Anchor Design vs Proof Testing

Frequently asked questions

Why does hole cleaning matter so much?

Because an adhesive anchor's capacity comes from the bond between the adhesive and the concrete, and dust prevents that bond forming. Drilling produces a fine powder that coats the walls of the hole; injecting adhesive into an uncleaned hole means the adhesive bonds to a layer of loose dust that is itself attached to nothing, and the capacity loss is large — large enough that manufacturers' published values are explicitly conditional on the cleaning procedure being followed. That procedure is specific: a defined sequence of blowing out the hole with compressed air and brushing it with a brush of the correct diameter, repeated a stated number of times, and it differs between products and between dry and wet or flooded holes. It is the most commonly skipped step in the whole operation, and it is invisible once the anchor is in.

Why does a concrete failure mode usually govern?

Because concrete is much weaker in tension than steel, and pulling an anchor out of it loads the concrete in tension. The anchor transfers its load into a volume of concrete around and below it, and that volume fails by a cone breaking out — the capacity depending on the embedment depth, the concrete's strength, and critically on whether the cone is complete. Nearby edges truncate the cone and nearby anchors' cones overlap, so an anchor near an edge or in a tight group has a fraction of the capacity it would have in open concrete. That is why spacing and edge distance appear so prominently in anchor design, why moving an anchor a little further from an edge can be worth more than a bigger bolt, and why the steel's grade is frequently irrelevant.

What does cracked concrete change?

The anchor's capacity and, in many cases, whether the anchor is suitable at all. A crack passing through the anchor's location disrupts the concrete it relies on — for a mechanical anchor it can open the hole so the expansion mechanism loses its grip, and for an adhesive anchor it breaks the bonded zone. Anchors are therefore qualified separately for cracked and uncracked concrete, with different and substantially lower capacities in the cracked case, and some anchors are not qualified for cracked concrete at all. The design decision is whether the concrete should be ASSUMED cracked, and in most structural applications the answer is yes, because a member in flexure is cracked in its tension zone by design. Using an uncracked value in a location that will crack is a common and consequential error.

What is a proof load and why not test to failure?

A proof load is a defined load below the anchor's design capacity, applied and held for a period, with acceptance based on the anchor sustaining it without excessive displacement — it demonstrates adequacy rather than establishing ultimate capacity. Testing a working anchor to failure would destroy it, which is the obvious objection, and it would also damage the concrete around it. So proof testing is non-destructive by design: the anchor is loaded to a value the specification sets, its displacement is measured, and it passes if it holds without slipping. Where ultimate capacity genuinely needs establishing — on an existing structure of unknown concrete, or to qualify a method — sacrificial anchors are installed specifically to be tested to failure, separate from the working ones.

How many anchors should be tested?

A proportion set in the specification before the work starts, along with the test load, the acceptance criterion and the response to a failure — and settling all four up front is what makes the regime useful. The proportion varies with the consequence: a higher rate for safety-critical fixings, for overhead adhesive anchors in sustained tension, and where the concrete's condition is uncertain, and a lower sampling rate for large numbers of routine fixings. What matters as much as the number is what happens when one fails: a defined escalation — test more, investigate the installation, remediate the affected area — rather than an argument on site about whether the test was fair. The tests should also sample across installers and across days, since installation quality varies by both.

Are there anchors that avoid the site variability?

Cast-in anchors do, which is why they are preferred where the layout can be known in advance. A cast-in bolt or channel is placed before the pour and is embedded in concrete cast around it, so there is no drilling, no hole to clean and no adhesive to mix — the capacity depends on the design and on the anchor being held in position during the pour, which is a far more controllable failure mode. The cost is that the layout has to be known before the concrete goes in, and moving one afterwards is not an option, which is exactly why post-installed anchors are so widely used. Where a layout is known and fixed, the comparison favours cast-in comfortably; where it is not, post-installed anchors with a proper installation regime are the answer.

What else goes wrong at installation?

Several things, all of them invisible afterwards. The hole drilled to the wrong DIAMETER, which for an adhesive anchor changes the bond area and the adhesive volume and for a mechanical one changes the expansion. The hole drilled to the wrong DEPTH, which reduces embedment and therefore the concrete cone directly. Drilling into reinforcement and either damaging it or stopping short. Adhesive not mixed properly, which for a cartridge system means discarding the first portion until the mix is uniform. Adhesive injected from the top rather than from the bottom of the hole, trapping air. Installation outside the temperature range, or loading before the cure is complete. Each is a departure from a stated procedure, and each is why the procedure is printed on the cartridge.

Does this apply to a domestic fixing?

The principles do, in proportion, and the two that transfer are edge distance and hole cleaning. A heavy fixing into concrete or masonry close to an edge has far less capacity than the same fixing in the middle of a wall, which is why a bracket anchored just below a slab edge or beside a window opening pulls out at loads the fixing's packet suggested it would hold. And blowing the dust out of a drilled hole before inserting a resin anchor takes seconds and makes a substantial difference. Neither requires a calculation or a test on a domestic job; both are the reason a fixing that should have held did not. Where the fixing is genuinely safety-critical — a stair balustrade, a heavy wall-mounted item above a seating area — it is worth treating as a design question.