EN

Compression & Euler Buckling Calculator

A column, a strut or a screw under a compressive load can fail in two very different ways: by crushing of the material (the stress σ = F/A exceeds the yield strength) or by buckling — a geometric instability that snaps the part sideways well before the material is anywhere near its limit. This calculator checks both failure modes: it computes the compressive stress and the Euler critical load, and tells you which of the two governs the design.

A slender compressed part almost always fails by buckling: the critical load falls with the square of the length. Doubling the length of a column divides its capacity by four.

Load, member and end conditions

Load and length

Effective length Lf = K × L.

Cross-section

Change the shape then hit "Calculate" to reveal the matching fields.

Material

E and Re fill in automatically (except "Custom").

Used only when "Custom" is selected.

Used only when "Custom" is selected.

An asterisk on first use points to the glossary at the end.

Compression is not just negative tension

Compression looks, at first glance, like tension with the sign flipped: σ = F / A, shortening in place of stretch. A short, stocky block does fail that way — crush, yield, a mushroomed end. A long member has a second way out, and that way does not ask Re for permission.

Buckling* is an instability of shape, not of the material. No bar is perfectly straight, and no compressive load is perfectly centred. A small eccentricity produces a small moment; the moment increases the bow; the bow increases the moment. Past the critical load the loop runs away, and the member snaps sideways, often while the stress is still far below yield. A tie has nothing like it. Checks in shear and in torsion answer different questions; neither one is this sideways departure.

Two safety factors come off this page. One is Re / σ, crushing. The other is Fcr / F, the Euler load*. The smaller one governs. A slender strut can sit at 20 MPa and remain a millimetre from folding. That is not a material problem. It is a length-and-stiffness problem.

Euler's formula: stiffness, not strength

Fcr = π² × E × I / Lf²   with   Lf = K × L

  • E — the only material number Euler’s load uses. Buckling is a stiffness event. A 42CrMo4 strut buckles at the same load as an S235 strut of the same shape. Raising the grade does nothing until the member is stocky enough to crush first.
  • I — the minimum second moment of area*. The column buckles about its weak axis, always. Metal set further from the centre (a tube, an H-section) raises I without adding mass.
  • Lf = K·L — the effective length*. End restraints rewrite the real length through the end-fix factor* K. With both ends fixed, K = 0.5: half the length, four times the load. A mast, fixed–free, takes K = 2 and carries sixteen times less than the same stick built in as a column. The textbook extremes really are a factor of 16.

Fcr falls with the square of the length. Double L, and the critical load drops to a quarter.

A 4 m prop and a 2 m prop are not “the same tube, twice as long”. They are different members.

Slenderness decides crushing versus buckling

The slenderness* ratio λ = Lf / i, with i = √(I/A) the radius of gyration*, is the switch. Below the critical slenderness λc = π√(E/Re) — about 94 for S235 — the member is stocky. It yields before Euler is allowed to speak, and the crushing check on this page is the one that matters. Above λc the member is slender. Euler governs, and Re watches.

λ = K·L / √(I/A)  ·  λc = π √(E / Re)

Real columns often live in the band between those two pictures. Neither pure Euler nor pure crushing. Design codes fill that gap with empirical curves — Eurocode 3’s χ factors, AISC’s tangent-modulus and parabolic branches — all of them below Euler, because residual stress and an initial bow are not optional extras. This calculator reports both raw checks, so the governing side stays visible. It does not apply the code reduction. That reduction is a different job.

λ is already a design tool on its own. A value of 40 says the worry is yield. A value of 200 says the worry is a breath of wind. λ falls by shortening L, by raising I, or by fixing the ends. It does not fall by ordering a stronger heat of steel — λc drops as Re rises, which is why high-strength steels are more buckling-sensitive at the same shape.

Worked examples

  • Ø30 jack leg, L = 2 m, pinned–pinned — λ ≈ 267, very slender. Euler critical load ≈ 20 kN. Crushing capacity of S235 ≈ 166 kN. Buckling governs by a factor of eight. A stronger steel does not save it.
  • Building prop, 48.3 × 4 mm tube — the tube maximises I at equal mass. Against a solid round of the same kilograms per metre, the tube carries about five times more in buckling. Scaffold tubes and props are hollow for that reason.
  • Table leg, fixed–free (K = 2) — the top connection is the whole story. The same leg bolted rigid at both ends (K = 0.5) carries sixteen times more. A “welded” foot that can rock is not K = 0.5, whatever the sketch claims.

λ comes first, before a diameter is settled. If λ is already 200, 42CrMo4 is money spent on the wrong axis.

Materials that prefer to be compressed

Compression suits materials that would never carry a hanging load. Grey cast iron and concrete take five to ten times more in compression than in tension. Pillars, machine bases, foundations: millennia of that observation. A GJL-250 iron at 600 MPa in compression is a different animal from the same iron in tension, where it snaps with almost no warning.

Wood is anisotropic. This page assumes an axial load, along the grain, the favourable case. Across the grain, crushing starts at a fraction of that stress, and the “column” is a block being mashed. Glue-laminated GL24h remains timber: E around 11.6 GPa, so a slender glulam post is an Euler problem long before it is a crushing problem.

Among metals, E often outranks Re for a column. An aluminium strut (E ≈ 69 GPa) buckles at roughly a third of the load of the same steel strut (E ≈ 210 GPa), even when the two yields sit close. Euler’s load is a reminder, not a slogan: buckling is stiffness. Where aluminium is chosen for the mass it saves, diameter does the work, not the grade.

AISC 360 and Eurocode 3: Euler is a ceiling

Euler’s load assumes a perfectly straight, pin-ended or ideally fixed, centred, elastic member. Nothing that leaves a mill is that member. AISC 360 (the American steel specification) and EN 1993-1-1 (Eurocode 3) both start from Euler and then cut it. Residual stresses from rolling, an out-of-straightness of L/1000, a load that never quite hits the centroid: the code folds those into a reduction.

Eurocode 3 uses a reduction factor χ ≤ 1 on the yield load, read from buckling curves a0, a, b, c, d according to the section type and the axis. A hot-finished hollow section on curve a is kinder than a welded I-section on curve c. AISC uses a pair of formulae on the slenderness KL/r: a parabolic branch for intermediate columns, an elastic Euler branch (with a 0.877 factor) for the slender ones. Both codes give a design strength well below the Fcr printed here.

This page is the right first pass. If Euler already fails, the code will not restore the member. If Euler passes with s = 8, a χ of 0.6 leaves a margin. The danger band is s ≈ 2 on Euler: the code may well pull the factor under 1. Timber carries its own reduction in Eurocode 5; concrete columns live in Eurocode 2, with slenderness limits that do not resemble steel at all.

Fcr is a theoretical ceiling: straight member, centred load, ideal ends. Real crookedness, eccentricity and residual stress make columns fail below Euler. A pre-design factor of 2.5 to 3 belongs on the result from this page; the member is checked to AISC 360, Eurocode 3 or Eurocode 5. This calculator is not that check.

Radius of gyration and the second-moment page

i = √(I/A) asks how far the metal sits from the axis. A thin-walled tube has a large i; a solid round of the same A has a smaller one. Catalogue columns are sorted by i (or by r in AISC notation) because λ = Lf/i is the number that picks the curve.

For a rectangle the weak-axis I is h b³/12, with b the smaller side — this form already takes the minimum, which is the one Euler wants. For a catalogue IPE, HEA or a square hollow, the tables supply Iz (the smaller), not Iy. Mixing the two axes is a classic way to “prove” a 6 m post that will fold about the web.

I taken from dimensions, rather than from a table, comes from the section modulus calculator, which returns I for the usual shapes alongside W. Paired with this page: I in, Fcr out. W itself is the bending cousin — W = I/v — and does not enter Euler at all.

K on the drawing versus K in the shop

The four textbook K values (0.5, 0.7, 1.0, 2.0) assume ideal pins and fixings of infinite stiffness. AISC recommended design values are already less optimistic: 0.65 for theoretically fixed–fixed, 0.80 for fixed–pinned, 1.0 for pinned–pinned, 2.1 for fixed–free. Eurocode 3 works in buckling lengths Lcr and in restraint at the joints, not in a single magic K, and the same caution applies.

A “fixed” column base on four grouted bolts in a thin slab is not K = 0.5. A beam-to-column connection with a pair of web angles is neither a pin nor a fix. It sits between the two, and the safe reading is a K closer to 1.0 unless a joint-stiffness calculation says otherwise. Bracing slender enough to buckle on its own does not count as a pin at mid-height.

K is a judgement about the structure around the member, not a property of the tube. Where the restraint cannot be defended, the larger K is the one to keep. Capacity here does not forgive optimism: going from 0.7 to 1.0 cuts Fcr by half.

Local buckling of a thin wall

Euler describes the whole member bowing. A thin-walled tube, or a slender flange of an I-section, can crumple locally — a ripple in the wall — while the axis stays straight. That is a different instability, driven by the width-to-thickness ratio, not by L. Class 4 sections in Eurocode 3, slender elements in AISC: the effective area shrinks, and the Fcr of this page is an overestimate twice over.

A 48.3 × 4 mm scaffold tube is stocky in the wall (D/t ≈ 12) and does not local-buckle in ordinary use. A 100 × 1 mm round, with nearly three times the I, is a drinks can: better on Euler, hopeless in the wall. A large D paired with a tiny t calls for a look at D/t before Euler is trusted. Steel rules of thumb put compact round tubes around D/t ≤ 0.11 E / Re; past that line, local buckling eats the section.

Member buckling combined with local buckling is a code interaction, not a hand formula. This calculator will still print a fine Fcr for a foil cylinder. The cylinder will not honour it.

Glossary

Buckling
A sudden sideways bow of a member in compression, once the load passes a critical value. The bar leaves the straight shape, often while the stress is still below the point at which the material would yield.
Slenderness
How long a member is compared with the spread of its cross-section: λ = Lf / i. A low value means a stocky piece that crushes; a high value means a slender piece that buckles.
Euler load
The theoretical critical load Fcr = π² E I / Lf² at which a perfectly straight elastic column becomes unstable. Stiffness and length set it. Yield strength does not.
Radius of gyration
The length i = √(I/A). It says how far the material of the cross-section sits from the axis the member bends about. A larger i means a less slender member for the same length.
End-fix factor
The multiplier K in Lf = K × L. It turns the real end restraints into an equivalent pinned length. Pinned–pinned takes K = 1; fixed–free takes K = 2.
Effective length
Lf = K × L, the length of a pinned–pinned column that would buckle under the same load as the real member with its actual ends.
Second moment of area
I, the geometric measure of how hard a cross-section is to bend. In Euler’s formula the smaller of the two axes is the one that counts, because the column buckles about the weak axis.

Related calculators