How to measure a compression spring, explained

A broken or unmarked spring carries its own drawing — four measurements recover it. The traps are not in the measuring but in the conventions: which diameter the math wants, and which coils actually count.

The four measurements

Wire diameter d — mic the wire a coil or two away from the ends (ground tips measure thin, plating measures fat). Outside diameter — calipers across the coil body; the math then wants the mean diameter:

D = OD − d · k = G·d⁴ / (8·D³·Nₐ)

Free length L₀ — uncompressed, standing on end. Coil count — follow the wire tip to tip; each full revolution is one total coil. Then convert to active coils: closed & ground or squared ends park one dead coil at each end (Nₐ = Nt − 2). The spring rate calculator turns the four numbers into the rate, and the full compression spring calculator adds stress, solid height and buckling.

Worked example — an unmarked die spring

Measured: d = 2 mm, OD = 20 mm, ten total coils with closed-and-ground ends, music-wire look. So D = 18 mm, Nₐ = 8:

k = 79,300 × 2⁴ / (8 × 18³ × 8) = 3.40 N/mm · solid height = d(Nₐ+2) = 20 mm · C = D/d = 9

The spring index of 9 sits mid-range (springs run ~4–12), which is itself a sanity check that the measurements are consistent. Solid height says the spring can stroke 30 mm from its 50 mm free length before coil-bind — minus whatever margin the deflection check demands.

Confirm with a load test

Geometry assumed the material; a two-point load test removes the assumption. Compress to two well-separated heights, read the forces, and k = ΔF/Δx — the spring constant calculator does the division. Measured ≈ computed: carbon-steel family confirmed. Measured ~13% soft: stainless (G ≈ 69 vs 79.3 GPa — the wire properties table has the values). Way off: recount the coils.

Common mistakes

• Using OD in the rate formula. Rate goes as 1/D³, so the 20 mm OD example computed on OD instead of D = 18 reads 2.48 N/mm against a true 3.40 — 27% under. Always subtract one wire diameter.
• Counting dead coils as active. Ten total ≠ ten active: closed ends bury two. Using Nt makes the spring look 20% softer than it is (rate ∝ 1/Nₐ).
• Micing the wire at a ground tip or over plating. d enters at the fourth power — 0.05 mm of error on 2 mm wire moves the rate 10%.
• Trusting one load point. End coils seat in the first millimeter of travel and fake a soft start; two separated points measure the real line.

Frequently asked questions

How do you measure a compression spring?

Four numbers define it: wire diameter (mic the wire mid-coil), outside diameter (calipers), free length, and the coil count. From those, mean diameter D = OD − d and the rate k = G·d⁴/(8D³Nₐ) follow. A spring with 2 mm wire, 20 mm OD and 8 active coils in music wire rates 3.40 N/mm.

How do you count active coils on a spring?

Count total turns by following the wire from tip to tip — every full revolution is one coil — then subtract the dead end coils: closed & ground ends lose one coil per end (Nₐ = Nt − 2), plain open ends deflect along their whole length. A 10-total-coil closed-and-ground spring has 8 active coils, and using 10 in the formula reads the rate 20% soft.

Do you use outside diameter or mean diameter for spring rate?

Mean diameter, always: D = OD − wire diameter. Rate scales with 1/D³, so the error compounds — computing the 20 mm OD example with OD instead of D = 18 mm underestimates the rate by 27%. It is the single most common spring-math mistake.

How do you identify the spring material?

Calipers cannot — but the rate test can narrow it. Compute k from geometry with G ≈ 79.3 GPa (the carbon-steel family: music wire, oil-tempered, chrome silicon all sit near it), then load-test the spring: if measured rate runs ~13% under the computed value, suspect stainless (G ≈ 69 GPa). Color and magnetism help; certainty needs the drawing.