Ravenscrag Slip really shines in its ability to produce a good floating blue glaze at cone 6, this is the GR6-M recipe. The speed of cooling in the kiln affects the appearance. The mug on the left was cooled faster, using our drop-and-soak PLC6DS firing schedule. The other one was slow-cooled using the C6DHSC schedule. The latter schedule is preferable for these because the G3914A black has a much smoother surface. The blue could be recovered by adding more cobalt.

The only difference between these two cone 6 glazes is the silica. Both are the G2926B recipe, both were thickly applied and fired in the same kiln. The left one employs the 90 micron (or 200 mesh) grade silica and the right one uses 45 micron (or 325 mesh). These test tiles are about 6 months old. There was no crazing out of the kiln. The porcelain recipe is 25% silica, 25% nepheline and the remainder kaolin and bentonite. It appears the finer particle size silica is dissolving in the melt much better, this narrows the difference between calculated and actual behavior, especially relating to coefficient of thermal expansion. While this grade is better in glazes it is not better in bodies, they most often depend on the thermal expansion increasing effects of the larger particles in the 200 mesh grade.

A reader noted that it is also a matter of the reaction between glaze and body. The original glaze having coarser silica would have smelt and reacted with the body more, the extra dissolution sourcing Na2O - thus increasing the COE of the glaze. Conversely, when the finer silica dissolves it increases melt viscosity thus reducing reaction with the body.

This mold is for a Medalta Potteries ball pitcher having a closed top with a teardrop-shaped spout (lower right). This plaster mould does not need a spare - 3D printing makes it possible to create a pour spout that inserts perfectly into the angled hole. The size of the pour spout reservoir and the degree of insert can tuned so that when the level drops to the bottom it is ready to pour and the hole is perfectly formed.

This blistering problem is common in high-temperature rutile blue glazes - this is not saleable. The reason relates to what it takes to create this kind of vibrant variegated aesthetic: Melting the crap out of the glaze and cooling it just right. This particular one is being fired to cone 11 down to get enough melt fluidity to make it crystallize and phase separate. It seems logical that if the glaze is melting so well it should be able to heal any bubbles that form and break (these are more than usual because the body is being overfired and generating gases). However, the fluidity comes with surface tension that can hold the bubbles intact. Each of these holes in the glaze is a product of that - plus another factor: Cooldown is rapid enough that the melt is not sufficiently fluid to heal after bubble breakage. The potter has been using this glaze for many years with success, but a small change in process or materials has occurred to push it past a tipping point. Solutions? A drop and hold firing. Add a flux (e.g. a little lithium or a frit) to make it melt fluid at cone 10R (where the body generates less gasses of decomposition). Replace any high LOI materials in the glaze itself with other materials to source the same oxides.

I have not made molds for years because I dread the process, the mess, all the supplies and tools. I am not a mold-making expert either, but I found a way to do it that is fun, rewarding and effective.
-I am wasting less plaster (it is not a green material). And PLA filament is corn starch or sugar cane. And I am not using rubber.
-I spend most time on design, pouring the plaster takes minutes.
-Many fewer tools are needed, the process is less messy.
-No natches make sanding of flat mating faces possible (for better seams than I've ever had).
-No spare is needed, the 3D-printed pour spouts works better.
-More shapes are possible.
-My molds aren't right until at least version 3. 3D makes do-overs or changes in design as easy as a reprint and plaster pour. I can make a mold just to test an idea!

This is what typically happens when applying a dipping glaze over an underglazed bisque-fired piece - it does not stick. Commercial underglazes impede the absorptive powder of the bisque but it still might be possible to make it work. Here are some options:
- Bisque the under-glazed ware to a low temperature (e.g. cone 010 or lower) and apply dipping glaze to that.
- Make your own underglaze for dipping, leaving out the CMC gum.
- Make your glaze thixotropic - because it is gelled it may hang on in an even layer over the underglaze.
- Make a brushing glaze and paint that over the underglaze. 1%-1.5% CMC gum will make it paintable. 1% additional Veegum will gel it enough that it can tolerate more water and go on thinner, this enables applying multiple coats and gives good control of the final thickness.

This is possible because Polar Ice, the casting version, behaves like rubber after draining - because of 1% added Veegum. The rims can be peeled away from the mold and the piece can be completely collapsed to forcibly peel it out of the mold. They could then be opened and dried. To add more insult to the wonky shape large pieces can be broken away at leather-hard stage and then reattached using the slip as glue - the piece will still dry normally! The 1% Veegum is solely responsible for all of this, without it the New Zealand kaolin based slip would be too fragile to even cast this.

Orton says “90 angular degrees is considered the endpoint of cone bending”. First, let's assume the normal: Examination of cones on kiln-opening to verify controller operation. Consider the cone on the left: The tip is touching. But it is also beginning to buckle, which means it was touching for a while before the firing ended. Who knows how long! The second one is not touching but has still fallen a little too far. Why do we say that? The third one, positioned on the Orton guide, has reached the recommended 90 degrees. This demonstrates a good reason why self-supporting cones are much better than standard ones: They are not touching when considered done. And standard cones, when sent in a 3/4" plaque, have a less consistent bending behaviour.

These cracks have been drawn because we were unable to get this to happen in our studio - likely because one condition was not met. This is M340, it is made from minimally processed natural mined clay. While inherently very fine it does have some sand particles. Consider the combination of conditions we created to try to make this happen:
-An unstable bowl shape (flared out to the edge of the clay's ability to support itself).
-Cutting the rim off with a needle tool and leaving that flat surface uncompressed and with water on it.
-Applying slip just after throwing the piece (rather than waiting until leather-hard).
-Not wedging the clay (or wedging it well).
Had all these been true and the clay was also soft and thrown using a lot of water (rather than slip) it could have split. Clays made from 200 mesh industrial minerals can even endure all of that. But to make this very unlikely just do the opposite of the above: Stable shapes, thrown rims, slipping later, wedging well and stiffness matching the shape being made.

This test mold is thin-walled yet I can cast three thick-walled mugs in three hours. This clay is L2596G, a buff burning cone 10 stoneware - the mug on the lower right has been fired to cone 10 oxidation. Achieving 4-5mm thick walls is not a problem if the casting slip employs a large particle kaolin intended for this purpose (e.g. Opticast). The flared lip works as expected, keeping the rim nice and round. No cracks have appeared at handle joins, even for pieces left in the mold overnight. The mold halves mate with each other very well and the seam is easy to remove. The seam on the base is an issue - I have to be careful to line up the halves well before clamping the mold strap - this is a warning for accuracy during the mold production stage. And the possible motive for a three-piece mold if I get more serious about this piece.