Deep Beams And Corbels

Ordinary beam theory assumes the section stays flat as it bends and the stresses vary linearly from top to bottom.

In a beam whose depth is large relative to its span, that assumption fails. The load does not travel along the member as bending; it travels directly to the support as a diagonal compression strut, held in equilibrium by a horizontal tension tie.

Designing that member with a bending diagram puts the steel in the wrong place.

When a beam becomes a deep beam

Roughly when the span-to-depth ratio drops below about two for a simply supported member, with a transition zone either side of that.

Which happens more often than people expect.

A transfer beam carrying columns over a wide opening, deliberately made deep for stiffness. The behaviour sits alongside the work in [transfer slabs and transfer beams](/blog/transfer-slabs-and-transfer-beams).

A pile cap spanning between two or more piles, which is almost always a deep member and is almost never analysed as one.

A wall acting as a beam over an opening, where the full height of the wall is participating.

A beam with a very short span, such as a lintel over a narrow opening in a thick wall.

A corbel or a nib, which is the shortest cantilever version of the same behaviour.

Strut and tie, which is the right model

The member is resolved into a truss: compression struts running diagonally from the load to the supports, tension ties running horizontally, and nodes where they meet.

That model tells you three things a bending analysis does not.

Where the tension steel goes, which is the horizontal tie across the bottom, carried all the way to the supports and anchored there. Not curtailed toward the ends the way a bending diagram would suggest, because the tie force is close to constant along its length.

How much the node can carry, because the concrete at the support, where the strut lands and the tie anchors, is under triaxial stress and it is the most highly loaded point in the member. Node capacity frequently governs rather than the strut or the tie.

That web reinforcement is structural, in both directions, to control the splitting of the strut and to distribute cracking. A deep beam with nominal web steel cracks diagonally.

Anchorage is the usual failure

The tie has to develop its full force at the support, and in a deep beam there is very little length in which to do it.

Which makes the anchorage detail the design. Hooks, bends, welded cross bars or mechanical anchors, rather than a straight bar relying on bond. A bar that stops short of the support has removed the tie at the one place it carries its peak force, and the member has no alternative path.

That is the same anchorage discipline described in [reinforcement detailing and congestion](/blog/reinforcement-detailing-and-congestion), applied where the length available is shortest.

Corbels

A corbel is a short cantilever projecting from a column or a wall to support a beam or a precast member.

It behaves as a strut and tie with the strut running diagonally down into the column and the tie running horizontally across the top, and the tie has to be anchored into the column rather than stopped at the back face.

Three things decide whether it works.

The top steel is continuous into the column, developed properly. The single most common corbel failure is tension steel that was lapped or stopped within the corbel itself.

Horizontal ligatures in the corbel body, which hold the strut together.

The bearing detail, because a precast member sitting on a corbel applies its load near the front face, and any horizontal restraint force from shrinkage or thermal movement adds a tension the corbel was not designed for. Which is why the bearing pad and the tolerance matter as much as the steel, and it connects to [precast connections and site tolerances](/blog/precast-connections-and-site-tolerances).

Openings make it much worse

A hole in a deep beam interrupts the strut, and the strut cannot simply divert around it without reinforcement designed to redirect the force.

Which means a penetration through a deep beam or a pile cap is a structural alteration of the load path rather than a local reduction in area, and it is a different problem from a hole in an ordinary slab. The general case is in [core drilling and penetrations in existing slabs](/blog/core-drilling-and-penetrations-in-existing-slabs), and in a deep member the answer is more often no.

On site

Deep beams and pile caps are heavily reinforced in a small volume, which makes them the hardest pours on a project.

Congestion, bar clashes at the nodes, limited space for a vibrator, and a pour depth that generates real heat. Each of those is a quality risk at the exact location where the design depends on the concrete performing, which is why the inspection regime in [construction supervision and pre-pour inspections](/blog/construction-supervision-pre-pour-inspections) matters more here than almost anywhere else.

And the heat is a genuine issue on a large pile cap, which is the subject of [what is the concrete temperature limit](/blog/what-is-concrete-temperature-limit).

ACSES provides structural engineering for residential and commercial projects across Sydney. Talk to us about a project.