Yield-Line Theory: How a Concrete Slab Chooses Its Collapse Pattern

The core idea: turn a plate into a folding mechanism

A reinforced concrete slab is a plate — a two-dimensional bending element carrying load by moments in two directions plus twisting moment. Solving the elastic plate governs the biharmonic equation ∇⁴w = q/D, where D = Et³/[12(1−ν²)] is the flexural rigidity, t the thickness, and ν Poisson's ratio (≈0.2 for concrete). That is genuinely hard for anything but a rectangle with idealized edges.

Yield-line theory sidesteps it entirely. At the ultimate limit state, an under-reinforced concrete section behaves rigid-plastic: it carries almost no rotation until the steel yields, then rotates freely at a nearly constant moment — the plastic moment of resistance m (units kN·m per metre of width). Johansen's insight: at collapse the slab breaks into rigid panels separated by lines along which the reinforcement has yielded. These yield lines are plastic hinges smeared into straight creases. The panels stay flat; all the rotation and all the energy dissipation happens on the lines.

• Yield lines pass through the intersection of the axes of rotation of adjacent panels — a purely geometric rule.
• A yield line under a supported edge lies along that support; over a continuous edge it forms a negative (hogging) yield line where top steel yields.
• The pattern must convert the slab into a mechanism with one degree of freedom, so the whole thing descends with a single deflection parameter δ.

The work equation — one energy balance solves the slab

Because the collapse mechanism has one degree of freedom, the principle of virtual work gives the capacity directly. Impose a virtual downward displacement δ = 1 at a control point and equate the work the load does to the energy the yield lines dissipate:

External work = Internal work → Σ (w · A · δ̄) = Σ (m · θ · L)

On the left, w is the ultimate distributed load (kN/m²), A the plan area of each panel, and δ̄ the deflection of that panel's centroid. On the right, for each yield line, m is the moment of resistance per unit length, θ the relative rotation across the line, and L its projected length. The projection trick makes this tractable: rather than integrating rotation along a skewed line, you sum m·θ·L using the components projected onto the axes of rotation, so a diagonal yield line contributes as its x- and y-projections.

• Step 1 — assume a kinematically valid pattern with unknown geometry parameters (e.g. the distance to a diagonal intersection).
• Step 2 — write external work as Σw·(volume of the deflected panel prism) — for a triangular panel rotating about its base, δ̄ = δ/3.
• Step 3 — write internal work Σm·θ·L, with θ ≈ δ/(perpendicular distance from line to axis of rotation) for small rotations.
• Step 4 — set the two equal and solve for w (or, in design, for the required m).
• Step 5 — minimize w with respect to the geometry parameters (∂w/∂x = 0). Because the method is an upper bound, the true collapse load is the lowest load over all patterns, so you differentiate to find the critical geometry.

Worked capacity: the square and rectangular slab

Take a square slab, side L, simply supported on four edges, isotropically reinforced (same m each way, each face as needed). The failure pattern is the classic diagonal 'envelope': four triangular panels folding about the four supports and meeting at the centre. Running the work equation with δ = 1 at the centre gives the tidy closed form:

m = w·L²/24

So a 6 m square slab carrying an ultimate w = 12 kN/m² needs m = 12 × 36 / 24 = 18 kN·m/m in each direction. For a general rectangle Lₓ × L_y (Lₓ ≤ L_y) with isotropic m, minimizing over the pattern geometry yields Johansen's result:

m = (w·Lₓ²/24) · [√(3 + (Lₓ/L_y)²) − Lₓ/L_y]²

which collapses to w·L²/24 when Lₓ = L_y. Real slabs are usually orthotropic — different reinforcement each way, m_y = µ·m. An affine transformation (scale one span by 1/√µ) maps the orthotropic problem back onto an equivalent isotropic one, so the same formulae serve. Typical numbers: a 200 mm slab, C30/37 concrete (f_ck = 30 MPa), with H12 bars at 150 mm (A_s ≈ 754 mm²/m, d ≈ 165 mm) develops m ≈ 0.87·f_yk·A_s·z ≈ 0.87 × 500 × 754 × 0.95 × 165 × 10⁻⁶ ≈ 51 kN·m/m — comfortably above the 18–25 kN·m/m a domestic floor demands.

Why it's an upper bound — and the safety trap that follows

Plastic theory has two bounding theorems. The lower-bound (static) theorem says any load supported by a statically admissible stress field that nowhere violates yield is safe. The upper-bound (kinematic) theorem — which yield-line theory embodies — says any load computed from a valid collapse mechanism is greater than or equal to the true collapse load. In plain terms: a guessed yield-line pattern always overestimates strength. The real slab will find the weakest mechanism nature allows, which may not be the one you drew.

This is the method's one genuine hazard. Miss a mechanism — a fan pattern under a concentrated load, a corner lever, a partial-width mechanism at a re-entrant corner — and your calculated m looks adequate while the slab is not. Best practice guards against it:

• Always check corner levers (corner fans) at simply supported corners; they can reduce capacity by 8–10% versus the straight-diagonal pattern, so slabs are often reinforced against corner uplift or the geometry is refined.
• Check local fan mechanisms around columns and point loads, where circular yield lines radiate and the punching-type collapse can govern.
• Because it's plastic, verify ductility: the section must reach its rotation without concrete crushing first, i.e. keep it under-reinforced with neutral-axis depth xᵤ/d ≲ 0.25 (EC2 redistribution) so the steel yields and holds moment while rotation accumulates.
• Cross-check against a lower-bound method (Hillerborg's strip method or an elastic FE moment field) for anything unusual — the strip method is inherently safe and complements yield-line's optimism.

Controlling variables, orthotropy, and the design trade-offs

The whole method turns on a handful of levers. Understanding how each moves the collapse load is what separates a fast, safe design from an over-optimistic one.

• Moment of resistance m — set by steel area A_s, yield strength f_yk (typically 500 MPa), and lever arm z ≈ 0.9d. Required capacity scales as m ∝ w·L², so span dominates: doubling the span quadruples the demanded m.
• Support conditions — fixing an edge adds a negative yield line and roughly doubles capacity versus simple support (for a fixed-fixed one-way strip, the collapse moment is w·L²/16 shared between span and support, versus w·L²/8 at midspan simply supported). Continuity is the cheapest strength you can buy.
• Orthotropy ratio µ = m_y/mₓ — lets you tune reinforcement to the aspect ratio. For a long rectangle you put more steel across the short span; the affine transform handles the maths, but the trade-off is congestion versus efficiency.
• Aspect ratio Lₓ/L_y — as a rectangle stretches, behaviour shifts from two-way (diagonal pattern) toward one-way (single central yield line), and the m·L²/24 coefficient migrates toward the one-way L²/8 regime.

The central trade-off: yield-line theory trades an unsafe optimism for enormous speed and design freedom. It gives no deflection, no crack width, no service stress — so it is an ultimate-limit-state tool only. You still owe a separate serviceability check (span/depth ratio, EC2 gives ~26 for a continuous two-way slab) and often an elastic analysis for cracking.

Where it earns its keep: real slabs and hardware

Yield-line analysis is the working method behind an astonishing range of real concrete, precisely because it eats irregular geometry that defeats tables and closed-form elastic solutions.

• Irregular and skew floor slabs — L-shaped bays, slabs with re-entrant corners, triangular panels, slabs supported on scattered columns. You just draw a plausible fold pattern and solve; no rectangular-slab table required.
• Flat slabs and flat plates — the folding-plate / fan mechanisms around interior and edge columns are yield-line problems, and the method underlies the moment coefficients in design codes.
• Slab-on-column punching interaction — combined flexural fan and punching cone define the true failure envelope of a flat slab; a punching shear reinforcement layout is checked against a fan yield-line mechanism.
• Bridge deck and blast/impact assessment — the rigid-plastic idealization is ideal for accidental and extreme loads where full plastic redistribution is credible; UK highway slab decks and protective structures use it directly.
• Assessment of existing slabs — because it captures reserve strength beyond elastic first-yield, yield-line analysis frequently 'passes' older slabs that an elastic check fails, avoiding needless strengthening. This is one of its most valuable modern uses.

The 'hardware' is deliberately dumb: straight yield lines, rigid panels, a deflection δ, projected lengths, and rotations. That austerity is exactly why an engineer can hand-solve a slab that would otherwise need a finite-element model.

Failure modes, limits, and best practice

The theory has sharp boundaries, and respecting them is the difference between a valid analysis and a dangerous one.

• Brittle (over-reinforced) failure invalidates it. Yield-line theory assumes the section rotates plastically at constant m. If the slab is over-reinforced, the concrete crushes at xᵤ/d beyond ~0.45 before the steel yields, moment redistribution never happens, and the plastic mechanism cannot form. Keep slabs under-reinforced (ρ well below balanced, ρ ≈ 0.4–1.0%).
• Membrane action — a hidden reserve you should not bank blindly. At large deflection, restrained slabs develop compressive then tensile membrane action that can lift capacity 1.5–3× above the yield-line value. It is real (exploited in composite floor fire design) but requires horizontal restraint; unrestrained edges give none.
• Anticlastic and corner effects. Twisting moments at corners lift the slab; without corner reinforcement the corner-lever mechanism governs and top steel must be provided in the corners.
• No serviceability information. A yield-line design can be strong yet crack and deflect excessively. Always pair it with a deflection/crack check.

Best-practice recipe: (1) sketch every plausible mechanism including corner and fan patterns; (2) solve and take the lowest collapse load / highest required m; (3) minimize over free geometry parameters; (4) confirm ductility (under-reinforced, adequate rotation capacity); (5) sanity-check against a lower-bound strip solution; (6) do serviceability separately. Follow that and yield-line theory is what it was in 1943 — the fastest honest way to know when a concrete slab will fold.

Yield-line (plastic) analysis versus elastic plate analysis for RC slabs

Attribute	Yield-line theory	Elastic plate theory
Bound on true capacity	Upper bound (≥ true, can be unsafe)	Serviceability / lower-bound moments
Governing equation	Work balance Σw·δ = Σm·θ·L	∇⁴w = q/D (biharmonic)
Output	Ultimate collapse load / required m	Elastic moment field mₓ, m_y, mₓy
Material assumption	Rigid-plastic, full moment redistribution	Linear elastic, uncracked or cracked-transformed
Effort for irregular slab	Minutes (guess pattern, solve)	FE mesh / tables, hours
Deflection & cracking info	None (collapse only)	Direct (needed for SLS)