Glaze & Recipe field guide
A Glaze Testing Workflow That Keeps Useful Records
A repeatable route from a 100-gram test batch to labelled tiles, application notes, firing records and a production decision.
A glaze test is not just a coloured tile. It is a small production record: recipe, batch, water, application, clay body, kiln position and heatwork. Keep those parts together and a promising surface can be reproduced—or rejected for a clear reason.
Write the question before mixing
“See what happens” produces attractive accidents and poor records. A useful test asks one bounded question: Does 2%, 4% or 6% rutile change this base at cone 6? Does the same glaze craze less on Body B? Does a slower cool alter the surface? Write the question at the top of the sheet and decide what will remain constant.
If several variables change together—different body, thickness, firing and colorant—you cannot tell which one caused the result. A line blend or grid can vary two ingredients deliberately, but the design should be planned before the cups are filled.
Normalize and preserve the original recipe
Keep a source copy exactly as received, including author, publication, date, intended cone and atmosphere. Work from a separate normalized version. A base recipe expressed as percentages is easy to scale because its base totals 100. Colorants are commonly recorded as additions to that base rather than forced inside the 100 total.
For a 100 g base batch, the percentage number is also the gram weight. A recipe containing 40% material A, 30% B, 20% C and 10% D becomes 40 g, 30 g, 20 g and 10 g. A 3% colorant addition adds 3 g, making 103 g of dry material in the bucket. Record the convention so another person does not normalize the colorant away.
Use a weighing sequence that can be checked
- Clean the scale area and use containers with legible test codes.
- Check the scale with a suitable reference weight and record its resolution.
- Weigh each material into a separate cup or use a check-off line before adding it to the batch.
- Record the actual weight when it differs from target; do not silently edit the recipe to match a mistake.
- Mix dry materials with appropriate dust controls. Review safety data sheets for every material.
Silica and many ceramic raw materials create hazardous respirable dust. Use wet-cleaning methods, local controls and suitable respiratory protection; do not sweep or blow dry powder through the studio.
Water is part of the test
“Add water until right” is not repeatable. Record the starting water mass or volume, mixing time, sieve, resting time and final specific gravity. Specific gravity describes slurry density, not viscosity. Two glazes at the same SG can apply differently because particle size, clay content, deflocculation and thixotropy differ.
When a test becomes a production glaze, establish a small control routine: mix fully, measure SG in a known container, run the same flow or viscosity check, then test application on a standard tile. Recheck after evaporation or material additions.
Make tiles that reveal more than colour
Use the intended clay body and firing. A good tile includes texture or a step so the glaze can be read thin and thick; a vertical face reveals movement; an unglazed area shows the body. Mark the code into the clay before firing. Iron-bearing and pale bodies should not share an anonymous tile.
Bisque schedule and surface preparation matter. Dust can cause crawling. Handling oil can create local defects. If production ware is once-fired, a bisque-fired test is not the same system. Match the process you intend to use.
Control application
Record whether the glaze was dipped, poured, brushed or sprayed. For dipping, note time and number of dips. For brushing, note coats and drying interval. Weighing a tile before and after application is a practical way to compare wet glaze pickup when consistent thickness matters.
A failed thick application does not necessarily condemn a glaze intended to be thin. Equally, a beautiful test tile does not prove the glaze will behave on a large bowl with a heavier application. Test the useful application window rather than one ideal spot.
Record the firing as heatwork
Write down the kiln, program, atmosphere, load, tile position and witness-cone result. The controller peak alone is incomplete. Orton’s cone FAQ explains that heating rate affects the temperature at which a cone bends; holds and atmosphere also affect ceramic response.
For a cooling test, preserve the complete schedule below peak. Crystalline and matte development, colour and defect healing can change during cooling even when the peak heatwork is similar.
Evaluate after the kiln excitement has passed
Record observations in neutral light and again after the tile is washed. Note surface, gloss, colour, opacity, movement, pinholes, blisters, crawling, crazing, cutlery marking and the interface at the edge. Photograph front, back, code and witness cones with a colour reference if the images will be compared over time.
Delayed crazing may not be visible at unloading. A production decision for functional ware can require thermal cycling, chemical durability or leach testing appropriate to the intended use. Calculated COE and colorant percentage do not establish food-contact safety.
Scale only after a confirmation batch
A successful 100 g test should normally be repeated before a 10 kg batch. Make a confirmation batch large enough to reveal mixing and application behaviour, then test it on the intended form. Material lot variation, water addition and mixing energy become more important at scale.
When scaling, calculate from the normalized master recipe—not from rounded gram weights written on the first test cup. Keep colorants and additions explicit. Label the production bucket with recipe revision, date, batch mass, SG and safety information.
Printable glaze test record
| Material | Base % | Target g | Actual g | Lot / note |
|---|---|---|---|---|
Water / mixing / sieve:
Fired observations:
Decision and next test:
A compact decision rule
- Reject when the result creates an unacceptable defect or safety concern in the intended process.
- Retest when the result is promising but a variable was uncontrolled or the record is incomplete.
- Confirm when the result meets the visual target and needs a repeated batch or application window.
- Adopt only with a named revision and defined clay, application and firing range.
Sources and further reading
Use these calculations with the procedure
Editorial note
Found a source or calculation that needs correction?
See how formulas are selected, how changes are recorded, and how to send a reproducible correction in the ClayCalc methodology.
Continue reading