Granite Processing - Batch 2

The initial milled "batch" of granite, Batch 1, consisted of a number of individual batches that exhibited a fair amount of variability in results, most notably due to the iron content present in each batch. A lab test confirms a sample of the hillside Inverness granodiorite contains 4% iron oxide. We also know observationally this can vary from sample to sample and therefore small batch to small batch.

The effect of iron was most noticeable with the cobalt versions of the granite glazes, which unpredictably ranged from blue to almost green in color. It was discovered through controlled testing that iron + cobalt combinations will produce a green glaze if there is enough iron present and that the hue can be controlled to a certain extent.

As of the end of 2025, the first round of millings, Batch 1, is nearly used up (no raw material is left), necessitating the creation of more processed granite.

The goal of production for the "second" batch, Batch 2, is to refine the milling process and to produce a larger quantity of material that can be mixed together entirely and thoroughly, making it consistent. Iron is naturally present in the granite and cannot be avoided, but larger production runs should even out its concentration variation in the rocks collected and processed. That's the hypothesis anyway.

Batch 2 was processed as follows:

• Mill calcined granite to pass through a 6mm sieve as described here.
• Ball Mill this output in a 5 liter porcelain vessel using 2.5 liters of charge and 2.5 liters of water for 8 1/4 hours. Sieve at 60M and dry, hold the residual aside. Repeat.
• When 2 liters of residual has been collected, mill that by itself with 2.5 liters of water for 12 hours. Sieve at 60M and dry, and discard any residual.
• Continue until about 15 kilos of dried material is processed. Then combine it altogether in a cauldron large enough to mix it thoroughly.

The table below shows the progression taken (starting 5/19/25 and finishing 6/5/21). Just under 14 kilos (~31 pounds) of dried crushed granite was produced.

Obviously things did not go accordingly to plan towards the end. Milling typically is performed over night and when the shut off timer broke, milling continued past 8 hours. So I called an audible and kept going since it was going to take some time to obtain a new timer.

A goal of the milling process is to capture as much of the granitic material as possible without over milling. Over milling can cause the loss of some important material, especially the feldspars, which are our fluxes. This said, what constitutes over milling in my situation is unknown.

The reprocessing the residual seems like good discovery as we are able to mill most of the charge in two passes, crushing and capturing the softer material in the first pass and then the harder material in the second pass. But when the timer broke, we achieved a similar result by just milling longer. Which is better? Without further testing I will assume the original plan is "safer".

Initial Glaze Tests

Just a few glaze tests have been performed to date using Batch 2.

A base glaze (no colorants) was formulated using 70% Batch 2, 30% Wollastonite and 4% Zinc Oxide. This typically produces a glassy version of the base granite glaze.

Tests on Nara 5 (Porcelain), Sedona Red and Sienna Red (Mt Vision sourced) test tiles are shown below.

If Batch 2 produced the same glaze as Batch 1 that would be a fine result, and this seems to be the case. See Base Glaze C in this article for a comparison.

The pictures do not do the glaze justice. It's a fine, fully melted, glassy glaze. A rich glaze is observed on all three tiles, with some micro-crystals forming on the iron based red clays. There is some crazing present that can be worked out.

The next test is a line progression using the same base recipe and Cobalt Carbonate. The progression is 0.25%, 0.5%, 0.75%, 1.0% and 1.25%. Most uses to date have been 0.5%.

This also produces results similar to those produced with Batch 1. Crazing persists. Refer to Base Glaze C in this article for one comparison and this article for others. What this test provides that is new are finer sample intervals between 0% and 1%.

The final test presented is a line progression starting with Test Tile 1 above, 0.25% Cobalt Carbonate and blending Red Iron Oxide in concentrations of 0%, 0.25%, 0.5%, 0.75% and 1.0%.

Results similar to Batch 1 are obtained once again. Crazing persists but decreases as iron is added. Test Tile 4, 0.25% Cobalt Carbonate and 1% Red Iron Oxide corresponds chemically to results shown here. Now that we know how to play with cobalt and iron combinations, the motivation for this test is to start to refine a new Green Celadon formula for 2026.