The compatibility of blowing colors from Reichenbach and Gaffer with blowing batches is well understood by most artists. Much less is known about compatibility when these materials are kilncast into slabs and related forms. I've tested various colors with my dual-purpose glass (see below) for compatibility and durability over 2014. Incompatible colors develop fissures (shallow cracks) at the interface of the color with the batch glass; these typical form quite quickly and are obvious.
Samples of rod were blown out and (reverse) overlaid onto batch glass, and then cased and pulled into thick (1"+) cane. This process works the solid color into a "tube" of color of sufficient density. The cane is then cross-sectioned into pieces ranging from 1/2" - 2" with a chisel, and cast into 1" thick slabs. Square samples are cut from the slabs with a diamond saw; fissures tend to develop along the interfaces at cut edges if the test-color is incompatible with the batch glass.
Some colors change when subject to the relatively high temps and long durations at high-fire in a kiln. Many opaques tend to become transparent or translucent. Copper reds can become brown or black. Gold-based yellows and oranges can acquire a pink tone. I've developed a sample set of kilncast colors to help remind me of these changes.
The majority of colors I've tested so far seem compatible when kilncast, provided that they are of an appropriate thickness (color density) and annealed properly. Generally they tend to have lead bases and lower viscosities when blowing. The incompatible colors tend to be opaque reds, orange, yellow, and some greens. These can sometimes be problematic with blow glass as well.
The following lists summarize compatible colors, with R for Reichenbach, G for Gaffer, K for Kugler, and Z for Zimmerman. All color tests utilize rods. An * denotes a "crazing" effect at the interface, without structural defect.
Reichenbach transparents: R3; R5; R6; R8; R12; R13; R14; R23; R24; R28; R33; R35; R43; R46; R49; R144; R146; R199; R212; R215*; R218; R231; R711; R730; R740; R770; R8904
other transparents: G62; G66*; G70; G71; G72; G75; Z640; Z661; Z672; Z721; K42
Reichenbach opaques: R60; R61; R87; R91; R120; R130; R191; R192; R223; R227; R241; R246; R316; R710; R712
Gaffer opaques: G123; G131; G139; G141; G171; G173; G176; G178
note: G62 appears compatible but requires verification
R67; R69; R71; R117; R141; R142; R143; R147; R156; R216; R315; R4291; G129; G175; G177; G183
note: R216 and G129 appear incompatible but require verification
Formulating a Dual-purpose Glass
I've spent the better part of 2013 trying to understand the technical aspects of glass. I want a batch for both blowing and kiln-casting. Commercially available options are decreasing in variety, increasing in cost, and only suited to one application and not both. Several commercially available glasses for blowing (batch, pellet, or nugget) were tested for devitrification in kiln applications. All devitrified significantly after two firings, and most after only one.
I sent a sample of a glass (no longer commercially available) to American Glass Research (AGR) for composition analysis. They use X-ray fluorescence spectroscopy to quantify most metal oxides in a 300g sample of glass; oxides of lithium and boron are quantified by wet-lab techniques. The sample was procured in the late 1990's as a batch for blowing, and showed no compatibility issues with Reichenbach and Zimmermann color over the trial period. A recent test-fuse with cullet from Spruce Pine 87 showed sufficient compatibility. More importantly, the sample showed no surface devitrification after several firings in plaster-silica molds. To my knowledge, it has not been commercially available for several years.
A recipe was created from the formulation results from AGR. The formula was batched in a cement mixer fitted with a simple lid to contain the dust. Initial batch sizes were about 40 kg, and mixing action seemed to deteriorate with batches over 60 kg. These small batches did not affect seem to affect the melt in a significant way (compatibility variation, chord formation, stones, etc). Addition of 2% - 4% water to the mixed batch helps to control the dust, and didn't seem to affect compatibility. A solution of sodium silicate will pelletize the batch; the recipe must be corrected for the silicate addition prior to mixing the batch.
Special additives were reduced or omitted in the formulation. An addition of 0.05% -0.1% As2O3 sufficiently refines the glass during cooking and squeezing. At 0.05%, a few pin-sized bubbles remained after a 5-hour squeeze at 1950 F, but these quickly vanished when the glass was brought to the working temperature of 2060 F. Decolorizer (as Er2O3) was omitted from the formula. Most Erbium oxide comes from China, which controls the pricing of rare-earth materials. Erbium has become increasingly important in glass for solar panels and related applications; prices for rare-earth oxides have soared in recent years. While the glass seems sufficiently colorless, there may be some room for improvement.
Rough estimates of compatibility were obtained by viewing samples with polarized light and filters. Each melt was assessed for compatibility with both Spruce Pine 87, and with the previous melt(s). All melts appeared fully compatible with one another, and with the Spruce Pine control. The formula does show some stress with System 96 sheet glass and with Gaffer batch. Gaffer offers trident-seal testing for batch glasses, an I plan to optimize the formula for their blowing colors with their help. Gaffer formulates their base glasses to a standard as determined by a trident-seal test, and they test their color batches as part of their quality control protocol. Numerous colors of Reichenbach and a few from Gaffer were asessed for compatibility. All colors seem fine in blown or hot-worked pieces; but some colors (about 10% - 15%) show stress and/or structural issues when cast, notably reds and some opaques. Work continues in this area to more thoroughly identify the shades and parameters associated with poor fit in castings.
The formula shows very little devitirification in casting. Samples without plaster/silica molds may be fired multiple times without any devitrification; samples fired with plaster/silica molds sometimes show some slight fogging around their edges. No internal devit was observed with multiple firings. Abraded surfaces (exterior) melt without devit; while interior surfaces exhibit a slight veiling, which probably arises from small bubbles imparted from the abraded surfaces.
As a batch for blowing, the formula works well. It has a nice working time and consistency, adequate thermal shock resistence, remains consistent thru the pot, and works at a reasonable temperature (2060 F). Its about as forgiving as Spruce Pine 87, but not as long as Gaffer batch. As a casting glass, it seems to flow adequately in the range of 1530 F to 1570 F. The slumping temperature has not yet been assessed.