Post-Installed Rebar Anchoring

Because the bar has to be surrounded by adhesive, and the adhesive needs an annulus around it to develop the bond. Each system publishes the hole diameter for each bar diameter, and that pairing belongs to the product approval rather than to a rule of thumb. Drilling a hole the size of the bar, or one that a convenient bit happens to give, takes the fixing outside its data and leaves the design without a basis.

To the anchorage length the design requires, taken from the system data for that bar diameter, concrete condition and loading. It is not automatically the same as the lap length used for cast-in bars, because a cast-in lap assumes fresh concrete flowing around the bar and curing against it. The depth is fixed before drilling, checked during the work, and not shortened because the drill met something.

It has to be deformed bar in the grade that the design and the system data cover, clean and free of oil, loose scale and flaking rust that would sit between the bar and the adhesive. Some systems state the bar standards and grades they have been assessed with. A smooth bar or an unidentified offcut has no bond worth designing for, however good the adhesive.

It can be, but not by assumption. A post-installed connection develops its anchorage by bond in concrete that has already cured, so its capacity follows the system data, the concrete condition, the cover and the embedment, and it is designed as a post-installed connection rather than as a cast-in one. Where the design intent is to reproduce a cast-in lap, that has to be established from the system's assessment for that case.

Because the bond is developed in the concrete around the hole, and the concrete between the hole and the surface has to resist being pulled away with the bar. Low cover, or cover reduced by a repair or a previous breakout, weakens that resistance and can govern the capacity instead of the bar or the adhesive. Edge distance at a slab or wall face is the same question in another direction.

Stop and re-site the hole. Cutting reinforcement in a structural member is not a decision to be taken on site, and drilling immediately alongside a bar leaves the new hole with less concrete to bond against. The member is scanned before drilling, and a clash is resolved by moving the position and rechecking the anchorage length available — not by cutting the obstruction out of the way.

Yes. The adhesive cures by chemical reaction, which slows as the concrete gets colder, so both the gel time and the full-cure time come from the cartridge data for the temperature measured at the hole. Bars tied, bent or loaded before full cure can lose their anchorage permanently, and nothing about the finished fixing shows it. Cold slabs in winter and shaded plant rooms are where this is most often missed.

Bending has to be allowed for before the bar goes in, because the anchorage length assumes the embedded portion stays straight, and bending a set bar can damage the bond. Welding a projecting bar puts heat into the adhesive, and not every system permits it. If either is likely on the job, it should be settled at design stage rather than discovered on site.

By the record and, where specified, by testing. The record covers the bar diameter, the hole diameter, the depth and embedment actually achieved, the product and batch, the cleaning method, the temperature and the cure time. Pull-out testing on a sample is asked for on many structural connections, and the results are recorded against the position so that a third party can check them later.

When the connection has to behave like reinforced concrete — a starter bar that will be bent into a new slab, a lap into a new wall, or a dowel that continues reinforcement across a joint. A threaded rod with a chemical anchor is a fixing; a post-installed bar is reinforcement, and the design, the embedment and the acceptance testing all follow from that difference. Which applies is settled by the structural designer.