Logo

MonoCalc

/

Thermal Expansion Calculator

Physics

Or start from a worked scenario:

Linear uses α, area uses 2α and volume uses β.
Leave the unknown blank; the rearranged formula is shown with the answer.
Presets carry a coefficient, the temperature it was measured at, and — where available — a Young's modulus from the elasticity table.
Filled in from Steel (mild); switch to Custom to edit.
1/°C and 1/K are identical — both are per-kelvin intervals. A per-°F figure is 5/9 of the per-K one.

Expansion

ΔL
0.054 m
New L
100.054 m
Thermal strain ε = αΔT
540.000 ppm
ΔL = α · L₀ · ΔTΔT = 45.000 Kα = 12.000e-6 /K

Change across units

UnitValue
m0.054
cm5.400
mm54.000
µm54000.000
in2.126

About This Tool

Thermal Expansion Calculator – ΔL = αL₀ΔT for Length, Area, Volume and Stress

Almost everything gets bigger when it gets hotter, and the amount is small enough to ignore right up until it is not. A thermal expansion calculator answers the two questions that follow from that: how much bigger does this get, and what happens if you do not let it? This tool works ΔL = α · L₀ · ΔT in one dimension, ΔA = 2α · A₀ · ΔT in two and ΔV = β · V₀ · ΔT in three, and rearranges any of them for whichever quantity you leave blank.

The linear case, and why expansion joints exist

A 100 m carbon-steel bridge girder taken from −10 °C to 35 °C sees ΔT = 45 K. With α = 12.0e-6 /K it grows 12.0e-6 × 100 × 45 = 54.000 mm. That is a thermal strain of ε = αΔT = 540 ppm— five hundredths of one percent, and also the reason every long span carries an expansion joint. Reporting the result in parts per million alongside millimetres is deliberate: ppm is how materials datasheets state α, and it makes the identity “the coefficient of expansion is just strain per kelvin” obvious.

Where the familiar formula starts to lie

ΔL = αL₀ΔT is the first-order expansion of L = L₀(1 + αΔT), and the truncation compounds with dimension. Take a 0.500 × 0.300 m aluminium plate through 80 K: the linear area formula gives 554.400 mm², while expanding each side exactly and differencing gives 554.912 mm². The 0.512 mm² gap is exactly the corner square A₀(αΔT)² that the linear form throws away, and its relative size is always half the thermal strain. The same thing happens in three dimensions: β = 3α is really β = 3α + 3α²ΔT + α³ΔT², which is 0.12 % off for steel over 100 K but 5.77 % off for HDPE over 500 K.

The bigger error is usually the coefficient
α is itself temperature-dependent. Steel runs about 11.0e-6/K near 0 °C and 13.5e-6/K near 500 °C — roughly ±10 %, which swamps a sub-percent truncation over the same span. Every preset carries the temperature its coefficient was measured at, and the calculator ranks that warning ahead of the arithmetic ones.

Thermal stress: what restraint costs

Prevent the expansion and it reappears as stress: σ = E · α · ΔT. Carbon steel with E = 200 GPa restrained through 45 K builds 108 MPa, which is 43.2 % of a 250 MPa mild-steel yield from one seasonal swing. Two things surprise people here. First, length does not appear— a 100 m girder and a 100 mm coupon carry the same stress. Second, aluminium expands 1.925× as much as steel yet builds only 71.7 MPa, because its modulus is far lower. The quantity that ranks materials for restrained service is the product , the thermal stress coefficient: 2.400 MPa/K for steel against 1.594 MPa/K for aluminium.

Liquids, containers and bimetallic strips

A brim-full tank overflows by the difference between what the contents gain and what the vessel gains. Sixty litres of gasoline (β = 950e-6 /K) in a steel tank (β = 36.0e-6 /K) warming 25 K spills 1.371 L, not 1.425 L — and for mercury in glass the vessel term hides 15 % of the true expansion, which is precisely how a thermometer is calibrated. Bond two metals with different coefficients and the strip curves toward the low-α side, which is the mechanism inside every mechanical thermostat and circuit breaker.

Practical uses

Size a rail expansion gap from a service temperature range and get both the total travel and the gap to actually set at the install temperature. Work out the shrink fit: a 50 mm steel ring with 40 µm interference needs only 66.67 Kof heating, a hot-water bath rather than a furnace. Or run it backwards — a 2.000 m sample that grew 1.64 mm over 68 K gives α = 12.06e-6 /K, which identifies carbon steel, and lands on the coincidence that makes reinforced concrete possible: concrete and steel share almost exactly the same coefficient, so rebar and concrete move together instead of tearing each other apart.

Frequently Asked Questions

Is the Thermal Expansion Calculator free?

Yes, Thermal Expansion Calculator is totally free :)

Can I use the Thermal Expansion Calculator offline?

Yes, you can install the webapp as PWA.

Is it safe to use Thermal Expansion Calculator?

Yes, any data related to Thermal Expansion Calculator only stored in your browser (if storage required). You can simply clear browser cache to clear all the stored data. We do not store any data on server.

How does this thermal expansion calculator work?

It works ΔL = α·L₀·ΔT for linear expansion, ΔA = 2α·A₀·ΔT for area and ΔV = β·V₀·ΔT for volume, and rearranges each for whichever quantity you leave blank — the change, the coefficient, the original dimension or the temperature change. A 100 m carbon-steel girder going from −10 °C to 35 °C grows 12.0e-6 × 100 × 45 = 54.000 mm. Every input is normalised to SI at full double precision and rounded only when it is printed.

Why does the calculator show an exact value alongside the formula answer?

Because ΔL = αL₀ΔT is a first-order truncation of L = L₀(1 + αΔT), and the dropped term grows with dimension. A 0.500 × 0.300 m aluminium plate heated 80 K gains 554.400 mm² by the linear formula but 554.912 mm² exactly — the 0.512 mm² difference is precisely A₀(αΔT)², and its relative size is always half the thermal strain. For a metal plate that is 0.09 % and irrelevant; for a polymer over a wide span it is percent-level and real.

Is β = 3α always correct for volume expansion?

Only to first order, and only for isotropic materials. Expanding (1 + αΔT)³ exactly gives β = 3α + 3α²ΔT + α³ΔT², so the error scales with αΔT — 0.12 % for steel over 100 K but 5.77 % for HDPE over 500 K. Anisotropic materials are worse: wood expands about 9× more across the grain than along it, so β = α₁ + α₂ + α₃ rather than 3α. Liquids and gases carry a measured β directly because they have no meaningful linear coefficient at all.

What happens if a heated part cannot expand?

The expansion it would have undergone shows up as strain, and therefore as stress: σ = E·α·ΔT. Carbon steel restrained through a 45 K rise builds 200e9 × 12.0e-6 × 45 = 108 MPa, which is 43.2 % of a 250 MPa mild-steel yield from a single seasonal swing. Notice that length does not appear — a 100 m girder and a 100 mm coupon, equally restrained, carry identical stress.

Does aluminium build more thermal stress than steel because it expands more?

No, and this catches almost everyone. Aluminium expands 1.925× as much as steel, but its modulus is far lower, so through the same 45 K it reaches 71.7 MPa against steel's 108 MPa — steel builds 1.51× the stress. The quantity that ranks materials for restrained service is the product Eα, the thermal stress coefficient: 2.400 MPa/K for steel against 1.594 MPa/K for aluminium.

How accurate are the tabulated expansion coefficients?

The coefficient itself is the largest source of error, not the formula. Steel's α runs about 11.0e-6/K near 0 °C and 13.5e-6/K near 500 °C — roughly ±10 %, which swamps the sub-percent truncation error over the same span. Every preset therefore carries the temperature its coefficient was measured at, and the calculator warns once the span exceeds 150 K for metals or 60 K for polymers. It also blocks outright where the model breaks down, such as water between 0 °C and 4 °C, where the coefficient changes sign.