Advertisement
Advertisement
PHYSICS

Thermal Expansion Calculator — linear, area and volumetric growth

Find how much a part grows or shrinks with temperature, and the stress it develops if something stops it moving.

Area expansion uses 2α and volumetric uses 3α, because a solid grows in every direction at once. Liquids are quoted with their own β directly.
Enter it in whatever unit you like — metres, millimetres, square feet, litres. Expansion is a fractional change, so the answer comes back in the same unit you typed.
Cooling is handled the same way — enter a lower final temperature and the change comes back negative, which is contraction.
Used only for the last grid figure: the stress that appears if the part is prevented from moving at all. Steel is about 200 GPa, aluminium 69, concrete 30.
Change in length
 
 
0
New size after the change
0
Temperature change
0
Fractional strain
0
Stress if fully restrained
Tip: the stress a restrained part develops does not depend on its length at all. A 10 mm steel block and a 100 m steel bridge, both held rigidly through the same temperature rise, reach exactly the same stress — which is why expansion joints exist on structures of every size.
Advertisement

The thermal expansion calculator above computes how much a length, an area or a volume changes when its temperature changes, from the material's coefficient of expansion and the temperature swing. It also computes the quantity that actually causes damage: the stress that appears when a part is heated or cooled and something prevents it from moving. That second figure is the one that cracks concrete, buckles rails and shears bolts, and it is missing from most expansion calculators.

Arb Digital builds free calculators that report the consequence as well as the quantity. Expansion on its own is usually harmless — a few millimetres over a long span is easy to accommodate if you design for it. It becomes a problem only when the movement is blocked, and at that point the numbers get large very quickly. A thirty-degree swing in restrained steel produces around seventy megapascals of stress, a significant fraction of its yield strength, with no external load involved at all.

What This Thermal Expansion Calculator Does

In linear mode it returns the change in a length, span, diameter or gap. In area mode it returns the change in a surface, using a coefficient of twice the linear value. In volumetric mode it returns the change in a solid body's volume, using three times the linear value. The distinction matters because a solid expands in all directions simultaneously, and the higher-dimensional coefficients are simply that fact expressed arithmetically.

The size field is deliberately unit-free. Thermal expansion is a fractional change, so a ten per cent growth is ten per cent whether the original figure was in metres, inches or litres. Enter your dimension in whatever unit your drawing uses and the change and the new size come back in that same unit, which removes an entire class of conversion errors.

The material selector loads a representative coefficient, and the field remains editable so you can enter a value from your own specification. Published coefficients vary between sources and with temperature range, so a supplier's figure for the specific alloy or mix you are using is always better than a generic one.

How to Use It

  1. Pick the expansion type first. It sets whether the coefficient is applied once, twice or three times, and the hero label renames itself accordingly.
  2. Enter the original size in your own unit. Metres, millimetres, square feet, cubic metres and litres all work, and the result returns in the same unit.
  3. Choose a material or type a coefficient. The value is in parts per million per kelvin, which is how expansion coefficients are almost always tabulated.
  4. Enter the two temperatures. Their difference drives everything, and a lower final temperature gives a negative result, meaning contraction.
  5. Set Young's modulus if you want the restrained stress. This is the figure to check against the material's allowable stress when the movement cannot be accommodated.

The Formula: How Thermal Expansion Is Calculated

Linear expansion is ΔL = αLΔT, area expansion is ΔA = 2αAΔT, and volume expansion is ΔV = βVΔT with β = 3α for an isotropic solid. OpenStax University Physics Volume 2, section 1.3 on thermal expansion, gives all three forms and states the relationship between the coefficients explicitly, alongside a table of representative values for solids, liquids and gases.

The reason area gets a factor of two and volume a factor of three is that each dimension grows by the same fraction independently. A square whose side grows by a factor (1 + αΔT) has an area that grows by (1 + αΔT)2, and expanding that gives 1 + 2αΔT plus a term in α2 that is utterly negligible when α is measured in parts per million. The HyperPhysics page on thermal expansion makes the same point that the fractional expansion of a uniform linear object is proportional to the temperature change.

Restrained stress follows from Hooke's law. If a part wants to strain by αΔT and is prevented entirely, the material carries that as an elastic strain instead, so σ = EαΔT. Length has cancelled out of that expression, which is the counter-intuitive result quoted in the tip above the article.

Work the default values. A 10-metre concrete element with α = 12 × 10−6/K warmed from 10 °C to 40 °C has ΔT = 30 K, so ΔL = 10 × 12 × 10−6 × 30 = 3.6 × 10−3 m, or 3.6 mm. The strain is 360 parts per million. Fully restrained at a modulus of 200 GPa, the stress would be 200 × 109 × 12 × 10−6 × 30 = 72 MPa.

Advertisement

Why Restrained Stress Does Not Depend on Length

This is the single most useful and least intuitive result on the page. Everyone expects a long member to be more of a problem than a short one, because it moves further. But stress is force per unit area, and it comes from strain, which is a fraction rather than an absolute distance. The fraction is the same for every length, so the stress is the same for every length.

What length does change is the movement you must accommodate. A 3.6 mm gap is trivial to design into a joint; a 36 mm one on a hundred-metre span needs a proper expansion device. So length drives the detailing while temperature range and material drive the stress. The two questions are separate, and mixing them up is how expansion joints get specified for the wrong reason.

It also explains why small components fail from thermal stress just as readily as large structures. A ceramic seal bonded to a metal housing sees the full stress from the mismatch in their coefficients regardless of how small it is. Differential expansion between bonded materials is a common failure mode in electronics packaging, glass-to-metal seals and coated components, and none of it depends on the parts being big.

Where the Simple Model Breaks Down

The coefficient is not truly constant. It varies with temperature, usually rising as the material gets hotter, so a value tabulated near room temperature will misstate behaviour across a wide range. Handbooks address this by quoting a mean coefficient over a stated temperature interval, and using one outside its interval is a common source of error.

Anisotropic materials expand differently along different axes. Wood expands several times more across the grain than along it, which is why timber panels are detailed to allow cross-grain movement while the length is held. Composites, single crystals and rolled metals can all show direction-dependent coefficients, and a single scalar α is meaningless for them.

Water is the famous exception to the whole picture, contracting rather than expanding as it warms between 0 and 4 °C. That anomaly is why ice floats and why lakes freeze from the surface down. For any liquid, use the tabulated volumetric coefficient directly rather than tripling a linear one, since a liquid has no shape of its own to expand linearly. Our water density calculator handles that temperature dependence for water specifically.

Designing for Movement Rather Than Fighting It

The practical response to thermal expansion is almost always to let it happen. Bridges sit on sliding bearings and finger joints; pipe runs include expansion loops or bellows; long buildings are divided by movement joints; rails are laid with a stress-free temperature chosen so that neither the summer compression nor the winter tension exceeds what the track can carry. Each of these is cheaper and more reliable than trying to restrain the movement.

Where restraint is unavoidable, the stress figure becomes a design case that must be added to the mechanical loads already present. Combining thermal stress with load-induced stress is what the von Mises stress calculator is for, and the shear stress calculator covers the shear component that differential expansion creates at a bonded interface. Where a spring or flexible element absorbs the movement instead, the spring rate calculator gives the force it generates.

The temperature range you design for should be the extreme service range, not the ambient one. A dark steel surface in direct sun reaches well above air temperature, and an unheated building in winter goes well below it, so the swing a component actually sees is usually wider than the weather record suggests. To find the heat energy involved in reaching those temperatures, use the specific heat calculator or the sensible heat calculator, and for how fast heat crosses the material, the thermal conductivity calculator.

Need a website that loads fast and actually works?

Arb Digital builds free tools like this one because useful pages earn attention. If you want tools, calculators or content built for your own audience, we can help.

Browse All Free Tools Talk to Arb Digital

Common Mistakes to Avoid

  • Using the linear coefficient for a volume — a solid needs three times the linear value, and using α where β belongs understates the volume change by a factor of three.
  • Tripling a liquid's coefficient — liquid coefficients are already volumetric, so multiplying by three again is the same error in reverse.
  • Applying a coefficient outside its stated temperature range — published values are means over an interval, and they drift with temperature.
  • Assuming stress scales with length — it does not. Length changes the movement to accommodate, not the stress a restrained member develops.
  • Designing for ambient rather than service temperatures — a surface in direct sun or an unheated space in winter both go well outside the air temperature range.

Related Free Tools From Arb Digital

Combine thermal stress with mechanical loading using the von Mises stress calculator or the shear stress calculator, and find the force a flexible element generates as it takes up movement with the spring rate calculator. For the heat side of the same problem, the specific heat calculator and sensible heat calculator give the energy involved, and the thermal conductivity calculator gives how fast it crosses the material. Liquid behaviour is covered by the water density calculator. The full free online tools hub lists everything.

Frequently Asked Questions

How do I calculate thermal expansion?

Multiply the original dimension by the coefficient of expansion and by the temperature change. For a length that is alpha times L times delta T; for an area it is twice alpha; for the volume of a solid it is three times alpha, because the material grows in every direction at once.

Why is the volume coefficient three times the linear one?

Because each of the three dimensions grows by the same fraction independently. Cubing the linear growth factor gives one plus three alpha delta T, plus higher-order terms that are negligible when alpha is measured in parts per million. The same reasoning gives two alpha for an area.

Does thermal stress depend on the length of the part?

No, and this surprises most people. Stress comes from strain, which is a fraction rather than a distance, and the fraction is identical at every length. A short block and a long span held rigidly through the same temperature change reach exactly the same stress.

What temperature unit should I use?

Celsius is fine, because only the difference between the two temperatures matters and a one-degree Celsius change is identical to a one-kelvin change. Coefficients quoted per kelvin and per degree Celsius are numerically the same for the same reason.

How do I handle liquids?

Use the liquid's own volumetric coefficient in volumetric mode and do not multiply it by three. A liquid has no fixed shape to expand linearly, so its coefficient is already volumetric. Water is also anomalous, contracting rather than expanding between 0 and 4 degrees Celsius.

Is the coefficient constant across all temperatures?

No. It generally rises as the material gets hotter, so published figures are means over a stated interval. Using a room-temperature value across a range of several hundred degrees will misstate the result, sometimes substantially. Check the interval your source quotes.

Can I use this for wood or composites?

Not with a single coefficient. Anisotropic materials expand by different amounts along different axes — wood moves several times more across the grain than along it — so each direction needs its own coefficient and its own calculation.

This tool is provided for educational and estimating use. It applies a constant-coefficient isotropic model and does not account for anisotropy, temperature-dependent coefficients, phase changes, creep or partial restraint, so treat its output as a physics result rather than a structural design.

Advertisement
Advertisement

Take it further