How I Developed the G2926B Cone 6 Transparent Base Glaze

How I found a pottery glaze recipe on Facebook, substituted a frit for the Gerstley Borate (using glaze chemistry), compared using a melt flow tester, added as much extra SiO2 as it would tolerate, and got a durable and easy-to-use cone 6 clear.

Click here to watch this at youtube.com or click here to go to our Youtube channel

Transcript/Notes

1
Every potter needs a reliable transparent glaze that fits the studio clay bodies, works well with colorants and has good mixing and application properties.
I want to show you a different approach on how to do more than just try one new recipe after another.
Here is the first thing that’s different: Treat it as a project worthy of being in your permanent testing history. The best way to do that is in an account at insight-live.com. I’ll start by adding a new recipe. I already have the new one I want to try on the clipboard, so I'll click the "Import button”, paste it into the box, check "Confirm", then "Interpret".

more..

2
Notice this material is not being found. That means it won't contribute to the calculated chemistry I am going to be doing shortly. In the Materials Manager I’ll search it in “Our Materials” (rather than the reference ones).
Notice that we have our own version of this material defined - this was done in order to include a cost. However it is still pointing to the reference one for the chemistry.
I'll click the blue edit button in the recipe panel and fix the spelling.
I'll also type in the name of the glaze.
One more thing to notice: I am not entering anything in the units column, this is only needed in special circumstances.

3
Here is the second thing I do that’s different: I assign it a code number: In this case "G2926A", it is the link between physical tests and their testing data. It gets written on every test sample and every container. Finally I save it by clicking “Done Editing”.

4
There is the third thing I do that is different: I sanity-check the recipe.
Notice this material: Gerstley borate (please forgive me for my pronunciation). This puts up an immediate flag, we'll learn more in a minute. This was a natural calcium borate mineral - actually, it was a "clay". Although complex chemically, the reason it was so popular is this oxide: B₂O₃ - this is boron. This oxide is the key enabler for glaze-melting at cone 6. If you need a reminder of why boron is so important, just click "here".
Back to the question here: Do I really want to use this material? Today the answer is "no" because it is no longer available. Yet you will find this in hundreds, perhaps thousands, of recipes. Years ago the answer was also no - but people endured all of its issues anyway, mostly it was because they didn't know the simple alternative we are about to use! If you would like to learn more, just click here.

5
There is the fourth thing I do that is different: I look at the chemistry.
Notice that, in addition to boron and minor amounts of other oxides this material sources significant calcia, magnesia, soda and silica. So, let’s replace it with a calcium borate frit - which also sources those other oxides.
To start, I’ll duplicate this recipe and click the blue edit button and name it "Studio Clear v2" with a code number of "G2926B".
I'll replace the material name with: “Ferro Frit 3134" and click "Done Editing".
Let’s open these two materials in side-by-side panels. The frit has a little less boron but the other oxides it delivers look familiar, although in quite different proportions.

6
Next, let's turn on chemistry calculation for the two recipes so we can compare them on the oxide level.
Let’s start with the boron, the yellow number. It is "down" in the new recipe, more frit is needed to supply enough.
While I could just repeatedly edit recipe amounts to zero in on the 0.33 target, there is a better way.
I'll start "calculation mode", set the increment to 2, and nudge up the frit amount until the boron matches.

7
Next, let's add a little talc to contribute the magnesia. I'll add a line, enter the name, an amount of one, and finish editing. I’ll change the “increment amount” to 0.5 and click its up-arrow until the MgO matches that in the chemistry of the original recipe.

8
Now, let's fine-tune the K-N-aO. Notice it is also supplied by the nepheleen syenite. I’ll reduce that, three at a time, until the oxide amount matches the 0.32 value of the original oxide formula.

9
Just a note that I have fixed an error here - the recipes should be code numbered G2926 and G2926A (rather than the A and B I had before).
We are down to matching the Al₂O₃ and SiO₂. Well, not quite matching - I decided to increase both by a little, since I was sure the frit would make the glaze melt better - and thus be able to tolerate, and benefit from them.
To do that I first increase the EPK by three-at-a-time to take the Al₂O₃ to 0.48 (that's the red number).
Last is the SiO₂, the blue number - I increase that enough to get a silica-to-alumina ratio of 9 to 1 (if you want to know about that ratio just click the word "Ratios").

10
There we have it, the chemistry of the improved recipe is doing what I want: Similar to the old, but better.
To test this I printed a batch mix ticket to make 1000 grams.
That was enough to fire test tiles, but also to do this melt fluidity test comparing it to an industrial glaze used in the tableware industry (that is the one on the left). Notice how well our new glaze is melting - this is good news as we will see in a moment.

This looked so nice in initial tests I decided to mix up the original recipe, the one I didn't want to even try, and this one, but with an addition of 10% zircon to opacify them white. Our new fritted version is the one on the left. Notice how much whiter it is - the frit is obviously lower in iron.
Let's look at another picture, this one compares the flow on this G2926A version to the original G2926 - it is flowing better even though I increased the silica and alumina.
Of course, as a normal part of our workflow, I would attach these pictures to the recipe using this button.

11
Then something bad happened: I did a 300 degree to ice water "stress test" on a clay I intended to use it on - unfortunately it crazed like this one.
However, since melting was good there was an opportunity to add silica, plenty of it. Silica is the glass-former in any glaze so the more you can put in the better. And, it reduces thermal expansion. Here is what I did.
First, I duplicated the recipe and code numbered the new one G2926B.
Then I retotaled it to 100, added 10 parts silica as another line and set it to added status.
Then I showed the line order column and brought the talc down to the bottom.
Finally, I clicked "Done Editing".

12
I fired a melt flow test comparing the A and B version. Notice the 10% silica is tolerated really well - only flowing a little less. And, as you can see here on the inside of this mug, it fires super glossy. Better yet, it survived the crazing stress test.
Later on, we made another important improvement, changing to 325 mesh silica. It is not a separate item in the materials database, as you can see I have put the generic name in the "material lookup" fields so the chemistry is still found.
In further testing we accumulated lots of notes. For example, I found that equal parts of powder and water produced a creamy 1.43-1.44 specific gravity slurry that worked well for dipping. And that 1000g produces about 1.4 liters. That would be about 3600g to make a US gallon.
This was intended to emphasize the process that led to the derivation of this popular cone 6 transparent glaze recipe. We did not end up with the exact same recipe (ingredients are at worst 0.5% different, likely because of rounding that was done).

Links

Articles	A Low Cost Tester of Glaze Melt Fluidity Use this novel device to compare the melt fluidity of glazes and materials. Simple physical observations of the results provide a better understanding of the fired properties of your glaze (and problems you did not see before).
Materials	Gerstley Borate Gerstley Borate was a natural source of boron for ceramic glazes. It was plastic and melted clear at 1750F. Now we need to replace it. How?
Recipes	G2926B - Cone 6 Whiteware/Porcelain transparent glaze A base transparent glaze recipe created by Tony Hansen, it fires high gloss and ultra clear with low melt mobility.
Glossary	Transparent Glazes Every glossy ceramic glaze is actually a base transparent with added opacifiers and colorants. So understand how to make a good transparent, then build other glazes on it.
Glossary	Glaze Chemistry Glaze chemistry is the study of how the oxide chemistry of glazes relate to the way they fire. It accounts for color, surface, hardness, texture, melting temperature, thermal expansion, etc.
Typecodes	Gerstley Borate Glaze Calculation examples Examples of how we use glaze calculations (in an Insight-Live.com account) to replace Gerstley Borate with other materials, especially frits, in various glaze recipes. In doing so we take the opportunity to improve the recipe in other ways (e.g. reduce thermal expansion, improve slurry properties, reduce bubbling and crawling).
Typecodes	Glaze Chemistry Case studies where glaze chemistry was used to solve a problem.
Troubles	Glaze Crazing Ask the right questions to analyse the real cause of glaze crazing. Do not just treat the symptoms, the real cause is thermal expansion mismatch with the body.