What Is Crazing in Ceramics? Definition & Food Safety Guide
A network of fine cracks in your favorite mug’s glaze does not mean it is ruined. It means the glaze and clay body are fighting each other at the molecular level. Understanding crazing is the difference between making food-safe functional pottery and creating decorative pieces that need a warning label.
By the Numbers
Crazing in Ceramics – What the Research Shows
Sources: Digitalfire Reference Library, ASTM C738 Testing Standards, Rhodes & Hopper Clay and Glazes
Crazing appears as a web of hairline cracks on the glazed surface of ceramic ware.
These cracks do not penetrate through the clay body itself.
Instead, they sit within the glassy glaze layer, creating a pattern that potters either love or dread.
This guide covers the full spectrum of crazing – from the glaze chemistry that causes it, to the difference between intentional crackle glazes and structural failures, to the food safety implications that every functional potter needs to understand before selling or serving from crazed ware.
What Is Crazing in Ceramics? The Complete Technical Definition
Crazing is a network of fine cracks in the glaze surface caused by a thermal expansion mismatch between the glaze and the clay body.
The glaze shrinks more than the clay during cooling, putting the glaze layer under tension it cannot withstand.
According to Clay and Glazes for the Potter by Daniel Rhodes and Robin Hopper, crazing occurs when the glaze has a higher coefficient of thermal expansion (CTE) than the body it covers.
A glaze with a CTE of 8.0 × 10⁻⁶/°C applied to a clay body with a CTE of 6.5 × 10⁻⁶/°C will almost certainly craze because the 1.5-point gap far exceeds the 0.5-1.0 threshold for compatibility.
The cracks form perpendicular to the clay surface and stop at the glaze-clay interface.
They do not crack the clay underneath because the clay body is under compression while the glaze is under tension.
Think of it like a sheet of glass glued to a rubber pad – stretch the rubber and the glass cracks.
Measuring glaze specific gravity consistently at 1.45-1.50 helps control application thickness but does not fix the underlying thermal expansion mismatch.
How Crazing Differs from Cracking, Shivering, and Other Glaze Defects
Crazing is often confused with other glaze-surface failures, but the cause and appearance are distinct.
A body crack goes through the clay wall itself – you can see it on the unglazed bottom or feel it as a structural failure.
Crazing stays within the glaze and rarely affects the pot’s strength.
Shivering is the exact opposite of crazing. The glaze has a lower thermal expansion than the body, so the clay compresses the glaze until it pops off in sharp flakes – especially at rims and edges.
Dunting is a cooling crack that splits the entire pot, usually from rapid temperature drop through the quartz inversion range at 573°C (1063°F).
Pinholing and crawling are application or firing atmosphere defects, not expansion issues.
With crazing, the glaze did melt, did bond, and did cool – it just did not fit.
For most home studio potters, a crazed cone 6 glaze on vitrified stoneware with under 1% absorption is a cosmetic issue, not a structural failure. Understanding this difference is what separates functional potters from those who throw away perfectly good work.
Why Does Crazing Happen? The Glaze Chemistry and Physics Behind It
Crazing happens because the glaze contracts more than the clay body during cooling.
When the kiln drops below the glaze’s glass transition temperature – roughly 500-600°C (932-1112°F) depending on composition – the glaze becomes a rigid solid.
Below that point, any difference in shrinkage between glaze and body creates stress.
The mechanism is simple at the molecular level. If the glaze’s thermal expansion coefficient exceeds the body’s by more than about 0.5-1.0 percentage points, the tensile stress in the glaze layer exceeds its strength.
The glaze cracks to release the stress, forming the characteristic craze network.
According to Tony Hansen’s Digitalfire Reference Database, the three flux oxides most responsible for high glaze expansion are sodium oxide (Na₂O), potassium oxide (K₂O), and lithium oxide (Li₂O).
A glaze high in soda feldspar (such as Minspar 200) will have a CTE near 8.5 × 10⁻⁶/°C – far higher than most cone 6 stoneware bodies, which cluster around 6.0-7.0 × 10⁻⁶/°C.
This only occurs when the glaze CTE meaningfully exceeds the body CTE – a difference of less than 0.5 × 10⁻⁶/°C rarely produces crazing.
If the glaze CTE is too low, the result is shivering, not crazing. The fix is adjusting the glaze formula by reducing high-expansion fluxes such as sodium and potassium feldspars and replacing them with lower-expansion sources like lithium carbonate or magnesium carbonate.
Which Clay Bodies and Glazes Are Most Prone to Crazing?
Porcelain bodies are the most prone to crazing because white-firing porcelains typically have low thermal expansion – often 5.5-6.5 × 10⁻⁶/°C – while many commercial glazes are formulated for stoneware with higher expansion.
The result is a persistent tension gap that appears as glaze cracking within hours or days of firing – sometimes weeks later in delayed crazing.
Low-fire earthenware (cone 06-04, 999-1060°C / 1828-1940°F) has higher absorption and generally higher expansion, so crazing is less common at low temperatures.
However, earthenware bodies rarely vitrify fully, which creates a separate food safety problem if the glaze crazes – bacteria can travel through the cracks into the porous body.
Mid-fire stoneware (cone 5-6, 2167-2232°F / 1186-1222°C) is the workhorse of studio pottery and where most crazing frustrations occur.
Commercial cone 6 stoneware bodies from Laguna or Standard Ceramic typically have absorption rates of 1-3% and CTEs in the 6.0-7.0 range, making glaze fit a matter of matching a compatible commercial glaze or adjusting a recipe.
Key Specifications: Firing range: cone 5-6 (2167-2232°F / 1186-1222°C). Absorption after firing: 1-3% for most commercial stoneware. Cost: $18-30 per 25-pound bag for mid-range bodies.
Myth vs Fact
Crazing – Common Myths Debunked
Separating fact from fiction on the most common crazing misconceptions
✗ Myth
Crazing means the pot is underfired or the kiln was opened too soon.
✓ Fact
Crazing is a thermal expansion mismatch. A pot fired to perfect cone 6 with a controlled cooling cycle will still craze if the glaze CTE exceeds the body CTE. Firing temperature does not fix expansion chemistry.
✗ Myth
A crazed pot is automatically not food-safe.
✓ Fact
Food safety depends on the clay body’s absorption rate, not just the presence of crazing. A crazed glaze on vitrified stoneware with under 1% absorption is generally safe. The same crazing on porous earthenware with 8% absorption is a health hazard because bacteria can colonize the body.
✗ Myth
Adding more silica to a glaze always fixes crazing.
✓ Fact
Silica additions above the saturation point do not dissolve into the glass. Excess silica remains as undissolved particles and can create a matte or dry surface without lowering glaze expansion. The correct fix is reducing high-expansion fluxes such as sodium and potassium.
✗ Myth
All crazing happens immediately after firing.
✓ Fact
Delayed crazing can appear days, weeks, or even months after firing. Moisture absorption into the clay body from use causes slight expansion, increasing tension on the glaze until it cracks. This is particularly common in ware used for hot liquids.
✗ Myth
A thicker glaze application prevents crazing.
✓ Fact
Thicker glaze layers are under higher stress because the volume of contracting glass is larger. A glaze applied at 0.5mm may not craze where the same glaze at 2mm cracks visibly. Thinner glaze layers distribute tensile stress more evenly.
Is Crazed Pottery Food-Safe? The Critical Health Question
A crazed glaze on a fully vitrified clay body with under 1% absorption is generally considered food-safe by most studio standards.
A crazed glaze on a porous body with 3% or higher absorption is not food-safe.
The difference is whether bacteria and moisture can enter the clay body through the cracks and survive washing.
According to ASTM C738 testing standards for ceramic foodware, the primary risk is that bacteria (including E. coli and Salmonella) can enter the crazed crevices, resist removal during household dishwashing, and multiply when the pot is used again.
The National Sanitation Foundation (NSF) has historically required that food-contact ceramic surfaces be smooth, non-absorbent, and free of cracks – which crazed surfaces technically fail.
The practical reality for studio potters is more nuanced. A crazed mug on a zero-absorption vitrified porcelain body presents minimal risk because there is no porous clay for bacteria to colonize.
The same crazed glaze on an earthenware body with 8% absorption presents a significant risk because the clay body itself is a hospitable environment for microbial growth.
For functional potters selling work: disclose crazing to customers, recommend hand-washing, and never sell crazed ware on porous bodies for food use. A food-safe ceramic sealer cannot fix this – it sits on top of the glaze and wears off.
Quick Reference
Crazing and Food Safety – Key Terms Explained
Quick reference for the terms used throughout this guide
How to Test Your Fired Ware for Absorption Rate
Boil a representative fired (but unglazed) clay sample in water for 2 hours, then let it soak for an additional 22 hours. Weigh it wet, then weigh it after drying at 110°C (230°F) for 24 hours. The absorption percentage is: (wet weight minus dry weight) divided by dry weight, multiplied by 100.
Under 1% absorption: your clay body is vitrified and food-safe even with crazing. Between 1-3%: marginal for food safety with crazing – limit to occasional use and hand-washing. Above 3%: do not use crazed ware for food under any circumstances.
This test takes about 48 hours but it is the only reliable way to know whether your crazed pots are functionally safe or only decorative. A digital gram scale accurate to 0.1g gives precise enough readings for this test.
How to Fix Crazing: Glaze Chemistry Adjustments That Work
The only permanent fix for crazing is adjusting the glaze chemistry to lower its thermal expansion coefficient. There is no firing schedule trick, no application method change, and no post-firing treatment that permanently eliminates crazing if the CTE mismatch is too large.
According to Mastering Cone 6 Glazes by John Hesselberth and Ron Roy, the most effective adjustments to reduce glaze expansion are: reduce sodium and potassium feldspar content, replace soda feldspar partially with lithium carbonate or spodumene, add 5-10% silica in 325-mesh increments, and increase alumina slightly through kaolin additions.
A typical high-expansion cone 6 gloss glaze might contain 40% Minspar 200 (soda feldspar), 15% silica, 10% whiting, 8% kaolin, and 5% zinc oxide, with the remainder being frit and colorants. To lower its CTE, reduce Minspar to 25%, add 5% lithium carbonate, increase silica to 18%, and test the new version on your specific clay body.
Always test on vertical tiles of your clay body before committing to production. A test tile rack holding 20-30 vertical tiles in a single firing saves months of guesswork.
Step-by-Step Guide
How to Fix Crazing – Step by Step Glaze Adjustment
7 steps · Estimated time: 2-3 firing cycles with test tiles
Confirm It Is Crazing – Not Dunting or Shivering
Examine the crack pattern with a magnifying glass. Crazing forms a fine web network within the glaze only. A structural crack goes through the wall. Shivering flakes off at edges.
Test Your Clay Body Absorption Rate
Boil-soak an unglazed fired sample for 24 hours. If absorption is above 3%, your body is not vitrifying and the expansion mismatch may be in the body, not the glaze. Switch to a body rated for your cone range before adjusting glazes.
Identify the High-Expansion Flux in Your Glaze
Look at your glaze recipe. If soda feldspar (Minspar, F-4) or nepheline syenite exceeds 30% of the formula, that is likely your problem. Sodium oxide has a CTE contribution of approximately 9.0 × 10⁻⁶/°C – roughly double that of silica.
Mix a Test Batch with Reduced High-Expansion Flux
Reduce the soda feldspar by 5-10% and replace the lost flux with a combination of lithium carbonate (0.5-2%) and additional frit. Alternatively, substitute 25% of the soda feldspar with spodumene – a lithium-bearing feldspar with much lower CTE.
Add 5-10% Additional Silica (325 Mesh)
Silica (SiO₂) has a low CTE of about 0.5 × 10⁻⁶/°C. Adding more silica dilutes the high-expansion fluxes. Use 325-mesh silica (Flint) for solubility. Stay under 5% addition per test or the glaze may go matte or underfired.
Fire Test Tiles and Check for Crazing Over 2 Weeks
Apply the adjusted glaze to vertical test tiles of your clay body. Fire to your standard cone. Check immediately after cooling, then check again at 24 hours, 72 hours, 1 week, and 2 weeks. Delayed crazing is real.
Repeat – Small Adjustments Over 3-4 Test Firings
If the adjusted glaze still crazes, further reduce the soda feldspar and add more lithium. If the glaze surface becomes too dry or matte, add a small amount (2-3%) of additional frit to restore gloss without pushing CTE back up. Track every variation with a notebook and test tile label.
For potters who buy commercial glazes rather than mixing their own, fixing crazing is harder because you cannot change the formula. Try a different commercial glaze line rated as low-expansion – Amaco’s Potter’s Choice series includes several glazes with documented lower crazing tendencies on porcelain. Alternatively, switch to a clay body with higher natural expansion to compensate.
Intentional Crazing: When Crackle Glazes Are a Feature, Not a Defect
Deliberate crackle glazes are glazes formulated specifically to craze in a controlled, predictable pattern for aesthetic effect. The mechanism is identical – the glaze has higher thermal expansion than the body – but the result is designed, not accidental. Traditional Chinese Guan and Ge ware from the Song Dynasty used iron-rich clay bodies with deliberately mismatched celadon glazes to produce the crackle effect prized by collectors.
Modern studio potters create crackle effects by applying a high-sodium glaze over a low-expansion porcelain body, then rubbing India ink or iron oxide into the cooled cracks to darken them visually. The key difference from unwanted crazing is control – the potter knows the glaze will craze and selects a vitrified body so the pot remains functional.
A black India ink applied with a cloth and wiped off the surface stays in the craze lines and highlights the crack pattern for decorative effect. This technique works on both functional and sculptural crackle-glazed pieces.
Intentional crackle glazes on functional ware follow the same food safety rules as unintentional crazing: the body must be vitrified. A crackle-glazed porcelain teapot with under 1% absorption is generally safe. A crackle-glazed earthenware bowl is decorative only.
Delayed Crazing: Why It Appears Weeks After Firing
Delayed crazing occurs when moisture from use or ambient humidity slowly absorbs into a marginally vitrified clay body. As the body expands from water uptake, the tensile stress on the glaze increases until it exceeds the glaze’s strength – sometimes long after the pot left the kiln. A mug that looks perfect out of the kiln can develop full crazing after a month of daily coffee use.
This is why absorption testing on a fired but unglazed sample is critical before selling functional ware intended for food or liquid. A body with 2% absorption may pass the immediate post-firing visual check but craze within weeks of use. Bodies under 1% absorption are far less prone to delayed crazing because there is almost no pore space for moisture to enter.
The only reliable test for delayed crazing susceptibility is to boil an unglazed fired sample for 2 hours (which forces water into any available pores), then compare the wet weight to the completely dry weight. If the absorption exceeds 1.5%, expect delayed crazing with use.
Does Crazing Affect the Strength of the Pot?
Crazing slightly reduces the strength of the glaze surface but does not typically weaken the clay body itself. The cracks are surface-level phenomena confined to the glass layer. A crazed mug holds liquid just as well as an uncrazed one – the structural integrity comes from the clay wall, not the glaze coating.
However, a crazed glaze is more susceptible to chipping at the rim because the crack network creates weak points where pieces of glaze can separate from the body. This is a functional concern for dinnerware that stacks or cutlery that scrapes. A set of felt separators between stacked crazed plates prevents rim chipping during storage.
For most functional pottery, the strength loss from crazing is negligible compared to other factors – clay body vitrification level, wall thickness consistency, and the absence of structural flaws (such as s-cracks or compression issues from throwing) matter far more.
How to Prevent Crazing Before You Fire
The most reliable prevention is matching a known compatible glaze to a known compatible clay body before you invest time in making the work. Commercial clay and glaze manufacturers test this for you – use their recommended pairings. A Laguna cone 6 stoneware matched with Amaco Potter’s Choice glazes has documented lower crazing rates because both are formulated for mid-expansion compatibility.
When mixing your own glazes, calculate the expansion using glaze chemistry software such as GlazeMaster or Insight-Live before mixing a single gram of material. The calculated CTE of your glaze should be at least 0.5 × 10⁻⁶/°C lower than your clay body’s published CTE. Clay body manufacturers such as Standard Ceramic and Laguna publish CTE data on request.
A final prevention tip: avoid applying glaze too thickly. A 2mm glaze layer has twice the thermal mass and stress of a 1mm layer. Most dipping glazes achieve optimal coverage at 0.5-1mm thickness – any thicker and you are adding stress without improving surface quality.
Buying Guide
Before You Buy Glaze Materials – Crazing Prevention Checklist
Check off each point before making your material purchase decision.
How Does Kiln Cooling Rate Affect Crazing?
A slower cooling rate through the glaze’s glass transition range (roughly 500-600°C / 932-1112°F) can reduce crazing marginally by allowing the glaze and body to adjust to thermal contraction more gradually.
A fast cool through this range shocks the glaze into rigidity while the body is still contracting – amplifying any expansion mismatch.
However, a controlled cooling cycle cannot fix a fundamental CTE mismatch. If your glaze CTE is 8.5 on a body of 6.5, no cooling curve will prevent crazing because the inherent expansion difference is too large. Cooling rate adjustments help when the expansion mismatch is borderline – 0.5-1.0 × 10⁻⁶/°C difference – and the glaze is just barely crazing on a standard cycle.
For potters with programmable electric kilns: add a controlled cooling segment from 600°C to 500°C at 60-80°C per hour (108-144°F per hour) to give the glaze a longer transition window. This adds 1-2 hours to your firing schedule but sometimes eliminates borderline crazing without changing the glaze formula.
Can You Seal Over a Crazed Glaze to Make It Food-Safe?
No permanent sealant fixes food safety concerns from crazing on a porous clay body. Commercial ceramic sealers and food-grade epoxies form a surface film that wears off with use and dishwasher cycles, exposing the crazed surface underneath within weeks or months. A food-safe ceramic spray sealer is a temporary barrier, not a permanent solution.
The only reliable approach is using a vitrified clay body (under 1% absorption) so the crazed glaze has no porous substrate for bacteria to colonize. Disclosure to customers is still important – some users will not accept any crazing, intentional or not, in functional ware.
How Do Commercial Glaze Manufacturers Test for Crazing?
Commercial glaze manufacturers such as Amaco, Mayco, and Spectrum test their glazes on multiple clay bodies using autoclave accelerated aging. An autoclave cycle subjects glazed test tiles to high-pressure steam at 150°C (302°F), which forces moisture into any available pores and accelerates delayed crazing from months to hours. A glaze that survives 3 autoclave cycles without crazing on a given clay body is considered compatible for that body.
This is why commercial glaze labels often specify compatible clay body types – the manufacturer has done the testing and knows which combinations work. Using a glaze outside its recommended compatibility range voids that testing and puts the crazing outcome on you.
What Is the Difference Between Crazing and Crackle Glaze?
Crazing is an unintentional glaze defect caused by thermal expansion mismatch. Crackle glaze is intentional crazing formulated for decorative effect, typically applied to a vitrified body to maintain functionality. The physical mechanism is the same – the glaze CTE exceeds the body CTE – but the potter’s intention and the end result are different.
Traditional Chinese crackle glazes (Guan, Ge, Ru) from the Song Dynasty used iron-bearing clay bodies with specific celadon formulations. Modern crackle glazes achieve the effect with high-sodium frit-based recipes fired on low-expansion porcelain. The crack pattern can be controlled by glaze thickness – thicker applications produce larger crack networks.
Why Does My Glaze Craze on One Clay Body But Not Another?
Each clay body has a unique coefficient of thermal expansion determined by its silica, alumina, and flux content. A glaze that fits perfectly on a high-silica stoneware body with CTE around 6.8 × 10⁻⁶/°C may craze on a porcelain body with CTE around 5.8 × 10⁻⁶/°C because the expansion gap is twice as large.
Even two “cone 6 stoneware” bodies from different manufacturers can have CTEs that differ by 0.5-1.0 × 10⁻⁶/°C. Always test your glaze on the exact clay body you intend to use. A successful test on Laguna B-Mix does not guarantee success on Standard 630 – the bodies are chemically different.
Does Adding Grog to the Clay Body Reduce Crazing?
Grog (pre-fired clay particles ground to specific mesh sizes) primarily affects workability, drying shrinkage, and thermal shock resistance. It has minimal effect on the thermal expansion of the fired clay body because grog and the surrounding clay matrix expand at similar rates once vitrified.
A body with 20% grog may have slightly lower overall expansion because grog is already fired and dimensionally stable. The effect is typically small – 0.1-0.3 × 10⁻⁶/°C – and not enough to fix a glaze that is crazing by a full percentage point. Grog helps with drying cracks, not glaze fit problems.
Can I Refire a Crazed Pot to Fix the Crazing?
Refiring to the same cone temperature will not fix crazing because the glaze remelts and re-enters tension on cooling – the same CTE mismatch still exists. The cracks seal temporarily during the melt phase, then reform on cooling because the underlying chemistry has not changed.
The only useful refire approach is applying a lower-expansion overglaze on top of the crazed surface and refiring. This overglaze must have a CTE lower than both the original glaze and the body. It creates a compression-sealed surface, but results are inconsistent and the double glaze layer can be too thick, causing crawling or blistering.
Is Raku Pottery Crazed?
Raku glazes almost always craze because raku firing involves rapid cooling from approximately 1000°C (1832°F) to room temperature in seconds or minutes – the most extreme thermal shock in ceramics. The crazing is accepted as part of the raku aesthetic and raku ware is almost never used for food due to the combination of crazing, high absorption, and sometimes toxic glaze ingredients.
Raku bodies are formulated for thermal shock resistance (high grog, open structure) and typically have absorption rates above 5%. This makes raku ware functionally decorative, regardless of whether the glaze crazes.
What Role Does Silica-Alumina Ratio Play in Crazing?
The silica-to-alumina ratio in a glaze strongly influences its thermal expansion. A high-silica, low-alumina glaze tends to have lower expansion because silica glass (SiO₂) has a CTE of approximately 0.5 × 10⁻⁶/°C. A high-alumina, low-silica glaze tends to have higher expansion because alumina (Al₂O₃) in the glass network increases rigidity and CTE.
According to the Digitalfire Reference Database, the optimal silica-to-alumina molar ratio for low-expansion cone 6 glazes is approximately 7:1 to 10:1. Ratios below 5:1 (more alumina relative to silica) increase the risk of crazing because the glass network becomes stiffer and less able to absorb thermal stress without cracking.
Can Dishwasher Use Make Crazing Worse?
Yes, repeated dishwasher cycles accelerate crazing development through thermal cycling and moisture absorption. A dishwasher heat-dry cycle raises the pot temperature to approximately 70-80°C (158-176°F), then the pot cools rapidly to room temperature during the drying phase – creating micro thermal stress cycles.
For functional ware with any crazing, hand-washing is recommended to extend the pot’s functional life and minimize progressive crack enlargement. Dishwasher detergent is also more alkaline than hand dish soap and can slowly leach oxides from the glaze along the crack edges, darkening the craze lines over time.
How Do I Know If My Crazed Pot Is Safe for Coffee or Tea?
A crazed mug is safe for hot coffee or tea only if the clay body is vitrified (under 1% absorption) and the glaze is certified lead-free and cadmium-free. Test the body absorption rate using the boil-soak method on a fired unglazed sample of the same clay. If the body absorbs under 1%, the crazing is cosmetic and the mug is functionally safe for hot beverages.
If the body absorbs 1-3%, limit use to occasional cold liquids and hand-wash only. Above 3% absorption, do not use for any food or beverage regardless of how the glaze looks. The hot liquid will expand the body slightly, opening the craze lines further each use.
Should I Tell Customers That My Pottery Is Crazed?
Yes, always disclose crazing to customers in functional ware listings and at markets. Describe it accurately – not as a defect, but as a characteristic of the glaze-body interaction. Some buyers specifically seek out crackle-glazed or crazed surfaces for their aesthetic appeal. Others will not accept it under any circumstances, and they deserve to know before purchasing.
A simple note on the tag or listing – “This glaze has a fine craze pattern. The clay body is fully vitrified and the piece is safe for food use with hand-washing” – both informs the customer and protects you from returns or complaints about a surface they did not expect.
Mastering crazing comes down to understanding two numbers: your glaze CTE and your clay body CTE. When the glaze contracts less than the body during cooling, the result is a durable, uncrazed surface. When the glaze contracts more, the result is the web of fine lines that potters have debated for centuries.
For the functional potter, the priority is vitrification first, then glaze fit. A crazed glaze on a zero-absorption body is cosmetic. An uncrazed glaze on a porous body is still not food-safe. For more on how clay bodies vitrify and what absorption rates mean for your work, explore our complete guide to ceramic materials science and clay body chemistry and our detailed breakdown of porosity in fired ceramics and its impact on functional ware. If you are working with clay that has not yet been bisque-fired, see our explanation of greenware stages and how drying affects final glaze results. For potters exploring decorative surface patterns that work with crazing effects, our guide to coil building techniques and textured surfaces covers how handbuilt forms interact differently with glaze fit.






