Anchors Into Concrete And Why They Fail

An anchor drilled into concrete is the most common structural connection on a site and the least often designed.

It is also the one where the published capacity and the achieved capacity diverge most, because almost every factor that reduces it is invisible after installation.

The failure is usually the concrete, not the bolt

Four modes, and only one of them involves the anchor breaking.

Concrete cone failure, where a cone of concrete pulls out around the anchor. The governing mode for most anchors in tension, and it depends on embedment depth and edge distance rather than on the anchor's own strength.

Pull-out, where the anchor slips out of the hole without taking concrete with it. A bond or expansion failure, usually caused by a dirty hole or an under-torqued expansion anchor.

Edge breakout, where a wedge of concrete shears off the face. The mode under shear load near an edge, and it governs more fixings than people expect because so many anchors end up near an edge.

Steel failure, where the anchor itself yields or shears. The mode the published bolt grade refers to, and the least likely of the four in practice.

Which means designing to the bolt's steel capacity, without checking the concrete, overstates capacity substantially.

The factors that reduce capacity most

Edge distance. Capacity falls away sharply as an anchor approaches an edge, because the cone cannot develop. An anchor at half the required edge distance can have a small fraction of its catalogue capacity.

Spacing. Two anchors close together share one cone rather than having two, so a group is worth less than the sum of its parts. Anchor group design accounts for this and a site decision to add a second bolt beside the first does not.

Embedment depth. Capacity scales strongly with depth. A hole drilled short because the drill hit reinforcement is a reduced anchor, and nobody can see it afterwards.

Cracked concrete. Anchors behave differently in concrete that is cracked, which in a reinforced member in service is the normal condition rather than an exception. Anchors are qualified for cracked or uncracked concrete and the two ratings are not close.

What it is fixed into. A solid slab, a hollow core slab, a masonry wall and a hollow block are four different substrates with four different anchor types. An anchor suited to solid concrete put into a hollow block has almost no capacity, and this is a frequent facade fixing failure, covered in [facade fixings and substrate capacity](/blog/facade-fixings-and-substrate-capacity).

The installation items that decide the result

Every one of these is invisible once the fixture is on.

Hole cleaning. The largest single installation variable for a chemical anchor. Drilling dust left in the hole can remove most of the bond capacity. The procedure is blow, brush, blow, to the manufacturer's specification, and it is the step most often skipped.

The right hole diameter, matched to the anchor rather than to the nearest available bit.

Torque, for a mechanical anchor, applied with a calibrated wrench. Under-torqued means it has not expanded; over-torqued means the cone has been pulled through and the concrete locally crushed.

Cure time for a chemical anchor, which depends on temperature and is longer in cold weather than most people wait.

Hole orientation. An overhead chemical anchor has different installation requirements from a downward one, and not every product is qualified overhead.

Not cutting reinforcement. Drilling through a bar to get the depth has removed structural steel to install a fixing, and on a slab near a column it can take capacity that mattered, which is the whole argument in [core drilling and penetrations in existing slabs](/blog/core-drilling-and-penetrations-in-existing-slabs).

Where anchors are asked to do too much

Retrofitted holding down bolts on a baseplate, where the footing was poured without provision. Workable with proper design and far weaker than cast-in, as noted in [steel baseplates and holding down bolts](/blog/steel-baseplates-and-holding-down-bolts).

Balustrade fixings, which take a horizontal load at the top of a post, which is a large moment at the base, which is tension on anchors near a slab edge. This combination is the single worst anchor condition in ordinary construction and it is why [balustrades and handrails](/blog/balustrades-and-handrails-structural-requirements) are an engineered item.

Rooftop plant and solar, where uplift puts anchors into tension on a roof whose structure was never checked for it, as in [rooftop solar and plant structural adequacy](/blog/rooftop-solar-and-plant-structural-adequacy).

Temporary works anchors, including brace feet for precast and scaffolding ties, where the anchor goes into young concrete at its weakest. Covered in [temporary bracing of precast panels](/blog/temporary-bracing-of-precast-panels) and [scaffolding ties and loads](/blog/scaffolding-ties-and-loads-on-existing-structures).

Proving it rather than assuming it

Anchors are testable, and on anything consequential they are tested.

A proof load applied to a sample of installed anchors confirms the substrate, the installation and the capacity together, which no amount of catalogue reading does. On a facade, a balustrade or a retrofitted structural connection, that test is the difference between a designed fixing and a hopeful one.

Where an existing anchor group has to be assessed, the test is often the only way, because the embedment and the hole condition cannot be established any other way.

ACSES provides structural engineering and fixing design across Sydney. Talk to us about a project.