Thermal expansion mismatch between different glasses is one of the bigger problems glass casters, glass blowers, bead makers and fusers face in their daily work. Several glass manufacturers serve the market, and produce many different colour properties, expansions and viscosities. In addition, there are many different clear batches and cullets. It can be a nightmare figuring out what will fit with a base clear glass. Besides, some of the manufacturers lack consistency within their own range, and glass blowers, in particular, are unsure how to effectively test their clear base against different colors. Thread and ring tests are not always reliable, particularly when comparing glasses from quite different glass families. Industry standard dilatometers are very expensive and have measurable accuracies no better than 1-2%. Dilatometers don’t yield useful information without knowing the set point or strain point of the respective glasses being measured and compared. That requires another, equally expensive instrument, that measures viscosity with only the same degree of accuracy as dilatometers.
Gaffer Glass carries out the same thermal mismatch test as Corning Inc, employing a polarimeter and a trident seal. This test bypasses dilatometers and viscosity meters, which are notoriously inaccurate. By contrast, the trident seal produces an easily read and very accurate control. Any studio can fashion it cheaply and accurately. The seal can be passed on to a third party, where the results can be compared under the same conditions. It takes care of different viscosities and variable properties such as Young’s modulus and
Each coloured melt is evaluated using apolarimeter, referenced against a transparent Gaffer master glass (see Gaffer Batch). The measurement of mismatch is read as a measured polarmetric number, i.e. degrees of retardation in nanometers/centimeter. We aim to have a mismatch reading no more than 30 nm/cm, which for most glasses is roughly a linear expansion coefficient (LEC) difference of ± 0.5 x10-7 up to the strain, or set point, of our coloured glass range. That is well inside a mismatch of ± 2.0 x 10-7, which is claimed to be the outer limit of acceptable thermal mismatch difference, (although we disagree; ±1.5 is safer) and clearly superior to the 7-8 point spread that we have measured taking random samples from our raw colour competitors.
Only after performing this daily test and seeing the result we require, does the melt get released onto the global market. The colour bars are for glass blowers and are clearly labelled “Tested: Compatible/Ultrasound”, meaning they have passed both the trident seal test and have been checked by an ultrasound scanner for excessive seed.
Be assured: all frits of blowing glasses are also compatibility tested.
All casting glasses, both frit and billets, meet an even higher standard: that is less than a 10nm/cm mismatch, which is less than 100psi strain in a casting format, nearly ten times below what is considered to be safe and acceptable (1000psi).
The coloured flashing or blowing glasses we produce should preferably be used in conjunction with a generic soda/lime etc. base clear, showing a measured expansion of around 96 x 10-7 (0-300oC) and an annealing point around 485-500oC (905-932oF).
Glassblowers: if you melt your individually designed recipe, send us your clear base cane to be tested against our standard. We require a 2-3 mm diameter cane (1/10th inch), not out-of- round or tapered (± 0.05mm –you will need a micrometer) and at least 300mm long. We offer a free evaluation service. If possible, send us the formula of the glass, so that we can estimate its properties, in advance of physical testing. We will post back to you a trident seal sample, matched with our control, and estimate the degree of strain with our standard coloured range. You can confirm the results with a simple polariscope. Large disparities in LEC’s (at around ±4 x 10-7) cannot survive this test. We offer a free service where we can adjust your batch formula to an optimum expansion coeffiencient theoretically, and then follow up with trident seal analysis on subsequent melts.