Why is mathematics important in pottery? A pot begins as a soft three-dimensional object, loses water while drying, changes again during firing, and receives a glaze mixed from a recipe. Every stage involves ratios, measurements and uncertainty. A maker who wants a particular fired size must reason backwards. A studio that wants repeatable batches must normalise recipes, record masses and compare evidence across firings.
This article explores pottery firing shrinkage, clay volume and glaze recipes as mathematical systems. It does not supply a production glaze formula or kiln schedule. Real clays and glazes can create dust, fumes, food-contact and firing hazards; use supplier instructions, appropriate controls and experienced kiln supervision.
Find Your Reading Route
- Understand linear shrinkage
- Move from length to volume
- Scale glaze recipes
- Read firing evidence
- Plan a student investigation
Linear Shrinkage: Start With a Reference Length
Suppose a clay test bar is marked 100.0 mm apart while plastic. After drying, the marks are 94.0 mm apart. After firing, they are 88.0 mm apart. The total linear shrinkage relative to the plastic state is:
(100.0 − 88.0) / 100.0 × 100% = 12.0%
The drying shrinkage relative to the plastic state is 6.0%. The firing shrinkage is often calculated relative to the dry length:
(94.0 − 88.0) / 94.0 × 100% ≈ 6.38%
These percentages have different denominators. Saying “6% drying plus 6.38% firing equals 12.38% total” is wrong because the second change acts on the already smaller dry bar. Sequential changes multiply their scale factors.
The dry scale factor is 94/100 = 0.94. The fired-from-dry factor is 88/94 ≈ 0.93617. Their product is 0.88, so total shrinkage is 12%. This is the same proportional reasoning used in compound discounts, population changes and repeated dilution.
Back-calculating a making size
If a tested clay body has a fired linear scale factor of 0.88 under a specific process, a maker seeking a 176 mm fired diameter would estimate a plastic diameter of 176/0.88 = 200 mm.
That answer is conditional. It assumes the new piece shrinks like the test bar, the forming method and moisture are comparable, and the firing delivers similar heatwork. It is a starting estimate, not a guarantee.
Percentage points versus percentage change
Suppose one tested batch shrinks 12% and another 13%. The second is one percentage point higher. Relative to 12%, it is about 8.33% greater because (13 − 12)/12 × 100% = 8.33%. Keeping these two comparisons distinct prevents inflated claims.
Why Volume Change Is Much Larger Than Linear Change
If shrinkage is uniform in every direction, each length is multiplied by the same scale factor s. Area is then multiplied by s², and volume by s³.
For s = 0.88, the area factor is 0.7744 and the volume factor is 0.681472. The modelled volume decrease is therefore about 31.85%, not 12%.
This surprises students because our intuition notices one edge at a time. Geometry reminds us that a three-dimensional object changes in three independent directions.
Worked cylinder example
Imagine a simple cylindrical cup with plastic outside diameter 100 mm and height 120 mm. Ignore wall thickness for the moment. Its bounding-cylinder volume is π × 50² × 120 = 300,000π mm³, about 942,478 mm³.
With uniform scale factor 0.88, the fired diameter is 88 mm and height 105.6 mm. The new bounding volume is the original volume multiplied by 0.88³, about 642,353 mm³. The ratio agrees whether we cube the scale factor or recalculate from dimensions.
This is not the cup’s clay volume, because the interior void and wall profile were ignored. The example is useful precisely because it distinguishes a bounding volume from material volume.
Unequal shrinkage changes shape
Real pieces may not shrink equally in every direction. Forming direction, particle alignment, wall thickness, drying gradient, support and firing can matter. Suppose diameter scales by 0.89 while height scales by 0.86. A plastic height-to-diameter ratio of 1.20 becomes 1.20 × 0.86/0.89 ≈ 1.160. The piece becomes relatively squatter.
A student should therefore record dimensions along labelled directions, not report one shrinkage number as if it describes every geometry. The mean can be useful, but it must not erase orientation.
Clay Volume, Wall Thickness and Mass
Geometry becomes practical when estimating clay amount. Model a straight-sided hollow cylinder with outside radius R, inside radius r and height h. Its wall-and-base volume needs a precise shape definition. For a simple open cylinder with a bottom thickness b and vertical walls, one model is:
V = π(R² − r²)(h − b) + πR²b
The first term represents the annular wall above the base; the second represents the solid bottom disc. Multiply volume by wet clay density to estimate mass.
If dimensions are in centimetres and density in grams per cubic centimetre, mass emerges in grams. The calculation is only as good as the geometry. A curved bowl, foot ring, handle and trimmed base require better approximations or direct measurement.
Thin-wall sensitivity
When wall thickness t is small relative to radius, a cylindrical wall volume is approximately circumference × thickness × height, or 2πRth. This approximation helps intuition: doubling thickness roughly doubles wall clay volume.
But “roughly” matters. The exact annular area is π[R² − (R − t)²] = π(2Rt − t²). The approximation drops the t² term. Students can calculate the relative error and decide whether the shortcut is acceptable.
Density is not one permanent constant
Plastic clay contains water and air; drying and firing change mass and volume. A density measured for one preparation is not automatically transferable. Record how volume and mass were obtained. NIST’s SI guidance on quantities and units lists volume in cubic metres and mass in kilograms; smaller coherent units may be more convenient, but conversions must remain explicit.
Glaze Recipes: Parts, Normalisation and Batch Mass
A common studio recipe expresses dry ingredients as parts by mass. Consider a fictional teaching recipe:
- material A: 45 parts;
- material B: 25 parts;
- material C: 20 parts;
- material D: 10 parts.
The total is 100 parts, so each number is also its mass percentage of the dry base recipe. To make a 750 g dry batch, multiply each fraction by 750 g: 337.5 g, 187.5 g, 150 g and 75 g.
If a recipe instead totals 125 parts, first normalise. A 50-part ingredient represents 50/125 = 0.40 of the dry batch. In a 750 g batch, it contributes 300 g.
Additions outside the base 100
Some studio records list a base recipe totalling 100, followed by an addition such as 2 parts of colourant. Then the dry total is 102 parts unless the record explicitly defines a replacement. For a 1,020 g combined batch, the base ingredients together make 1,000 g and the addition makes 20 g.
Ambiguous notation is a production risk. A good record says whether an addition is included in, replaces part of, or sits outside the base total.
Water is not automatically part of the dry recipe
If 800 g of water is added to 1,000 g of dry materials, the wet slurry mass is 1,800 g. But the dry recipe percentages still refer to the 1,000 g dry total unless another basis is stated.
This is a denominator lesson. “44.4% water by total wet mass” and “80 g water per 100 g dry material” describe the same example using different bases. Neither phrase should be shortened to “80% water” without explanation.
Scale factors preserve ratios
To scale a 100-part recipe to a desired dry mass B, use scale factor k = B/100. Multiply every part by k. If the original total is T rather than 100, use k = B/T.
After rounding to the available balance resolution, sum the rounded masses. The difference from the target is a rounding discrepancy. It can be assigned deliberately rather than hidden.
Specific Gravity as a Process Number
Slurry specific gravity is the ratio of slurry density to a reference water density under stated conditions. In a simplified room-temperature studio calculation, a tared 100 mL cup holding 148 g of slurry gives an approximate density of 1.48 g/mL and a numerical specific gravity near 1.48.
This number can support consistency checks when the method is standardised. It does not reveal particle-size distribution, chemistry, suspension stability or application behaviour by itself.
Air bubbles, an unlevel fill, material on the cup exterior and temperature introduce error. Repeating the measurement and keeping the raw masses makes the result more defensible.
Water adjustment by mass balance
Suppose 2,000 g of slurry contains 1,200 g of dry solids and 800 g of water. Adding 100 g water produces 2,100 g total with the same 1,200 g solids. The dry-solids mass fraction changes from 60.0% to about 57.14%.
This is not enough to predict the new specific gravity without density assumptions, but it correctly tracks material amount. A mass balance is often the safest first step before a more detailed model.
Heatwork Is Not One Temperature Number
Firing is affected by both time and temperature. The Edward Orton Jr Ceramic Foundation’s pyrometric cones FAQ explains that cones indicate heatwork, that heating rate influences their endpoint, and that large or self-supporting cones near the ware help assess whether enough heatwork was delivered.
This matters mathematically because a kiln display’s peak temperature is one data point, not a complete description of the ware’s thermal history. Heating rate, hold time, atmosphere, load arrangement and location can change outcomes.
Orton’s TempCHEK explanation describes calibrated ceramic pieces that shrink in response to heatwork and stresses that their equivalent temperature is not simply the actual temperature. That is a beautiful example of a measurement model: an observable dimension is converted through a calibration created under defined conditions.
A kiln map is a spatial data set
Place test evidence at several documented locations. Record shelf, height, front/back position and nearby load. A 3 × 3 map can show whether shrinkage or cone response has a spatial pattern.
The mean summarises the centre of the data. Range and standard deviation describe spread. A heat map shows location. None should replace the raw observations. If one corner repeatedly differs, averaging everything may conceal the information the studio most needs.
Shrinkage is an outcome, not a perfect thermometer
Orton explains that ceramic materials densify and shrink with heatwork, but a piece’s shrinkage also depends on its composition and process. A calibration developed for one controlled material cannot simply be transferred to another clay body.
This is an important scientific habit: the instrument, calibration and specimen form one system. A number becomes meaningful through traceability, not appearance.
Worked Example: Plan a Fired Bowl and a Test Batch
A student’s tested clay has plastic-to-dry scale factor 0.95 and dry-to-fired scale factor 0.93. The desired fired outside diameter is 167.4 mm.
The total scale factor is 0.95 × 0.93 = 0.8835. The estimated plastic diameter is 167.4/0.8835 ≈ 189.47 mm. Given measurement and process variation, specifying 189.468… mm would be false precision. A sensible making target might be 189.5 mm, followed by testing.
The fictional glaze base has 40, 30, 20 and 15 parts, totalling 105. For a 630 g dry base, k = 630/105 = 6 g per part. The masses are 240 g, 180 g, 120 g and 90 g. They sum to 630 g.
Now add a fictional 3 parts outside the base. On the base scale, that addition is 18 g, making 648 g combined dry material. Writing “3% addition” would be slightly misleading because 18/648 is about 2.78% of the final combined mass. “3 parts per 100-ish base parts” and “2.78% of combined dry mass” use different denominators.
Uncertainty and Repeatability
Suppose five fired test bars show total linear shrinkages of 11.6%, 11.9%, 12.1%, 12.0% and 12.4%. The mean is 12.0%. The range is 0.8 percentage points.
A product target calculated with exactly 12.0% shrinkage should not be presented as exact. Variation can come from specimens, measuring, moisture, placement and firing. The student can propagate a plausible scale-factor interval or run a Monte Carlo simulation to show the distribution of final size.
Measurement resolution also matters. If marks are only 50 mm apart, a 0.5 mm reading uncertainty represents 1% of that length. Longer marked distances reduce relative reading error, provided the specimen stays practical and representative.
Common Misconceptions
“Drying shrinkage plus firing shrinkage equals total shrinkage”
Only if both percentages use the same original reference, which they often do not. Multiplying scale factors is safer.
“Twelve percent linear shrinkage means twelve percent volume loss”
Not for uniform three-dimensional scaling. Volume follows the cube of the linear factor.
“A cone number is a temperature setting”
Cones respond to heatwork. Heating rate and cone type affect endpoint interpretation.
“A recipe totalling 100 is automatically a chemical analysis”
No. It is usually a batch proportion by mass. Chemical composition requires information about each raw material.
“More water simply lowers specific gravity by the same percentage”
The relationship depends on masses, volumes and densities. Use a mass balance and measure rather than guessing.
A Safe Mathematical Investigation
Use supplier-prepared, suitable materials under experienced supervision. Avoid generating or inhaling dust. Let the mathematical work begin with existing studio data or synthetic values if practical controls are unavailable.
Create several identical labelled test bars. Mark a sufficiently long gauge length. Record plastic, bone-dry and fired dimensions with the same instrument and direction. Photograph placement and record the firing schedule and cone evidence.
For recipes, use fictional non-hazardous components such as coloured beads or beans to practise normalisation and batch scaling. The mathematics transfers without turning a classroom into a glaze laboratory.
Parents and students can focus on explanation: What is the reference length? Which factors multiply? What does the recipe total mean? Which quantities were directly measured and which were inferred?
Did You Know?
A small difference in linear scale can become a larger difference in volume. A clay body scaling to 0.90 in each direction retains 72.9% of its original bounding volume, while one scaling to 0.88 retains about 68.15%. Only two percentage points separate the linear factors, but the volume factors differ by about 4.75 percentage points.
Recipe normalisation is also a form of vector scaling. The ingredient list has a direction—the ratios among components—and batch size changes only its magnitude. That same idea appears in computer graphics, chemistry, finance and data preprocessing.
Surface area changes too
Uniform linear scale factor s multiplies every surface area by s². If a piece scales to 0.88, an ideal corresponding surface scales to 0.7744 of its former area. That matters when students compare glaze quantity on geometrically similar test pieces: halving every length does not halve surface area; it reduces area to one quarter and volume to one eighth. Real glaze pickup also depends on slurry properties, porosity and application, so geometry supplies a baseline rather than a guaranteed consumption rate. Recording dipped area and wet mass gain can turn that baseline into a testable studio question.
A labelled sketch should accompany the calculation so readers know whether the measured area includes the foot, rim, interior and exterior. Clear boundaries are always as important as correct arithmetic.
Frequently Asked Questions
Can one shrinkage percentage predict every form?
No. It is evidence for a defined material and process. Shape, thickness, forming, drying and firing can change the outcome.
Why use a 100 mm test-bar mark?
It makes percentage interpretation convenient: each millimetre of loss corresponds to one percent of the original gauge length. Other lengths work if the denominator is recorded.
Is glaze recipe mathematics the same as glaze chemistry?
No. Batch proportions say how much raw material is weighed. Chemistry considers oxides and material analyses, which requires additional data and expertise.
Does a kiln thermocouple replace cones?
They provide different evidence. Follow the kiln and material guidance; do not treat one measurement as a complete description of heatwork.
What should a student learn first?
Master percentage change with a stated reference, then sequential scale factors, area and volume scaling, and recipe normalisation.
Useful Next Reading
Read Why Mathematics? | Cooking, Baking and Recipe Scaling for another careful look at ratios and batch size. Continue with Why Mathematics? | Tolerance Stack-Ups, Fits and Manufacturing Assemblies to see how small dimensional variations combine in production. The Mathematics Learning Hub provides wider school and pathway context.
Final Perspective
Pottery gives mathematics a wonderfully physical life. Percentage change becomes a smaller fired form. Cubic scaling explains why volume changes faster than one edge. Normalisation turns a recipe into any batch size. Statistics transforms scattered test pieces into evidence.
The strongest maker is not the person who expects every equation to be perfect. It is the person who uses a clear model, measures carefully, records context and lets the next test improve the next decision.
A Practical Investigation Studio
Use synthetic data and harmless classroom materials. These investigations expose the mathematics and its limits; they do not authorise kiln, glaze-safety or studio-production decisions.
Investigation 1: Measure a test bar
Mark a fictional 100.0 mm wet test bar, then record dry and fired lengths. Calculate drying, firing and total linear shrinkage with the correct reference length. Use synthetic or openly released teaching data. Begin by naming the response variable, input variables, units and prediction before calculating. Preserve raw values and unrounded intermediates in a table, then make one graph whose axes communicate the model clearly. Add an independent hand estimate and a dimensional or limiting-case check. Deliberately introduce one unit error and describe the numerical signature rather than merely correcting it. Label every quantity observed, assumed, fitted or derived. Record the model domain and one condition that would require a richer model. Ask a peer to reproduce the result from the written procedure without seeing the answer. If the reproduction differs, locate whether the ambiguity arose in definitions, units, rounding or software defaults. End by explaining the result to a younger student in three sentences. This remains a classroom calculation, not professional validation or a safety, production or security decision.
Investigation 2: Diameter planning
Back-calculate a making diameter for a target fired bowl. Compare sequential shrinkage factors with the incorrect practice of simply adding percentages. Create a data dictionary before entering a spreadsheet or script. State how each row was obtained, which values are exact by definition and which are measurements or simulated observations. Calculate once with full precision and once with premature rounding, then compare the final difference. Vary the principal input across at least five sensible values and look for linearity, curvature or instability instead of reporting only two endpoints. Keep signed residuals if a model is fitted, because absolute errors hide direction. Include a graph and a small results table that another person can audit. Write a one-paragraph uncertainty note distinguishing input uncertainty from model-form limitations. A peer should be able to reconstruct one row from the formula and raw data alone. State what evidence would falsify the interpretation. Do not present the exercise as a certified engineering, manufacturing, metrology or cryptographic assessment.
Investigation 3: Volume effect
Model uniform shrinkage in three dimensions. Cube the linear scale factor and explain why volume loss is not equal to linear loss. Treat the investigation as a comparison of hypotheses, not a hunt for an attractive number. Write a baseline model and at least one plausible alternative, then predict where their outputs should diverge. Hold all unrelated inputs fixed while varying the chosen factor. Preserve coordinate, sign, phase, size-weighting or bit conventions beside the data. Use a ratio, residual or normalised error that has a declared denominator. Repeat the calculation with that denominator changed and explain why the percentage changes. Check one extreme case where the answer should approach zero, unity or another known limit. Make the graph before writing the conclusion and note any outlier without deleting it. Separate repeatability of one prepared sample from coverage across positions, times or populations. The final claim should match the observed range and must not be extrapolated into real operational approval.
Investigation 4: Unequal directions
Assign different shrinkage to height and diameter. Compute the fired proportions and identify the risk of assuming isotropy. Plan the computation so that errors leave visible fingerprints. Save the raw input, transformed input, intermediate result and final result in separate columns or variables. Insert one deliberate sign reversal, one missing observation and one hidden reset, each in a separate copy, and document how the plots or checks respond. Compare an analytical shortcut with the more complete calculation over a range where the shortcut is expected to fail. Report the largest absolute difference and where it occurs. Include conservation, balance, boundedness or monotonicity checks appropriate to the topic. If a fitted curve is used, inspect residual clusters and changing spread instead of relying on one fit statistic. Have a peer review only the assumptions first, then the arithmetic. Keep the conclusion educational and conditional on the fictional data.
Investigation 5: Recipe normalisation
Convert five fictional glaze parts into a 100-part recipe. Distinguish parts by mass from a percentage of the final wet slurry. Build a sensitivity map rather than changing several settings at once. Choose a baseline, vary one input downward and upward, and express output change in both original units and a dimensionless ratio. Then select a second input and repeat. If interactions seem likely, test a small two-dimensional grid and identify where the one-factor interpretation breaks down. State why the chosen ranges are plausible for the teaching model. Preserve failed or surprising runs and annotate them. Compare computational resolution or sample length at three levels so numerical artefacts are not confused with physical behaviour. Provide a hand estimate for the order of magnitude. Finish with a decision table using cautious language—stable, sensitive or unresolved—rather than safe or unsafe. Classroom evidence cannot authorise a product, structure, process or security control.
Investigation 6: Batch scaling
Scale a dry recipe from 100 parts to 750 g. Keep water addition outside the dry-batch total unless the recipe explicitly includes it. Make uncertainty visible at every stage. Assign each input a source and a plausible interval, then identify which inputs are correlated rather than automatically treating them as independent. Propagate uncertainty with a simple high-and-low calculation or transparent simulation, and retain the seed or full synthetic samples for reproduction. Compare the spread caused by measurement variation with the shift caused by changing the model assumption. Plot both if possible. Explain why more decimal places do not narrow an interval supported by weak inputs. If a result lies near a fictional decision boundary, rewrite the conclusion to reflect the chance of crossing it. Ask a peer to challenge the largest assumed uncertainty and rerun the analysis. Report the conclusion as evidence about the model, never as professional certification.
Investigation 7: Specific-gravity check
Use slurry mass and container volume to calculate a fictional density. Treat the result as a process observation, not proof of glaze chemistry. Audit representation choices. Recalculate after changing the coordinate origin, phase wrapping convention, histogram bins, mesh labels or binary mapping while preserving the underlying situation. A correct physical or probabilistic conclusion should change only where the definition genuinely changes. Show one example where a visual choice exaggerates or hides a pattern. Use identical axes for fair comparison and include sample count or resolution in captions. Test an invariant such as total mass, probability, force, energy sign or average ratio. Where values are averaged, identify whether the mean is number-, mass-, volume-, time- or state-weighted. Ask another student to explain the plot without reading the conclusion; revise labels if their interpretation differs. The display supports reasoning but cannot replace domain review.
Investigation 8: Heatwork comparison
Compare cone observations at two kiln locations. Do not translate a cone number into a single universal temperature without the firing-rate context. Separate verification from validation. First verify arithmetic or code against a case with a known answer, including units and at least one manually calculated row. Next ask whether the teaching model represents the intended phenomenon and list the omitted mechanisms. Refine numerical resolution or increase synthetic sample length and record whether the chosen quantity stabilises. A stable answer can still arise from an unrealistic model, so compare with an alternative assumption or openly documented benchmark. Record software version, solver or library settings and stopping rules. Avoid tuning settings after seeing the desired result; predeclare the comparison instead. Summarise numerical error, input uncertainty and model-form uncertainty separately. No classroom agreement with a benchmark should be described as product or system validation.
Investigation 9: Kiln map
Create a 3 by 3 grid of fired test-tile shrinkage. Use mean, range and position to discuss spatial variation. Design the investigation to reveal dependence over position, time or sequence order. Keep the original ordering, then compare with a deliberately shuffled copy and explain which statistics change. Examine at least two lags, locations or neighbourhood scales rather than only the overall average. Plot local values alongside the aggregate so compensating errors are visible. State whether successive observations are plausibly independent and why that matters to the calculation. If repeated measurements share one specimen, source or initial state, do not count them as independent coverage. Add a block or subgroup analysis and note any drift. Preserve the random seed for simulated data but use more than one seed before generalising. The conclusion should describe detected structure without claiming causation from the pattern alone.
Investigation 10: Uncertainty budget
Combine calliper resolution, mark placement and specimen variation. Report an interval and avoid false tenths of a percent. Create an adversarial edge-case test. Identify the smallest, largest, most symmetric and most irregular inputs allowed by the classroom model, predict the expected behaviour, and then compute it. Include zero or a limiting value only when mathematically meaningful. Watch for division by a small number, logarithms of invalid values, wrapped angles, negative geometry, impossible probabilities or solver singularities. Replace silent software errors with explicit checks and explanatory messages. Compare the edge cases with the ordinary baseline on the same scale and describe why a method that works in the middle may fail near a boundary. Record the first assumption that breaks. The exercise is about model literacy; it is not permission to explore hazardous physical extremes.
Investigation 11: Studio record
Archive clay batch, dimensions, drying conditions, firing schedule, cone evidence and recipe masses. Use compatible tested materials and qualified kiln supervision for real firing. Prepare a compact reproducibility package: raw synthetic data, formulas or code, version information, one chart, a results table and a short read-me file. Give every file and column a meaningful name and avoid values copied manually between tools. Re-run from a clean state and confirm that outputs are regenerated rather than cached. Ask a peer to alter one declared input and predict the direction of change before execution. Compare the prediction with the result and investigate any disagreement. Add a limitations section naming data coverage, numerical resolution, model assumptions and the next useful measurement. Archive corrections instead of erasing them so the reasoning trail remains visible. Conclude with what the calculation demonstrates, what it does not demonstrate and which qualified professional or authoritative standard would govern a real application.