the piece.  If you have added another layer to a full fused piece, for example and are using a tack fuse, you will need to anneal for longer, because the style and thickness have been altered.  Not because it is a second firing.  If you are slumping a fired piece, the annealing does not need to be any different than the original firing.

The only time the annealing needs to be altered is when you have significantly changed the thickness of the piece, or the style of fusing (mainly tacking additional items to the full fused piece).  This is when you need to look at the schedules you are planning to use to ensure your heat up is slow enough, that your annealing soak is long enough, and the cool slow enough for the altered conditions.

Determining the annealing point of unknown glass

You don’t have to guess at the annealing temperature for an unknown glass.  You can test for it.  It is known as the slump point test.

This test gives the softening point of the glass and from that the annealing point can be calculated.  This test removes the guess work from choosing a temperature at which to perform the anneal soak. The anneal temperature is important to the result of the firing.  This alone makes this test to give certainty about the annealing temperature worthwhile.

You can anneal soak at the calculated temperature, or at 30°C below it to reduce the anneal cool time.  This is because the annealing can occur over a range of temperatures.  The annealing occurs slowly at the top and bottom of the range. But is at least risk of "fixing in" the stress of an uneven distribution of temperature during the cool when the annealing is done at the lower end of the range.

Do not be fooled into thinking that CoE determines annealing temperatures.  Use published tables, especially the Bullseye table Annealing for Thick Slabs to determine soak times and cooling rates.  Use the standard test for determining the softening and annealing points of unknown glasses.

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

Needle Points

Often fused glass has prickles or needle points around the edges and especially at corners after firing.

This illustration is from Glass Fusing Made Easy

The nature of glass and its interaction with the separators is the cause.  As you heat glass it expands. Once the cooling starts, the glass contracts. Often a particle of the glass sticks to the separator while the rest continues to contract. This dragging of the glass along the separator results in the creation of little sharp points developing as the glass retreats to its final dimensions.

The best solution I have found to reducing the points at corners is to blunt any points or corners before assembly. Only a tiny amount of glass needs to be removed from the corners to reduce the possibility of these points being developed.

Small needle points can also develop along the sides of the glass too.  These are more difficult to avoid.  The most successful method for me is to use a loose separator.  This can be Thinfire, Papyros or a fine dusting of alumina hydrate or powdered kiln wash.  Although less widely available, talc can be used. Talc is known to be carcinogenetic with high exposure, so breathing protection is needed. All these powders provide enough lubrication to allow the runny glass to slide without sticking. 

Of course, you can use boron nitride, which is very slippery, but the cost of it makes it expensive in comparison to the other methods, including using fine diamond pads to remove the needles.

An additional consideration is the temperature you use.  The higher the temperature, the more the expansion.  Expansion rates are almost exponential above the brittle phase of the glass.  Reducing the temperature by 20C and doubling time or more means the glass does not expand so much and the additional time allows the desired profile to be achieved.  

Of course, paying attention to volume control - using 6mm or more thickness - will help to reduce the needle points.  A 3mm sheet both expands and becomes thicker at the edges by drawing more glass from the interior and the edge while attempting to reach 6mm.  This means there is an increase in the needling effect.  Although a 6mm piece retreats on cooling, it does not have the additional thickening effect of a 3mm piece.  Even a 9mm piece retreats on cooling, although the final piece has a larger area than at the start. 

- - - -

There are various preventive measures that can be taken to avoid needle points on fused glass.  These range from altering the edges of the glass, using fibre papers that turn to powder, using refractory powders, or boron nitride. Post firing solutions relate to cold working.

Deep moulds

What is the relation between the diameter and depth of a mould and the diameter of the blank to be slumped into it?  Is there an equation? For example, a mould of 6.5 inches diameter and 3.5 inches deep.

The above example is a deep mould. When compared to the diameter it is more than half the diameter deep. It is a difficult style of mould for many kiln formers.

To explain, some of the differences between deep and shallow moulds needs to be noted.  Deep and shallow are descriptions for the relationship of the span (or space from side to side) to the depth (the distance the glass must slump into the mould).

A 150mm square bowl that is 50mm deep

       Shallow moulds are easier as there are no steep curves to form into. Gentle compound curves generally provide no greater challenges. Shallow moulds with angular corners or abrupt changes of curve to the bottom are moderately more difficult than simple curves.

       Deep moulds are difficult. An example of this is the successful deep vessels of Karl Harron where it is necessary to form the glass through three successively deeper moulds.  His vessels are often deeper than the diameter, making them among the most difficult kilnformed vessels to achieve except roll-ups, which are a combination of kilnformed and blown.

The difficulties with deep moulds are multiple.

·        The glass must stretch more than in a shallow one where only a change in shape is required.

·        As the depth increases, the upper rim is heated more than the centre, being closer to the heat source.

·        The edge develops needle points and stretch marks.

·        The blank becomes smaller in diameter than the rim of the mould.

·        As it slips down into the mould the softened rim of the glass catches on the mould and produces stretch marks.

·        If the upper rim is significantly hotter than the centre, needle points are left on the edge where it catches the mould and stretches to very thin points.

Because of the depth, you will find the finished glass diameter will be smaller than the rim of the mould.

An example of an apparently shallow bowl until you take account of the size of the inner part, making this a deep bowl.  

Over hanging blanks

Do not be tempted to make the blank much bigger than the mould. The most you should risk is 12mm larger diameter than the mould. This means only 6mm is over the edge of the mould.

In deep moulds where I am doing only one slump, I find better success in making the glass slightly smaller than the mould diameter, so the glass does not hang up on the edges.

A blank larger than this mould is likely to hang up on the narrow rim of the mould, and the blank should be slightly smaller than the interior of the rim.

Multiple Moulds

Because deep moulds are difficult to do in one firing, it is best is to start with a larger blank in a shallower mould and transfer the formed piece to successively deeper moulds. 

Some experimentation will be needed to determine the starting size, of course.  A guide would be to measure the length of inside curve of the deep mould from one edge along to the bottom and up to the other side with a flexible measuring tape.  This will give the approximate diameter needed for the blank.  

I start off with a blank of about 25mm less than the measured diameter.  This is to allow for the stretching that is going to occur even at low temperatures. Cut and fuse a clear blank for the test. Find a shallow ball mould of that diameter or larger and slump into it. A second mould may be needed that is deeper and of the size of the now slumped piece with a smaller than original span.  By this second slumping, the piece may fit into the deep mould for the final slumping.

Firing for deep moulds

To fire deep moulds successfully, you need to go slowly (maybe only 75C per hour for 6 mm) up to a low slumping temperature (maybe 630C for 60 to 90 minutes).  This slow rate of advance allows the heat to be distributed evenly throughout the piece.  A low temperature avoids over-heating the rim of the piece during the slumping.  The long soak allows the glass to gradually conform to the shape of the mould without excessive marking.

It is important to monitor the firing of deep moulds.  It is common for the glass to slump unevenly.  Peeking and being prepared to reach into the kiln to shift the glass or even tilt the mould so the lowest part is elevated to receive more heat is important to succeed in slumping into deep moulds. 

Deep moulds require a lot of effort to achieve successful results. You must give special attention to temperatures, rates of firing, soaks and consider the use of multiple slumps – each deeper than the previous.

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

Vitrigraph Pot Liners

Stainless steel vitrigraph pots are durable replacements for ceramic pots that do not last many -  if more than one - firings. But cleaning is not straight forward. Most recommendations seem to concentrate on cleaning by banging the metal to break the glass away from the sides and bottom.  This seems more brutal and noisier than necessary. It will eventually dent the metal and possibly become unusable.

An unlined ceramic pot

An alternative is to line the sides with 1mm or 2mm fibre paper.  Paper of this thickness has enough fibres that the paper will stick together and not contaminate the pull.  It will still protect the metal from glass sticking At the conclusion of the firing and after the cool down, you can remove the fibre paper and have clean sides.

Instead of placing the glass in a bare pot, you can  line the pot with fibre paper 

 

It is possible to put a piece of fibre paper on bottom with a hole in it to match the pot’s hole.  There is a slight risk of drawing fibres into the pull, although I have not experienced it yet.

This method also works with ceramic pots.

Viscosity of Colours

“I have been advised in the past, that blue fires quicker. I was told this by a Master glass maker.”

Viscosity has some relation to colour and intensity.  But you should note black & stiff black are both of the same intensity, and are fusing compatible, but have different viscosities.  This shows that colour is not the only determinant of viscosity, as the stiff black shows the viscosity can be adjusted within the same colour.  The quotation above indicates that the reasons behind any declarative statements need to be investigated.

Some factors in viscosity

Opalescent colours tend to be more viscous than their transparent counterparts.

It is the metals that develop the colours that produce much of the difference in viscosity.  The same metal can produce different colours in different furnace conditions, so viscosity cannot be assumed to be directly related to colour. 

Some people in the past have done their own tests on viscosity and colour relationships, but I have no access to them.  More recently Bob Leatherbarrow shows (Firing Schedules for Kilnformed Glass, 2018, chapter 7.2.5, p.88) some slumping tests done with opalescent glass. It shows how much less viscous black is than white, and that white is the most viscous.  The other results show red a little less viscous than white, then some greens, yellows and oranges, other greens, purple, pinks (in that order) and of course, the least viscous is black.

Transparent glasses tend to be less viscous than opalescent glasses.

How does this information relate to kilnforming practices?  It indicates that a piece with the less viscous glasses requires lower temperatures or less heat work to complete the forming of the glass than more viscous glasses.

When you have a combination of more and less viscous glasses in a piece you need to fire more slowly to ensure all the glass is thoroughly heated through and will deform equally.  You will need to observe and be prepared to move the piece on the mould to straighten it up.

Do your own viscosity tests

You can do your own tests for viscosity differences by arranging 10mm wide strips all the same length (about 30cm) of different colours. These should be placed on a kiln washed pair of narrow batts set parallel to each other 25cm apart and about 15cm high.  Fire at about 150°C per hour to about 650°C, setting the soak to 30 minutes.  Observe at intervals from 620°C.  Stop the firing when the least viscous has almost touched the floor of the kiln. When fired all together at the same time you can see the relative viscosity of the colours tested.  You can label these and store them, or tack fuse these labelled curves to a piece of base glass for future reference.

Achieving the Striking Colour

"Is there anything special I have to do to fire striker glass?  Can I mix striker and non-striker in the same kiln or piece?"

Strikers generally need a two-hour soak at slumping temperatures, about 660C.  This heat soak helps ensure full development of the colour. If the soak is not long enough, the colour may not achieve the target colour at all, or be paler than anticipated.

The rate of advance to the heat soak is not critical.  But it does need to be the appropriate rate for the thickness and nature of the assembly of glass being fired.

If you were to have too short a heat soak, you can fire again to help mature underdeveloped colours.  This will, of course, change the profile of the finished piece.

Strikers are compatible within their manufacturer’s own range. So, they can be combined in the same piece as any other of the glass in the fusing compatible range.  That means that they can be fired in the same kiln load as non strikers.

The two-hour soak at slumping temperature will not harm the later stages of firing, but it might lead to use of a slightly lower temperature tack fusing than without the long heat soak.  That is because of the heat work put into the glass at the lower temperature.   Only observation will tell you how much less temperature is required.  It may be possible that only a little less time at the forming temperature is required.  Again, only observation will tell you that.

Strikers require a heat soak to mature the final colour.  These striking glasses are compatible with the rest of the fusing range from a single manufacturer. Glass from different manufacturers must be tested for compatibility before combined into a project.

More Information on Citric acid

Vinegar is not recommended for cleaning of glass, and especially not to soak glass in to remove kiln wash or investment materials.

The reasons for avoiding vinegar are that

·        Dilute vinegar - as culinary vinegar is - attacks glass, giving a mild etch to the surface similar to devitrification.  Concentrated vinegar – oddly - does not attack glass as strongly.

·        It is of variable quality – due to uncontrolled strength, various culinary additives, etc.,

·        It has a strong odour, and

·        It takes a long time to work.

Citric Acid

However, there is an acid which works very well to remove investment materials and kiln wash without affecting the glass.  It is the humble citric acid.

Citric acid is a weak organic acid that has the chemical formula C₆H₈O₇. It occurs naturally in citrus fruits, although that is not the best source for cleaning purposes.

More than two million tons of citric acid are manufactured every year. It is used widely as anacidifer, as a flavouring agent and chelating agent. It is the last that is of most interest to kilnformers.

Chemical characteristics

A citrate is a derivative of citric acid. There are many formulations. Two examples are a salt that is named trisodium citrate (also known as sodium citrate); and an ester called triethyl citrate. We are more interested in the first as it is cheap and widely available.

The citrate ion forms complexes with metallic cations.  It forms complexes even with alkali metal cations. This makes citric acid an excellent chelating agent, especially of interest in removing kiln wash and refractory materials from glass.

Chelation 

This is a type of bonding of ions and molecules to metal ions. The agents are usually organic compounds. Chelation is useful in applications such as providing nutritional supplements, in chelation therapy to remove toxic metals from the body, in MRI scanning, in chemical water treatment to assist in the removal of metals, and in fertilisers, among other things. 

Citric Acid as a Cleaning and chelating agent

Citric acid is an excellent chelating agent, binding metals by making them soluble. Among many cleaning uses are:

    to remove and discourage the build-up of lime scale, from boilers and evaporators. 

    to treat water by chelating the metals in hard water, cleaners produce foam and work better without need for water softening. Citric acid is the active ingredient in some bathroom and kitchen cleaning solutions. 

    A solution with a 6% concentration of citric acid will remove hard water stains from glass without scrubbing. 

    Citric acid can be used in shampoo to wash out wax and colouring from the hair. 

    In industry, it is used to dissolve rust from steel and to form a coating on stainless steels to resist corrosion.

Its use in kiln forming is to make use of the chelation properties when dealing with kiln wash and investment material residues.  Aluminium hydrate is the main ingredient of all kiln washes.  When it becomes bound to glass, it is impervious to almost all chemicals.  The chelating property of citric acid enables the bond between the glass and the kiln wash to be broken by incorporating the molecules within its own, making a colloidal solution.  This process is approximately 6 times faster than any vinegar solution and without the odour and etching risks.

A sample of the affected glass followed by 4 hours in citric acid and 24 hours in vinegar.
Credit: Christopher Jeffree

A 5% solution made up with 50gm of granular citric acid in 1 litre of water is all the strength that is required. The affected glass can be soaked in this solution for the time required to complete the chelation without the risk of etching, and without needing ventilation to remove smells.  Unless you are using a lot of cleaner, it is better to make up much smaller amounts as mould can grow on this organic solution.

A 5% solution made up of 50gms citric acid in 500ml of water and 500ml of isopropyl alcohol makes an inexpensive and effective glass cleaner. However, if left for a length of time, it becomes sticky.  Apply the solution, scrub the glass and immediately wipe off the solution.  Then polish the glass dry. The alcohol in the solution makes keeping large quantites possible. 

This post was compiled with the assistance of Wikipedia, Christopher Jeffree and my own experience.

Although this post remains valid, there is another chemical for long soaks to remove mould material or kilnwash.

Citric Acid Cleanser

Christopher Jeffree has kindly outlined the reasons for the effectiveness of citric acid as a cleaner for removing refractory mould residue and acting on kiln wash stuck to glass.  This is his work (with a few personal notes removed).

"Citric acid works well for removing the plaster scale that builds up in vessels used to mix plaster, and it helps to remove traces of investment plaster and kiln wash from glass.  Its metal-chelating properties probably help with dissolution of calcium deposits, but I am less clear why it is so good at removing kiln wash.  The constituents of kiln wash are kaolin and alumina hydrate, neither of which I would expect to be soluble in dilute acids.  Equally, the refractory materials in investment formulae I would expect to be insoluble.  However, kaolin forms layered structures in which flakes, molecular layers, of alumina hydrate and silica interact through hydrogen bonding. It is possible (I am guessing here) that citric acid can disrupt those hydrogen bonds, thereby disaggregating the clay.  All we can say is that empirically, it works.

"I prefer to use citric acid partly because it has a defined composition, but also because it is safe and pleasant to handle – no odour, and comes in the form of easily-dissolved dry crystals like granulated sugar.  Vinegar stinks, and glacial acetic acid is  an aggressive flammable, corrosive liquid with a chokingly acrid smell.

"Calcium sulfate has low solubility, but is not completely insoluble in water - gypsum (calcium sulfate dihydrate) has a solubility of about 2.5g per litre (0.25%)  from 30-100 C. Its solubility is retrograde, meaning that it decreases, rather than increasing, with temperature.  Natural gypsum is an evaporite, a type of rock that often forms by evaporation of lake water in a geological basin with little or no outflow. It can also be produced hydrothermally in hot springs, when water containing sulfuric acid passes through limestone.  

"Calcium citrate is not very soluble either, only in the order of about 0.85g per litre, but the important thing from our point of view is not to get the material into solution but to separate its crystals and make it detach from the glass.

"In other contexts, warm citric acid is used by jewellers and silversmiths as a pickle for dissolving copper oxide (firestain) from silver and gold alloys  after heating / soldering.  It is a safer alternative to the traditional jeweller's pickle of 10% H₂SO₄.

"Citric acid also dissolves rust from iron, without much etching the iron itself, so is good for cleaning rust off tools etc.

"These pictures show a plaster mixing bowl with (presumably) CaSO₄-rich deposit on the surface, cleaned by soaking with 5% citric acid for 4 hours,

and flash from the pate de verre castings with tightly adhering kiln wash, cleaned using 5% citric acid soaked for 4 hours, and vinegar (white wine) soaked for 24 hours.

"I'm not sure about reaction products - I was speculating a lot there, running through hypotheses that I can't support. We don't really have data on the composition of the layers that are stuck to the glass, or a clear idea of why they sometimes stick and sometimes don't (e.g. the differences between transparent and opal glasses in this respect). Maybe this would be a topic to discuss with technical people at Bullseye."

Hope this helps
Best wishes

Chris Jeffree

Subsequent to this work Christopher has done more work and found that Tri-sodium citrate is an even better chemical for cleaning glass of kiln wash and mould material.

Placing of glues

The placing of glues to hold the glass pieces temporarily is important.  Often unsightly black marks appear due to inappropriately placed glues.  Bubbles can form between layers  and even appear to come from underneath the glass for the same reason.  Placing is often more important than the amount of glue used.  Still, the amount used should be the minimum to hold the glass from moving from bench to kiln.

Place glues at edges of the pieces to be secured during movement.  This allows the burn-off of the glue to evaporate without being trapped under the glass.  If you use very runny or diluted glues, the capillary action will draw the required amount of glue under the glass piece to form a secure adhesion.

Glues burn off and leave the glass pieces unsecured long before the glass becomes tacky enough to stick together.  This means that if your stack of glass will not stay in place without glue as you build it, the glass will collapse or move in the kiln.  Glues are only suitable to stabilise the glass pieces while moving to the kiln.

Two recommended glues that burn off cleanly are the Bullseye Glasstac (more fluid) and the Glasstac gel (more viscous)