Fixing a glaze mistakenly made using frit 3134 instead of 3124
Share from Insight-live.com (Lab Documentation and Calculation System) by Digitalfire. https://digitalfire.com
This is a fascinating example of how valuable a little basic glaze chemistry can be in dealing with issues like this.
Project Name
A Glaze Mixing Error Fixed Using Chemistry
Project Codenumber
UnAssigned
Notes
Glazes fire the way they fire because of their chemistry. The kiln fires build the glass using the oxides supplied by the materials. While the mineral presences of the materials themselves also affect the way they combine in melting, when frits are the major melter, chemistry is the most relevant.
This project centered on a mistake made by a potter. She mixed 5 gallons of a glaze and made a big mistake: She used frit 3134 instead of 3124. Fortunately, these frits are similar. Except for one thing: Frit 3134 has almost no Al2O3 (3124 has lots) and significantly less SiO2. That is a game changer, making it melt dramatically more and lower. Not surprisingly then, her glaze melted way too much, running off the ware.
But, the mistake had an element of luck. Frit 3134 can be viewed as a "chemistry subset" of 3124. Adding the right amount of Al2O3 and SiO2 moves its chemistry toward 3124. That is what made it possible, by an addition of material to supply these oxides, to fix the glaze. What materials? Kaolin supplies Al2O3 and SiO2 and silica supplies SiO2.
To actually do this I entered the original glaze as a recipe, then duplicated it and simply added kaolin until the Al2O3 values matched. Then I added silica to match the SiO2. As it turns out, the addition turned out to be:
For each 100 grams of powder of the original, add 15 calcined kaolin, 10 EP kaolin and 17 silica. Thus, if her original batch was 10,000 grams it would be 1500 calcined kaolin, 1000 EPK and 1700 silica. Based on the results of the fired testing I did, this new recipe has better transparency.
You may have noticed something I did not mention: As a cone 6 glaze, this one has triple the amount of frit one would normally see (which means triple the level of B2O3). My adjustment adds even more B2O3 (reason why explained later). That is obviously why it has the degree of melt fluidity it does. Is this good? Yes, but as long as one knows. If this recipe were shared it would be important to label it as having "high melt fluidity" (to protect people's kiln shelves). Melt-fluid glazes like this are very useful to create reactive effects with color, opacity and variegation. But their application thickness must be carefully controlled. Such glazes commonly have high thermal expansion (and therefore craze), while this one is not high it is high enough that it certainly could craze on some bodies. This one will melt well at much lower temperatures, likely down to cone 02. One asset that it has is that the Al2O3 and SiO2 levels are what is normal in a cone 6 glaze, meaning that it should be reasonably durable.
*This is the original recipe that was intended to be mixed. A pottery made a large quantity. But she mistakenly used Ferro Frit 3134 instead of 3124. Of course, the glaze turned up to be much different, running off the ware because it was melting so much more.
Pictures
G3904, G3904A flow test, glazed tites
These were fired using the drop-hold-slow cool schedule. Fired behaviour almost the same. The A is more transparent and runs a little more. It is applied a little thicker on the tiles so has more bubbles.
*Here is the same recipe, with the Frit 3134. I have added enough EPK and Silica to restore the levels of Al2O3 and SiO2 in the original recipe. There are a couple of considerations in doing this:
1. If I would have added regular raw kaolin the total percentage would have been raised enough to be a cause of excessive shrinkage during drying. That would produce drying cracks and fired crawling. I have employed a combination of raw and calcined kaolin to alleviate this situation.
2. The switch to Frit 3134 has increased to B2O3 content somewhat. That was unavoidable. That means the glaze should melt more. And that is what happened. But this is not enough to be an issue. A benefit is the extra melt fluidity has enable the glaze to shed many more micro-bubbles, so it is more transparent.
3. The level of KNaO is a little higher. That would normally raise thermal expansion (and risk of crazing). But the higher B2O3 has countered that and kept the expansion the same (6.8).
Pictures
G3904, G3904A flow test, glazed tites
These were fired using the drop-hold-slow cool schedule. Fired behaviour almost the same. The A is more transparent and runs a little more. It is applied a little thicker on the tiles so has more bubbles.