Buckling is a stiffness failure, not a strength one
A short column fails by CRUSHING — it reaches the material's compressive strength and gives way. A long one fails by BUCKLING: it bows sideways at a load well below its crushing capacity, and the material never comes close to its strength.
Euler's critical load describes the second, and it contains no strength term at all. It is built from the elastic modulus, the second moment of area and the length — stiffness and geometry only. That is why a higher-strength material does not make a slender column carry more, exactly as a higher-grade steel does nothing for deflection.
Between the two extremes lies a transition where both matter, and real design curves are empirical blends of the crushing limit and the Euler one, reduced further for the imperfections every real member has: initial crookedness, residual stresses from rolling or drying, and load that is never perfectly centred.
- P_cr
- elastic critical buckling load — no strength term in it
- k
- effective length factor, set by the END CONDITIONS
- L
- actual unbraced length
- r
- radius of gyration, the square root of I over A
- λ
- slenderness ratio — the number that says which failure mode applies
The ends are worth more than the member
The effective length factor converts a real column into the equivalent pin-ended one Euler's expression describes, and it depends entirely on what restrains each end. A member fixed at both ends has a factor near a half; pinned at both ends, one; fixed at one end and free at the other — a cantilever — two.
Because the factor is SQUARED, that range is a factor of about sixteen in critical load between the best and worst end conditions on an identical member. Nothing else on this page moves the answer that far.
It is also the assumption most often taken optimistically. A base plate with four bolts looks fixed and behaves as something between fixed and pinned; a beam framing into a column flange restrains one axis and not the other. Codes give theoretical values and then recommend higher design values precisely because real restraint is never as good as the sketch, and a column designed on a theoretical fixed-fixed factor is designed on a condition that does not exist.
Height-to-thickness rules are slenderness in disguise
Masonry walls, parapets and shaftwalls are commonly checked against a simple ratio of height to thickness rather than against a buckling calculation, and the ratio is a proxy for the same physics. A thin wall is a slender member in one direction, and the limiting ratio is a conservative stand-in for the full check.
The proxy's value is that it is checkable on a drawing in seconds, and its cost is conservatism and silence about the end conditions. A parapet is the sharpest case: it is a CANTILEVER above the roof, with no restraint at its top and wind able to act on both faces, so its permitted ratio is much lower than that of the wall below it — which is propped at every floor.
For a masonry prism the same ratio has a different job. A short stocky prism tests stronger than a tall slender one of identical masonry, so the measured strength is corrected back to a standard proportion — the ratio here is a correction factor rather than a limit, and it is the reason the prism's as-built height has to be measured rather than assumed.
An arch's rise-to-span is the same kind of number
An arch carries load in compression along a curved line, and the flatter the curve the harder it pushes sideways at its springings. The horizontal thrust is inversely proportional to the RISE: halve the rise of an arch carrying the same load and you double the outward thrust on its abutments.
So a semicircular arch is nearly self-contained — its thrust is modest and largely vertical — and a shallow segmental arch over a wide opening delivers a horizontal force that must be resisted by something: a buttress, a substantial return, a tie rod, or the mass of the wall alongside. An arch that works structurally and has nowhere to put its thrust is not an arch, it is a slow-motion failure.
This is why the rise-to-span ratio is quoted as a proportion rather than a dimension, and why jack arches and flat arches were always built with iron ties or into heavy masonry. It is the same conceptual move as the height-to-thickness rule: a single ratio standing in for a force calculation that would otherwise need the whole structure.
Where it fails, and the alternative
The buckling load above assumes an axially loaded, initially straight, perfectly elastic member. Every real one violates all three, which is why design codes do not use Euler's expression directly — they use curves fitted to tests that already include imperfections, and the calculation here is the idealised reference rather than a design value.
The other failure is applying a slenderness check about the wrong axis. A member buckles about its WEAKEST axis unless something restrains it, and bracing that restrains one axis and not the other changes which slenderness governs. A stud braced by sheathing in one direction and by nothing in the other is two different columns depending on which way you look at it.
The alternative to all of this is a second-order analysis that models the deflected shape directly and finds the load at which it runs away. It is standard in analysis software and it is the honest answer wherever the restraint conditions are ambiguous — which, given the paragraph above about end conditions, is more often than the sketches suggest.
Calculators that use this method
Basis
- Euler, L. (1744), on the buckling of elastic columns — the origin of the critical load expression above.
- AISC 360, Chapter E and Appendix 7: effective length factors, the recommended design values against the theoretical ones, and the direct analysis method as the second-order alternative.
- National Design Specification for Wood Construction (NDS), Chapter 3 — the column stability factor, which blends the crushing and buckling limits empirically.
- TMS 402 / ACI 530, height-to-thickness limits for masonry walls and the slenderness reduction on axial capacity.
- ASTM C1314 and TMS 602 for masonry prism testing, including the height-to-thickness correction applied to measured strengths.
- Heyman, J., The Stone Skeleton (1995), on arch thrust, the rise-to-span relationship and the line of thrust within the masonry.