IPE vs HEA vs HEB: How to Choose the Right European I-Beam

IPE, HEA, HEB and HEM are the workhorse families of European hot-rolled I-sections. They share one silhouette (two parallel flanges joined by a web), but they distribute steel very differently. IPE is the narrow-flange family: a deep, slender web with flanges roughly half as wide as the section is deep. HEA, HEB and HEM are wide-flange sections with an almost square footprint: for HEA and HEB the flange width equals the nominal size through the small and mid range, and HEM is a few millimetres wider still.

One trap for the unwary: the nominal size equals the actual depth only for HEB. An HEA is slightly shallower than its nominal size and an HEM is noticeably deeper. The flange width is shared by HEA and HEB alone; HEM is wider than both, by 6 mm at nominal size 200 and by 10 mm at nominal size 300. When you swap families during design development, recheck the true depth and the true weight, never just the name.

For a floor beam or rafter bending about its strong axis, the property you are paying for is the major-axis inertia I_y and the section modulus W_el,y or W_pl,y. Bending stiffness grows with the square of the distance of material from the neutral axis, so a deep, light section beats a compact, heavy one. That is exactly what IPE is optimised for: a thin web holding two flanges as far apart as possible, with almost no steel wasted near mid-depth.

Read that table carefully, because the same nominal size is the wrong basis for this comparison. At size 200 the wide-flange sections do carry far more steel per metre, but they are also deeper or stockier, so their stiffness per kilogram comes out slightly higher, not lower: about 87 cm⁴ per kg/m for both IPE 200 and HEA 200, 93 for HEB 200 and 103 for HEM 200.

What you actually buy a beam for is a required stiffness or a required bending resistance, and on that basis IPE wins decisively: the weight it saves on flange width it spends on depth. Matching the I_y of an IPE 300 at 42.2 kg/m takes an HEA 260 at 68.2 kg/m or an HEB 240 at 83.2 kg/m. Matching its W_pl,y takes an HEA 240 at 60.3 kg/m or an HEB 200 at 61.3 kg/m. That is why laterally restrained purlins, floor beams and composite beams are almost always IPE. Remember too that every structural steel shares E = 210 000 MPa: where deflection governs, a higher grade buys nothing. Only geometry helps.

Flip the loading and the ranking flips with it. A column fails by flexural buckling about its weak axis, so the governing properties are the minor-axis inertia I_z and the radius of gyration i_z that sets the slenderness entering the EC3 buckling curves. Narrow IPE flanges mean a small i_z, a high slenderness and a punishing reduction factor χ, so little of the squash load N_pl = A·f_y/γ_M0 survives the buckling check. Wide HE flanges keep the weak-axis slenderness low and the reduction mild.

In practice, HEA is the economical first pick for ordinary building columns. Its weak-axis inertia I_z is lower than the HEB's roughly in proportion to its weight, but the property that drives the buckling check is the radius of gyration, and i_z stays within about two percent of the HEB of the same size. Slenderness and the reduction factor χ therefore come out almost the same, while the section weighs 20 to 30 % less per metre. HEB (for example an HEB 300) earns its extra steel when the member sees genuine biaxial demand: columns of unbraced frames carrying moment, crane runway girders with horizontal surge loads, or any beam bent about both axes, where the weak-axis modulus of an IPE would be hopeless.

A beam bent about its strong axis wants to escape sideways: the compression flange deflects laterally while the section twists. The elastic critical moment M_cr grows with the minor-axis inertia I_z, the torsion constant I_t and the warping stiffness I_w, precisely the properties an IPE has traded away for depth. An unrestrained IPE therefore loses a large share of its bending resistance to lateral-torsional buckling, and the loss grows quickly with unrestrained length.

• If the compression flange is continuously held (by a composite slab, decking fixed to the top flange, or closely spaced secondary members), use IPE and take the weight saving.
• If restraint is sparse (long spans between purlins, cantilevers and hogging regions where the bottom flange goes into compression, or members standing alone during erection), a wide-flange HEA or HEB often ends up lighter than the IPE it replaces once the reduction factor χ_LT is applied.
• Always check the erection stage: many IPE beams that are fine in service are critical before the deck is fixed.

Fabricators often hold the casting vote, and it usually favours the wide-flange families:

• Bolting. A wide flange gives room for standard gauges, larger bolts and two bolt lines where needed; on narrow IPE flanges the geometry of end plates and flange cover plates gets tight quickly.
• Column faces. In bolted moment end-plate joints designed by the EN 1993-1-8 component method, a thin column flange in bending is a classic weak component. Thick HEB or HEM flanges frequently eliminate stiffeners and backing plates that an HEA column would need.
• Webs. The slender IPE web is sensitive to concentrated transverse forces (reactions, purlin cleats, temporary supports) and may need bearing stiffeners per EN 1993-1-5; the stockier HE webs rarely do.
• Handling. IPE beams are more flexible about the weak axis during transport and lifting; wide-flange sections sit flat, stack well and tolerate site abuse.

A slightly heavier section that removes a dozen welded stiffeners is often the cheaper structure.

HEM is the heavyweight sibling of HEB: the same wide-flange geometry with a much thicker web and flanges, and a depth above the nominal size. It is rarely the economical answer, but in a few situations it is the right one:

• Heavily loaded columns where the floor plan fixes the column footprint: HEM packs the largest area, and therefore the largest N_pl, into a given envelope.
• Capacity design in seismic frames, where a deliberately strong, stocky column with thick flanges satisfies strong-column/weak-beam checks and panel-zone demands without doubler plates.
• High steel grades. Cross-section class limits scale with ε = √(235/f_y), so parts that are Class 1 in S235 can slip a class in S460. The stocky HEM outline stays plastic where slenderer families would not.
• Fire. A low section factor (massive steel, little exposed perimeter) heats slowly, which can reduce or simplify fire protection.

Set against this: a high price per metre, longer mill lead times for the rarer sizes, and heavier lifts. Specify HEM deliberately, not as a lazy upgrade from HEB.