Teaching glaze calculation

About this pack

This pack is for anyone who teaches glaze calculation: a school, a studio class, a study group of working potters, or a potter learning alone. It has six sessions, and each one has its aims, what to prepare, a worked example with real materials, how to do it in Glazecalc, a worksheet and questions to check.

The order of the sessions follows a published study group: the 陶磁器釉薬技術研究会 (ceramic glaze study group) of the Ibaraki Prefectural Industrial Technology Center in Japan, which ran six sessions for 12 young potters in 2015 and reported them in its research report. Its follow-on course at the Kasama college of ceramic art set three tile assignments of 42, 90 and 78 glazes, each fired in oxidation and in reduction. Glazecalc is not connected with the centre or the college. We follow the order of their published outline, and cite it below; the text, examples and worksheets here are ours. Courses in other countries teach the same steps in nearly the same order: Bilecik Şeyh Edebali University's SCT305 in Turkey, Glendale Community College's ART 195 in the US, the Rhineland-Palatinate syllabus for ceramics schools in Germany, and 李堅萍's 2006 paper for teachers in Taiwan.

How to use it

• Each session takes about 2 hours. Sessions 4 and 5 need calculators, and a computer or phone with Glazecalc for each pair.
• The worked examples use materials sold in Japan and the figures in Glazecalc's standard library: 福島長石 (Fukushima feldspar), whiting (石灰石, calcium carbonate), a kaolin and silica (珪石). Use your own supplier's analysis where you have one: add it on the Materials page, and the numbers change, but not the method.
• The firing is SK 8, 1250 °C (2282 °F) nominal. Use the cone your class fires to; Firing to SK cones explains SK numbers.
• To print a worksheet or a tile map, print this page from the browser. Saving recipes, blends and notes needs a free account.

Session 1: Materials and triaxial tiles

Aims

• Name the three jobs in a glaze: glass former, flux and stabilizer.
• Sort 4 common materials by the oxides they bring.
• Set up a triaxial blend, and read its tiles as a map.

Prepare

• A bag each of 福島長石, whiting, a kaolin and silica, or the materials your class has.
• The triaxial tile map below, printed, and 21 test tiles per student for each triangle.
• Fired tiles from an earlier triaxial, if you have them, to pass round.

Worked example

Each material's analysis, from Glazecalc's library, gives the oxides it leaves after firing. Loss on ignition (LOI) is what burns away.

Feldspar alone melts at SK 8 into a stiff glass; whiting alone does not melt; silica alone does not melt. Mixed, they melt far better than any one of them. A triaxial blend shows where: the tile map below steps the three in 20% steps, with 10% kaolin in every tile.

In Glazecalc

1. Open each material on the Materials page, or its public page, such as Fukushima Feldspar. Note which oxides it brings.
2. Make three recipes: 福島長石 90 with kaolin 10, whiting 90 with kaolin 10, and silica 90 with kaolin 10. On Blends, choose Triaxial (3 recipes) and 6 rows: 21 tiles, the same as the map below.
3. Choose the batch for each tile and Print what to weigh. After the firing, save the tiles you like as recipes from the drawing of the tiles.

Worksheet

Check

• Which material brings most of the alumina, and why does a glaze need it?
• Why does the whiting corner of the triangle not melt on its own?
• Where on the triangle did your tiles melt best, and which corner was that nearest?

Session 2: The chemistry you need

Aims

• Read a chemical formula: SiO₂, CaO, Al₂O₃, CaCO₃.
• Work out a molar mass from atomic weights.
• Turn grams into moles, and back.

Prepare

• A periodic table for each student, and calculators.
• The atomic weights in the table below (IUPAC's standard values, as Glazecalc uses them).

Worked example

A formula counts atoms. SiO₂ is one silicon and two oxygen, so one mole of it weighs 28.085 + 2 × 15.999 = 60.083 g. Al₂O₃ is 2 × 26.982 + 3 × 15.999 = 101.961 g. CaO is 56.077 g, and whiting, CaCO₃, is 100.086 g; Glazecalc's library rounds it to 100.08.

A mole is a count, like a dozen: the same number of molecules whatever they are. Moles are grams ÷ molar mass:

• 25 g of silica is 25 ÷ 60.083 = 0.416 mol of SiO₂.
• 20 g of whiting is 20 ÷ 100.08 = 0.200 mol. In the kiln it loses its carbon dioxide (CO₂, 44.009 g/mol) and leaves 0.200 mol of CaO, which weighs 0.200 × 56.077 = 11.2 g. That loss is its LOI: 43.97%.

In Glazecalc

1. Open Whiting and Silica: each page gives the formula and its weights.
2. Check one molar mass by hand against the page.

Worksheet

Check

• How many grams does one mole of SiO₂ weigh, and how is that worked out?
• Why does 20 g of whiting leave less than 20 g in the fired glaze?
• Why do we compare oxides by moles, not by grams?

Session 3: Reading a unity formula

Aims

• Read the three columns of a unity formula (ゼーゲル式, the Seger formula): fluxes, stabilizer, glass former.
• Find a glaze on a silica–alumina chart, and know what the chart can and cannot tell you.
• Colour a ready-made glaze in steps, and record it.

Prepare

• The unity formula below, written up large.
• A ready-made transparent glaze, and a colorant such as red iron oxide (弁柄).

Worked example

The glaze of session 4, 福島長石 45, whiting 20, kaolin 10 and silica 25, has this unity formula (Glazecalc's figures, rounded):

• The fluxes add up to 1: that is the rule of the formula. Each other number is moles for each mole of fluxes.
• Silica to alumina is 3.45 ÷ 0.43, about 8: a typical glossy glaze.
• Calcium is most of the flux, as in most Japanese lime glazes (石灰釉).

A chart with silica across and alumina up places each glaze as a point. Seger's tests and Stull's 1912 chart are of this kind, and a 釉性状図 maps a class's own fired tiles onto it. Glazecalc's recipe page draws the Stull chart, and this glaze sits in its bright region. Stull's regions come from his own fluxes and firing: a guide, not a prediction. The tile grid below lets a class draw its own chart at its own firing.

To colour a ready-made glaze whose recipe you do not know, Glazecalc cannot work out its formula. Add the colorant on top of the 100 instead: 100 g of the dry glaze with 1, 2 and 3 g of iron oxide, one tile each, fired in oxidation and reduction.

In Glazecalc

1. On the recipe page, enter the recipe above and read its Unity formula and Stull chart.
2. Change the silica from 25 to 35, and watch the point move to the right.
3. For a base whose recipe you have, make the colorant an addition and use a line blend on Blends to step it.

Worksheet

Check

• In a unity formula, which column adds up to 1?
• What is this glaze's silica to alumina ratio, and what does it suggest?
• Why can a point in the chart's bright region still give a matte tile in your kiln?

Session 4: From recipe to unity formula

Aims

• Work out a unity formula from a recipe by hand.
• Check it against Glazecalc's own working.

Prepare

• Worksheet 4, the four analyses from session 1, and calculators.

Worked example

The recipe: 福島長石 45, whiting 20, EPK kaolin 10, silica 25. Take each amount as grams. For each material, grams × the oxide's percent ÷ 100 ÷ its molar mass gives its moles. The table gives the main oxides; Glazecalc counts the traces too (the kaolin's iron and titanium, the feldspar's magnesium).

The fluxes add up to 0.0518 + 0.0275 + 0.2041 + 0.0014 (magnesia) = 0.2847 mol. Divide every total by it:

• K₂O 0.0518 ÷ 0.2847 = 0.18; Na₂O 0.10; CaO 0.72.
• Al₂O₃ 0.1229 ÷ 0.2847 = 0.43.
• SiO₂ 0.9820 ÷ 0.2847 = 3.45.

That is the formula of session 3.

In Glazecalc

1. Enter the recipe on the recipe page.
2. Open How the unity formula is worked out. It shows these steps with this recipe's numbers: grams, the moles of each oxide in each material, the totals, and the division by the fluxes.
3. Compare each line with your worksheet. Then change one amount and work out only the column it changes.

Worksheet

Check

• Why does any batch size give the same unity formula?
• Which oxides does the kaolin bring, and which of them matters most here?
• Why is whiting's LOI left out of the formula?

Session 5: From unity formula to recipe

Aims

• Choose materials for a target formula, and work out how much of each.
• Use Solve for a target, and say how near it comes.

Prepare

• The target below, worksheet 5, and calculators; Glazecalc with an account.

Worked example

The target: 0.2 K₂O, 0.1 Na₂O, 0.7 CaO, 0.4 Al₂O₃ and 4.0 SiO₂. Take the materials in this order, each for the oxide only it can bring:

1. The feldspar, for the potash and soda. One unit of 福島長石 (536.8 g, the weight that holds one mole of fluxes) brings 0.94 mol of K₂O and Na₂O. The target needs 0.3, so 0.3 ÷ 0.94 = 0.319 units: 171.4 g. It also brings 0.015 CaO, 0.328 Al₂O₃ and 1.865 SiO₂.
2. Whiting, for the rest of the calcium. 0.7 − 0.015 = 0.685 mol × 100.08 = 68.6 g.
3. Kaolin, for the rest of the alumina. 0.4 − 0.328 = 0.072 mol. EPK kaolin holds one mole of Al₂O₃ in 272.6 g, so 19.6 g, which also brings 0.149 SiO₂.
4. Silica, for the rest. 4.0 − 1.865 − 0.149 = 1.986 mol × 60.083 = 119.3 g.

The batch is 379.0 g. As percent: 福島長石 45.2, whiting 18.1, kaolin 5.2, silica 31.5. The feldspar sets the ratio of potash to soda, so the recipe gives 0.195 K₂O and 0.104 Na₂O: near the target, not on it.

In Glazecalc

1. Open Solve for a target, choose A unity formula, and type the target.
2. Under The materials, give it 福島長石, whiting, EPK kaolin and silica, and choose Solve.
3. It gives 福島長石 44.9, whiting 18.2, kaolin 5.6 and silica 31.3, and a table of how near each oxide comes. Compare it with your hand working: the solver counts every trace, and finds the best fit the four materials allow.
4. Open as a new recipe, and check it on the recipe page.

Worksheet

Check

• Why is the feldspar worked out first?
• Why can these four materials not give exactly 0.2 K₂O and 0.1 Na₂O?
• What would you add to the materials to bring in magnesia?

Session 6: The clay body and glaze defects

Aims

• Measure a body's shrinkage and water absorption.
• Name the common defects, their usual causes and a first fix.
• Know which defects make ware unsafe.

Prepare

• Test bars of your clay body, marked 100 mm apart while wet, fired and unfired.
• Tiles with defects, if you have them: crazing, crawling, pinholes, running.

Worked example

• Shrinkage: (marked length − fired length) ÷ marked length × 100. Marks 100 mm apart that measure 88 mm after firing: 12% shrinkage.
• Water absorption: (weight after soaking − dry weight) ÷ dry weight × 100. A fired bar of 50.0 g that weighs 50.6 g after soaking: 1.2%.

Glaze fit and defects explains each one, with the ice-water test for crazing. Do not use crazed or shivered ware for food or drink.

In Glazecalc

1. On the recipe page, read the Limits panel for a glaze of yours, and its calculated expansion.
2. Copy the recipe, swap some feldspar for whiting, and compare the two: which way did the expansion move?

Worksheet

Check

• Which defect comes from a glaze that shrinks more than its body, and what is a first fix?
• Why must shivered ware never be used for food?
• Can a calculation tell you a glaze is safe for food? What can?

Tile maps to print

Triaxial map: 42 glazes

Two triangles of 21 tiles, from Blends' layout with 6 rows: tile 1 at the top (feldspar), the bottom row from whiting (left) to silica (right). Triangle II is triangle I with 1 g of red iron oxide on top of each 100 g, which turns toward celadon (青磁) in reduction. Fired in oxidation and reduction, that is 84 tiles. Amounts are grams per 100 g.

Seger grid: 24 glazes

Every tile has the same fluxes, 0.2 KNaO and 0.8 CaO, from 福島長石 and whiting. Alumina steps down the grid and silica across it. The recipes were worked out by the method of session 5; Glazecalc's unity formula of each comes within 0.01 of its alumina and 0.06 of its silica. Mark each tile on a chart of silica against alumina, and draw your own regions: matte, glossy, underfired, running.

Glazecalc's Blends makes triangles and squares of 3 to 66 tiles, not this grid; open any tile as a recipe on the recipe page and print its batch list from there.

Sources

• 吉田博和, 吉田彩美, 寺門秀人, 陶磁器釉薬技術研究会, 茨城県工業技術センター研究報告 第44号 (2015): the six sessions.
• 茨城県工業技術センター, 陶磁器釉薬技術研究会, 平成27年 成果集: the sessions' dates and subjects.
• 茨城県産業技術イノベーションセンター, 陶磁器用釉薬に関する研修, 令和元年 成果集: the follow-on course's assignments of 42, 90 and 78 glazes in oxidation and reduction.
• 岡部弘文, 釉薬計算プログラムの開発, 茨城県工業技術センター研究報告 vol33 (2005): the centre's own calculation program.
• Bilecik Şeyh Edebali University, SCT305 course information: the Seger formula in weeks 5, 6 and 8.
• 李堅萍, 國立臺北教育大學學報 19(2), 2006, pp. 51–68: glaze calculation for school teachers in Taiwan.
• Glendale Community College, ART 195 course outline: the unity formula in weeks 5 to 7 of 15.
• Ministerium für Bildung Rheinland-Pfalz, Lehrplan Fachschule Keramikgestaltung (2003), Lernmodul 6, Glasuren.
• 佐賀県陶磁器原料株式会社, 釉薬原料の化学分析表: the analysis of 福島長石 in Glazecalc's library.
• IUPAC, standard atomic weights, as Glazecalc's chemistry uses them.