Secant Formula Stress⚠ unverified
Mechanical / Columns · Compute the maximum stress in an eccentrically loaded column (secant formula)
Parameters
| Input | Symbol | Unit | Default | Description |
|---|---|---|---|---|
| P | P | N | 1.0 | Axial compressive load |
| A | A | m**2 | 1.0 | Cross-sectional area |
| M | M | N*m | 1.0 | Applied moment |
| c | c | m | 1.0 | Distance from the neutral axis to the extreme fibre |
| I | I | m**4 | 1.0 | Second moment of area of the cross-section |
| e | e | m | 1.0 | Load eccentricity |
| L | L | m | 1.0 | Unsupported length of the column |
| E | E | Pa | 1.0 | Young's modulus of the material |
| Output | Symbol | Unit | Description |
|---|---|---|---|
| result | σmax | Pa | Maximum compressive stress, in pascals (Pa). The secant term is infinite when the argument reaches pi/2 |
The science & history
Understanding the Parameters
-
Axial load $P$ — enters three ways: the direct stress $P/A$, the eccentric moment $Pe$, and the amplification argument. This triple role is what makes a beam-column nonlinear: the axial load both causes and amplifies the bending.
-
Eccentricity $e$ and applied moment $M$ — the two sources of primary bending. $Pe$ is the moment from the off-centre load; $M$ is any additional applied end moment. Their sum $(M+Pe)$ is the first-order moment that the secant then amplifies.
-
Extreme-fibre distance $c$ and inertia $I$ — convert the amplified moment into a bending stress via the flexure formula $\sigma = Mc/I$ (see Beam Bending Stress).
-
The secant argument — $\text{arg} = \dfrac{KL}{2}\sqrt{P/EI}$ is half the column's "buckling phase." At $P = P_{cr} = \pi^2 EI/(KL)^2$ the argument equals exactly $\pi/2$, where $\sec\to\infty$. So the formula predicts unbounded stress at the Euler load — the mathematical signature of buckling.
Registry note: the calculator lists $\text{arg}$ implicitly; it is the standard beam-column phase $\tfrac{KL}{2}\sqrt{P/EI}$ (equivalently $\tfrac{L}{2r}\sqrt{P/EA}$). Confirm the exact argument and whether $K$ is applied in the ToolBox implementation. Flagged in Known Issues.
Derivation (Approaching a Proof)
Model a pin-ended column of length $L$ loaded by $P$ at eccentricity $e$. The lateral deflection $v(x)$ obeys the beam-column equation with the load acting through its own eccentricity:
$$EI\,v'' + P v = -P e.$$
This is a linear ODE with constant forcing. Its solution, applying the symmetric boundary conditions, gives the mid-height deflection
$$v_{\max} = e\left[\sec\!\left(\frac{L}{2}\sqrt{\frac{P}{EI}}\right) - 1\right].$$
The maximum bending moment at mid-height is $M_{\max} = P(e + v_{\max}) = P e\,\sec(\text{arg})$ — the eccentric moment $Pe$ amplified by the secant factor. Adding the direct axial stress $P/A$ and the bending stress $M_{\max}c/I$ (with any extra applied $M$ folded into the eccentric term) gives
$$\sigma_{\max} = \frac{P}{A} + \frac{(M + Pe)\,c\,\sec(\text{arg})}{I}.$$
The secant amplification $\sec(\text{arg})$ is the exact (not approximate) moment-magnification factor for a beam-column; it reduces to the familiar $1/(1 - P/P_{cr})$ approximation for small loads and diverges at $P = P_{cr}$.
Dimensional check. $\left[\dfrac{(M+Pe)c}{I}\right] = \dfrac{\text{N}\cdot\text{m}\cdot\text{m}}{\text{m}^4} = \text{Pa}$; $\sec$ is dimensionless; $[P/A] = \text{Pa}$. ✓
History and Development
The secant formula descends from the beam-column theory developed in the 19th and early 20th centuries (Ayrton–Perry, Timoshenko). It was central to allowable-stress column design — AISC used a secant-based column formula for decades — because it rationally accounts for the unavoidable initial eccentricity of real columns through an equivalent eccentricity ratio $ec/r^2$. Modern codes have largely replaced it with empirical column curves, but the secant formula remains the clearest analytical bridge between bending and buckling.
Related Concepts: Beam Column Moment, Euler Buckling Load column, Beam Bending Stress, Slenderness Ratio, Interaction Formula, Factor of Safety
Notes: The secant argument reaches $\pi/2$ at the Euler load, where $\sigma_{\max}\to\infty$ — restrict use to $P < P_{cr}$. The classic form uses the eccentricity ratio $ec/r^2$; confirm the exact argument used by the calculator (see registry note).