10.3 Buckling, Effective Length, and Second-Order Stability

Buckling is a loss of stability, not simply crushing at high stress. A slender member can deflect laterally and lose capacity before its material reaches a short-member limit. For July 2026, use the April 2024 Civil: Structural specification, current PE Civil Reference Handbook, the AISC Steel Construction Manual 15th edition, ACI 318-14, and NDS 2018 using ASD only for wood. Do not use the April 2027 editions.

Euler Behavior and Its Limits

For an ideal straight elastic prismatic column,

P_cr = π^2EI/(KL)^2

where I is about the axis of buckling, L is the relevant segment length, and K represents rotational and translational restraint for that mode. The corresponding slenderness parameter is KL/r, with r = √(I/A). Because critical load varies with the inverse square of KL, modest restraint or length errors create large force errors.

Euler theory assumes ideal geometry, centered load, linear elasticity, and idealized end restraints. Real design includes residual stress, initial crookedness, load eccentricity, inelasticity, local buckling, connection behavior, and code resistance factors. Treat P_cr as a mechanics benchmark and stability indicator, not automatically φP_n or P_n/Ω.

Worked Effective-Length Sensitivity

A steel column segment has E = 29,000 ksi, weak-axis I = 80 in^4, area A = 20 in^2, and physical segment length L = 15 ft = 180 in. First use K = 1.0.

r = √(80/20) = 2.00 in

KL/r = (1.0)(180)/2.00 = 90

P_cr = π^2(29,000)(80)/(180^2) = 707 kips

If justified restraint changes the effective-length factor to K = 0.70, then KL = 126 in and

P_cr = π^2(29,000)(80)/(126^2) = 1,443 kips

If a sway-prone idealization justified K = 2.0, KL = 360 in and P_cr = 177 kips. These values illustrate sensitivity, not three factors to choose for convenience. The frame, end connections, bracing, axis, and selected code analysis method establish K. The AISC design strength will also reflect the applicable column curve and resistance format.

Physical, Unbraced, and Effective Lengths

Physical member length is a geometric dimension between its ends. Unbraced length is the distance between points that actually restrain the displacement or twist relevant to a mode. A 24 ft column with a qualified weak-axis brace at midheight may have a 12 ft weak-axis unbraced segment, while its strong-axis restraint pattern differs. Effective length is K times the applicable segment length and represents the ideal pin-ended length with equivalent elastic buckling behavior.

A brace counts only if it has adequate stiffness, strength, anchorage, and a complete load path. A connection that restrains translation but permits rotation affects the model differently from a moment connection. For beams, lateral-torsional unbraced length between compression-flange restraints is not a column KL; it belongs to a different buckling mode. Never copy total floor height into every unbraced-length field without identifying axis and restraint.

Local, Member, and System Buckling

Local buckling involves an individual plate element, flange, web, tube wall, or wood component distorting before the entire member bows. Width-thickness or slender-element limits affect usable section strength. Member buckling includes flexural, torsional, or flexural-torsional modes over an unbraced segment. System instability involves story sway or interaction among members and connections. Passing an Euler flexural calculation about one axis does not clear local elements, the other axis, torsional modes, bracing, or frame stability.

Concrete columns require ACI 318-14 treatment of cracked stiffness, slenderness, creep-related sustained effects, moment magnification or second-order analysis, and reinforcement. Steel uses its AISC stability method and member provisions. Wood uses the NDS ASD column and beam-column provisions with applicable adjustments. Do not mix stiffness assumptions or resistance formats among these systems.

P-Delta and P-delta

Both are second-order effects: axial load acting through a displacement creates additional moment. Capital Δ denotes global frame or story sway. If story gravity load P = 300 kips acts through story drift Δ = 0.50 in, the geometric moment scale is

PΔ = (300)(0.50) = 150 kip-in = 12.5 kip-ft

Lowercase δ denotes local curvature between member ends. If the same axial load acts through a local bow δ = 0.20 in,

Pδ = (300)(0.20) = 60 kip-in = 5.0 kip-ft

These products illustrate the source of amplification; a code analysis does not necessarily add these two isolated numbers directly. A nonsway or braced frame can still have P-delta member effects. A sway frame may have both. As displacement grows, secondary moment increases, which can cause more displacement and approach instability.

Consistent Stability Workflow

1. Define the frame, load combinations, axes, and actual restraint points.
2. Identify local, member, and system modes.
3. Select the AISC, ACI, or NDS analysis route before assigning stiffness and length factors.
4. Include required imperfections, notional loads, stiffness reductions, cracking, or sustained-load effects.
5. Perform first- or second-order analysis as that route permits, then use its compatible member resistance equations.
6. Check every brace and connection that the model relies on.

Do not combine a direct-analysis stiffness reduction with an unrelated effective-length assumption merely to improve capacity. Finish by labeling physical length, unbraced length for each axis, K, and the exact KL used. Stability depends on that chain of assumptions as much as on arithmetic.

An eigenvalue buckling result also requires interpretation: its mode shape identifies a possible instability pattern, while its ideal load factor is not automatically a code design strength.