Section Modulus and Beam Design

Why section modulus matters

Almost every horizontal structural member — a floor joist, a bridge girder, a crane boom, an aircraft wing spar — earns its keep by resisting bending. When you load a beam, one face is squeezed and the opposite face is stretched, and the material in between transitions smoothly from compression to tension. Somewhere across the depth lies a surface that is neither stretched nor squeezed: the neutral axis, which for a symmetric homogeneous section passes through the centroid. Bending stress grows linearly with distance from that axis and peaks at the outermost fiber. Section modulus is the single geometric number that packages "how far away the worst fiber is" together with "how the area is distributed," so that a designer can turn a bending moment straight into that worst-case stress.

The practical payoff is that beam selection collapses to one inequality. Compute the largest moment the loads impose, divide by the stress you are willing to accept, and you have the section modulus you need. Every steel handbook — the AISC Steel Construction Manual, the Eurocode section tables, the aluminium and timber design guides — lists S and Z for hundreds of rolled and extruded shapes precisely so you can read off a member that clears your required value. Get the section modulus right and the beam is safe; get it wrong and you are either wasting steel or courting a yield failure.

• Buildings. Floor beams and roof purlins are sized so bending stress under service load stays below the allowable; W-shapes dominate because their S per kilogram is unbeatable.
• Bridges. Plate girders and box girders are proportioned by section modulus, then checked for shear, fatigue, and lateral-torsional buckling.
• Machines. Shafts, arms, and frames use S (about the bending axis) to keep working stress inside the endurance limit.
• Aerospace. Wing and fuselage spars are built up with caps far from the neutral axis to maximize S while shaving mass.
• Timber and cold-formed steel. Joist and stud tables are indexed by section modulus and span.

How it works, step by step

The whole edifice rests on the Euler–Bernoulli bending relation. For a beam in pure bending, plane sections remain plane, so the longitudinal strain at a fiber a distance y from the neutral axis is proportional to y. Multiply by the elastic modulus E and the bending stress at that fiber is:

σ(y) = M·y / I

where σ(y) is the normal (bending) stress at height y, in pascals (Pa) or psi; M is the internal bending moment, in newton-metres (N·m) or lb·in; y is the perpendicular distance from the neutral axis, in metres or inches; and I is the second moment of area about the neutral axis, in m⁴ or in⁴. The stress is largest at the extreme fiber, where y reaches its maximum value c. Setting y = c and defining the elastic section modulus S = I/c collapses the expression to the working formula every engineer memorizes:

σ_max = M·c / I = M / S, with S = I / c

Here S has units of length cubed (mm³ or in³) because it is a length⁴ quantity divided by a length. The four steps in practice are:

1. Locate the neutral axis. For a homogeneous section it is the centroid. Compute the centroid by area-weighting the component shapes.
2. Compute I about that axis. Use the parallel-axis theorem, I = Σ(I_c + A·d²), summing each part's own inertia plus its area times the square of its distance to the neutral axis.
3. Divide by c. For a symmetric section, c = half the depth and there is one S. For an unsymmetric section (a tee, a channel), the top and bottom fibers sit at different distances, so you get S_top = I/c_top and S_bottom = I/c_bottom; the smaller S governs.
4. Convert moment to stress. σ = M/S. Compare against the allowable stress, or invert to size the member: S_req = M/σ_allow.

The critical insight lives in step 2. Because inertia weights area by distance squared, a thin flange placed at the edge of the section contributes wildly more to I than the same area placed near the middle. That is why the depth of a beam is so powerful — and why hollowing out the middle costs almost nothing.

Elastic vs plastic section modulus

The elastic modulus S assumes a linear stress distribution in which only the outermost fiber has just reached the yield stress F_y. The moment that produces first yield is the yield moment M_y = S·F_y. But a ductile steel beam does not fail at first yield. As the moment grows, yielding spreads inward from both faces until the entire section is at yield — a fully developed plastic hinge. In that limit the stress block is uniform at F_y in compression above the plastic neutral axis and F_y in tension below it. The moment carried is the plastic moment M_p = Z·F_y, where the plastic section modulus Z is the first moment of the two equal half-areas about the plastic neutral axis:

Z = A_c·ȳ_c + A_t·ȳ_t

with A_c, A_t the compression and tension half-areas (each equal to half the total area for a solid section) and ȳ_c, ȳ_t the distances from their centroids to the plastic neutral axis. Z is always larger than S, and their ratio is the shape factor k = Z/S. It quantifies the reserve strength between first yield and collapse. Modern steel codes (AISC 360 LRFD, Eurocode 3 for compact sections) design directly on Z: the nominal flexural strength is M_n = Z·F_y, applied with a resistance factor φ = 0.90. Allowable-stress (ASD) practice and brittle or slender sections that cannot reach the plastic hinge still design on S.

Shape efficiency: comparing cross-sections

The table below shows why the I-beam won the structural-steel century. All values are for bending about the strong (horizontal) axis. "Efficiency" here is a rough S-per-unit-area figure — how much bending strength you buy per kilogram of material — normalized so the solid rectangle is 1.0.

Cross-section	Elastic S	Plastic Z	Shape factor Z/S	Relative S-per-area
Solid rectangle (b×h)	b·h²/6	b·h²/4	1.50	1.0
Solid circle (dia d)	π·d³/32	d³/6	1.70	0.83
Thin circular tube	≈ π·r²·t	≈ 4·r²·t	1.27	≈ 2–3×
Rolled I-beam (W-shape)	tabulated	tabulated	1.12–1.18	≈ 3–4×
Rectangular hollow box	tabulated	tabulated	≈ 1.1–1.3	≈ 2–3×

The lesson from the last column is that hollowing and spreading the material outward multiplies bending strength for the same weight. The I-beam takes this to its logical end: two flanges as far apart as buckling and manufacturing allow, joined by a thin web that mainly carries shear. Its shape factor is low precisely because it is already efficient — nearly all its area is at the extreme fiber, so first yield of the flange is very close to full plastic capacity; there is little "lazy" central material left to recruit.

Worked example: sizing a floor beam

A simply supported steel floor beam spans L = 6 m and carries a uniformly distributed load w = 20 kN/m (factored). The maximum bending moment for a simply supported beam under uniform load is:

M_max = w·L² / 8 = 20 × 6² / 8 = 90 kN·m

Using A992 steel with yield stress F_y = 345 MPa and an allowable-stress approach with an allowable bending stress of σ_allow = 0.66·F_y ≈ 228 MPa, the required elastic section modulus is:

S_req = M_max / σ_allow = 90 × 10⁶ N·mm / 228 N/mm² ≈ 3.95 × 10⁵ mm³ = 395 cm³

From the steel table, a W250×33 (nominal 250 mm deep, 33 kg/m) provides S_x ≈ 380 cm³ — just short — while a W310×39 provides S_x ≈ 549 cm³ with room to spare, so the W310×39 is selected. Notice the depth-squared payoff: the deeper 310 mm section clears the requirement easily at only marginally more weight, because S rises with the square of depth. In LRFD you would instead check φ·M_n = 0.90·Z·F_y ≥ M_max, using the plastic modulus Z, and then verify deflection (typically span/360 for live load), shear on the web, and lateral-torsional buckling of the unbraced compression flange before the design is final.

Common misconceptions and failure modes

• Confusing S with I. Moment of inertia I (length⁴) governs deflection through EI; section modulus S (length³) governs stress. A stiff beam is not automatically a strong one — check both.
• Ignoring the weak axis. An I-beam's S about its weak axis is a fraction of its strong-axis value. Load it the wrong way, or let the compression flange buckle sideways, and it fails far below the tabulated capacity.
• Assuming symmetry. For a tee or channel the neutral axis is not at mid-depth; the fiber farther from it yields first, so the smaller of S_top and S_bottom controls.
• Forgetting lateral-torsional buckling. A high section modulus is useless if the beam twists and buckles before it yields; the unbraced length must be checked separately.
• Using Z when the section is not compact. Slender flanges or webs buckle locally before the plastic hinge forms, so the plastic modulus overstates capacity; only compact sections reach M_p.
• Neglecting shear near supports. Section modulus addresses bending stress only; short deep beams and points near reactions can be shear-critical.
• Deflection failure. A beam sized purely for strength may still sag past serviceability limits; deflection often governs long spans and light live loads.