Spreader beam or lifting beam: which one your lift actually needs

Both devices are a beam that hangs under a crane hook and spreads a load. They are hired from the same yards, stored in the same racks, and drawn almost identically on a lift plan. They fail in completely different ways.

A spreader beam has slings running from the crane hook down to a lift point at each end. Those slings are not vertical, so each one pulls inwards as well as up. The beam's job is to hold the two ends apart against that inward pull, so its primary action is axial compression along its length.

A lifting beam hangs from a single point above it, or from a short bridle close to the top, and the load hangs from points along its length. Nothing is pulling the ends together. The beam is a simple beam: the load hangs off it, the support is somewhere else, and the difference between them is bending moment.

That is the whole distinction, and every practical consequence follows from it.

Take the vertical share that a leg has to carry. The tension follows from the angle, and the inward pull follows from the tension.

Sixty degrees is comfortable. Forty-five is where people start paying attention. Thirty is where most rigging practice draws the line, and below it the relationship turns hostile: the sine term is shrinking while the cosine term is growing, so the horizontal component the beam has to carry accelerates away.

The worked example makes it concrete on one beam and one load.

Now flatten the slings and change nothing else.

The load never changed. The leg tension went up by 84 percent, the compression in the beam by more than three times, the declared sling-angle limit was breached, and the first strength check to fail was not on the beam at all but the weld attaching its padeye.

That last point generalises. Flattening the slings loads the attachment harder than it loads the beam, because the attachment sees the full leg tension at a worse angle while the beam sees only the horizontal component.

The beam in the example is a circular hollow section, and that is not an arbitrary default.

A spreader is a long strut. Its capacity in compression is governed by buckling, and buckling happens about whichever axis is weakest. Take an I-section, which is superb in bending about its major axis, and stand it up as a strut: its minor-axis stiffness is a small fraction of its major-axis stiffness, and that small fraction is what decides the capacity. Restraining it is awkward, because a spreader hangs in the air with slings at each end and no bracing to attach to.

A hollow section has the same stiffness in every direction. There is no weak axis to find, no lateral-torsional buckling mode to check, and no restraint to argue about. For a member whose job is to be squeezed along its length while hanging free, that is the right answer, and it is why the yards are full of tubular spreaders.

The reverse is true for a lifting beam. There the action is bending, an I-section is efficient at bending, and the awkward mode is lateral-torsional buckling of the compression flange. So the two devices tend towards different sections for the same reason: each one is shaped around the action it actually carries.

One thing a spreader is not is a pure strut. The sling pin sits above the beam axis, so the inward horizontal force acts at an eccentricity and hands a moment into the beam. In the 60 degree example that transfer couple is the entire source of the 8.2 kN·m in the beam, and at 30 degrees it grows to 18.3 kN·m. A calculation that models a spreader as axial force only is missing a real action, and it is missing more of it exactly when the geometry is worst.

Take the identical beam and the identical 10 t load, and hang it from one point at midspan instead.

The moment went from 8.2 kN·m to 151.5 kN·m on the same span with the same load. That is the price of bending, and it is why nobody builds a 20 m lifting beam.

Look also at what got easier. The top sling is vertical, so there is no inward pull, no transfer couple, and the padeye weld that failed at 30 degrees is now barely worked. A lifting beam trades a hard beam problem for an easy attachment problem, and a spreader does the reverse.

Work through it in this order.

Do you have height above the load? A spreader needs the sling legs above it to develop the geometry, and at a sensible angle that height is roughly the same order as the half-span. If the hook cannot get that high, the choice is already made.

How long is the span? Bending gets expensive with span far faster than compression does. Past a few metres the weight difference between the two devices stops being a detail.

Where can the load actually be attached? A lifting beam can pick up at points along its length, including asymmetric ones, and can be trimmed by moving its top lug. A spreader wants its load at the two ends. A load with awkward pick points may not suit a spreader at all.

Is the load being turned, tilted or landed on an angle? Both devices are analysed differently once the beam is not level, and an arrangement that is fine level can be a different structure at 15 degrees of tilt.

What is the consequence of getting it wrong? A spreader that loses a sling angle assumption overloads its attachments quietly, as the example showed. A lifting beam that is overloaded shows it in deflection first, which people notice. That is not a reason to prefer one, but it is a reason to be stricter about the assumptions on a spreader.

Canada is the interesting case, and it comes up the moment somebody asks for a spreader beam designed to CSA. There is no Canadian below-the-hook device standard. A CSA route therefore has to take its demand-side load factor as a declared project value, because there is no document to take it from. That is not a gap in a calculator; it is a gap in the standards landscape, and the honest response is to make whoever owns the project declare the factor rather than to quietly borrow one from a country with different rules.

Offshore, a marine warranty scope will usually impose its own factors on top of whichever route you are using, and those are not optional.

1. A spreader analysed as a simple beam. It puts the moment in and leaves the compression out, which is backwards on both counts. The compression is the primary action and the moment comes mostly from the pin eccentricity, not from the payload.

2. A lifting beam analysed as a strut. The mirror image, and rarer, but it happens when a spreader spreadsheet gets reused.

3. The sling angle taken from the drawing rather than the solve. The worked example lost 2.2 degrees to lug eccentricity alone, and landed below the 30 degree floor while the drawing said 30.

4. The pin eccentricity ignored. Modelling a spreader as pure axial force removes a real moment, and removes more of it precisely when the sling angle is worst.

5. The attachment assumed to follow the beam. Flattening the slings failed the padeye weld while the beam was still comfortable. The beam and its attachments do not degrade together, because they see different components of the same force.

6. A device standard assumed to cover the steelwork. EN 13155 gives you the load basis, not member resistance. A calculation that cites it alone has not checked the beam against anything.