Why Does Glass Crack When Heated? Thermal Shock Explained

Glass can crack during rapid heating or cooling because different parts change temperature at different rates. Their unequal expansion or contraction creates stress. A sufficiently large tensile stress can open a flaw and cause fracture. This mechanism is called thermal shock. SCHOTT explains the connection between temperature gradients, thermal expansion and stress in its technical note Thermal loads on optical glass.

The useful question is therefore more specific than “How hot can this glass get?” It is also “How will the temperature change across this particular component?”

What happens during a sudden temperature change?

Imagine a hot glass part whose outside is suddenly cooled. Its surface tries to contract while the warmer interior restrains it. The resulting tensile stress at the surface can promote cracking. Heating produces a different stress distribution, so the two situations should not be treated as identical. SCHOTT's thermal-load analysis describes these differences.

Material properties are only part of the explanation. The same analysis identifies component geometry, support conditions, the surrounding medium and the rate of temperature change as factors in the developing stress. A glass name alone cannot supply a universal thermal-shock limit.

Why does borosilicate glass help?

Low-expansion borosilicate glass changes dimensions less for a given temperature change. For example, SCHOTT's BOROFLOAT 33 thermal data compare mean linear expansion coefficients of:

• 3.25 × 10⁻⁶ K⁻¹ for BOROFLOAT 33.
• 9.06 × 10⁻⁶ K⁻¹ for the soda-lime glass in that comparison.

Both values refer to the 20–300 °C interval. The borosilicate value is roughly one-third of the soda-lime value, helping reduce thermally generated stress under comparable conditions. These are specific reference values, not specifications for every glass in either family.

Even “borosilicate” covers different compositions. DWK's glass-property tables list borosilicate grades with different expansion coefficients. For a real component, identify the exact grade and its supplier data.

Is a high temperature rating a thermal-shock rating?

A service-temperature figure needs its conditions. The same BOROFLOAT 33 datasheet distinguishes short- and long-duration operating temperatures. Those figures do not describe an arbitrary jump between hot and cold environments.

For laboratory glassware, DWK's heating guidance recommends gradual heating and cooling and avoiding localized hot spots. A material described as heat-resistant still needs a suitable heating method and a product-specific operating limit.

This is also separate from the glass transition. A thermal-shock problem should not be diagnosed from the glass-transition temperature alone.

Why condition and product design matter

An undamaged laboratory flask and a scratched storage bottle are different engineering situations, even when both use borosilicate glass. Corning's PYREX laboratory-glassware guidance calls for inspecting glassware for scratches, chips and cracks. It also distinguishes vessels intended for direct heating from thick-walled storage bottles that should not be placed directly on a hotplate or flame.

That example shows why composition cannot replace a product specification. The manufacturer's instructions for the actual item take priority over a generic comparison between glass families. Laboratory guidance should not be transferred automatically to cookware, windows or pressure equipment.

How to investigate a recurring glass failure

For a materials-development project, I would organize the first discussion around the following questions:

1. What is the glass? Record the grade, supplier, drawing and any documented thermal treatment.
2. What changed? Describe the heating or cooling step, contact medium, timing and the last successful operating condition.
3. Where did failure begin? Keep photographs and record whether the first visible crack was near an edge, contact point or other identifiable feature.
4. What comparisons are available? Note differences between failed and unaffected parts, including dimensions, mounting and handling history.
5. What would count as success? Define the intended thermal cycle and an acceptance criterion before choosing a test.

This is a proposed investigation framework, not evidence that any one cause has already been established. Fracture-surface analysis can help test an explanation: NIST's introduction to fractography describes how fracture patterns help engineers investigate failures in glasses and ceramics. A crack appearing during heating is a starting observation, not a complete diagnosis.

For a research or industrial question, the contact page is a place to describe the material, the operating conditions and the outcome you need. An initial non-confidential description is enough to begin.

Sources and further reading

• SCHOTT: TIE-32, Thermal loads on optical glass — thermal gradients, stress direction and the role of geometry and boundary conditions.
• SCHOTT: BOROFLOAT 33, Thermal Properties — the expansion comparison and duration-dependent operating-temperature data.
• DWK Life Sciences: Glass Types & Properties — differences between specified borosilicate and soda-lime glass grades.
• DWK Life Sciences: Tips for safely heating laboratory glassware — gradual heating and cooling, heat distribution and product limitations.
• Corning: PYREX Brand Glass Products — laboratory product distinctions and inspection for damage.
• NIST: Guide to analyzing brittle-material failures — an introduction to the role of fractography in failure analysis.