Interesting to see if it passes the shock testing. I would think maybe you have slowed the problem not rid the problem but I am always surprised in ceramics. Maybe the cone6 instead of 5 has changed the expansion of the glaze.
I will let you know. I assume the best testing would be
> dishwasher run several times - like over two weeks, just leave it in there?
> freezer for 1/2 hour then boiling water?
for now, I will advertise these as not microwave or oven safe. my intention was for them to be decoration or light use, but you know people wont listen.
I use a slightly harsher version of the oven / cold water shock test to test for crazing. It’s a good indicator if the glaze is going to have delayed crazing. I also do porosity / absorption tests for the clay (before doing the crazing tests) If the clay is too porous the glaze will have delayed crazing as moisture soaks into the pot over time.
This is what I do, test pieces in oven at 300F for around 20 minutes to make sure they are heated all the way through then plunged into cold water. If there is no visible crazing I bump the temp up to 310F and repeat then repeat again at 320F. After the last cycle I brush calligraphy ink on the test pieces then rinse them off, hopefully no crazing lines. I know others say going to 300F X 3 cycles is enough so I might be a bit extreme.
The freezer test is for shivering, glaze in compression. Thin walled cylinder, glazed thickly on the inside only, right up to the top and just over the rim. Pot in coldest part of freezer overnight then put in the sink and filled with boiling water. If the pot splits open or shivers of glaze come off the edge of the pot the coe is too low. [(glazes in compression are far less common than ones in tension (crazing ones)]
Yup on using the pots for a while and seeing how they standup, a couple months in the dishwasher is good too for checking glaze durability.
I agree about adding silica and alumina to fix crazing, if the base can support the amount needed to do so. Otherwise you need to change the fluxes, obviously not possible with a commercial glaze.
I think there is another way of looking at coe figures and their relationship to crazing. That being to look at glaze “families” or “systems”.
This is from digital fire:
Results are determined by the set of expansion numbers (different values are available from different sources) and the method of additive calculation method chosen (based on formula or mole%). Thermal expansion values predicted by calculation are relative (not absolute) and apply within 'systems'. Thus, if a glaze calculates to a higher expansion than another, and is in the same system, then it is more likely to craze. For example, if you have a dolomite, whiting, feldspar, kaolin, silica glaze and you try a bunch of variations, the calculated expansions will give you an indication of which variations have higher and lower expansions. But if you introduce lithium carbonate, or boron frit, or zinc, for example, now you have a different system. Also, some oxides, like Li2O or B2O3 do not impose their expansions in a linear fashion, thus they do not calculate as well.
I use a clear with a coe of 5.76 (using Insight). It is bombproof and does not craze on my clay, have used it for many years. Using the same clay I recently ran some tests for a friend with a glaze that has a coe of 7.20. I tested it on the same porcelain clay that I use the low expansion glaze on. To my amazement it does not craze either. I ran exhaustive crazing tests on it and all samples are craze free.
Neil E. got me thinking about this when he posted his clear glaze for porcelain a few months ago. At the time it didn't seem logical to me why it didn’t craze since the coe came in fairly high. Now it makes sense.
I think glaze blisters are one of the harder things to narrow down the cause of since there are so many possibilities for what the culprit is. If you feel like doing some reading there is a link to a Ceramics Industry 4 page article on blisters below, hopefully something seems a logical avenue to explore to fix the problem.