Post-installed anchor pull-out testing is the principal method for verifying the tensile capacity, bond performance, and installation quality of anchors and bonded reinforcing bars installed in hardened concrete. This technical guide explains the governing standards, sample categories, measured parameters, test procedures, acceptance criteria, and typical engineering applications of post-installed anchor pull-out testing. It is intended for structural engineers, quality inspectors, and construction professionals responsible for anchorage design verification and on-site acceptance.
Technical Overview
Post-installed anchoring refers to fastening systems installed into hardened concrete (or masonry) by drilling, as opposed to cast-in anchors that are placed before concrete pouring. The pull-out test is the most direct and widely accepted method for evaluating whether a post-installed anchorage can safely transfer tensile loads into the base material. During the test, an axial tension force is applied to the anchor or bonded rebar through a calibrated hydraulic loading system until either a specified proof load is sustained without distress or the anchorage reaches failure.
Post-installed anchor pull-out testing serves two distinct purposes. The first is product qualification and certification, in which systematic laboratory programs establish characteristic resistance values for design; these programs are governed by standards such as ASTM E488, ACI 355.2 for mechanical anchors, ACI 355.4 for adhesive anchors, and the European assessment documents derived from ETAG 001, with design rules harmonized in EN 1992-4. The second purpose is construction quality acceptance, in which on-site tests verify that anchors installed under actual project conditions meet the specified design capacity; in China this is controlled by JGJ 145 and GB 50550, and such tests are commonly performed by accredited third-party laboratories.
A central objective of pull-out testing is to identify the governing failure mode, because design resistance is calculated differently for each mechanism. The principal failure modes include:
- Steel failure of the anchor rod or reinforcing bar, generally the preferred ductile mode for well-designed anchorages.
- Bond failure (pull-out), in which the anchor withdraws from the adhesive layer or the expansion mechanism loses grip.
- Concrete cone breakout, in which a cone of base material separates from the member.
- Splitting failure of the concrete member, typically caused by insufficient edge distance or spacing.
- Pull-through failure, in which the anchor sleeve pulls past the expansion element.
Testing Samples
Mechanical Expansion Anchors
Torque-controlled expansion anchors expand against the drill hole wall as a nut is tightened, while displacement-controlled anchors expand when a conical mandrel is driven into the sleeve. Pull-out testing of these products verifies expansion force development, sensitivity to hole diameter tolerance, and torque-limited installation behavior in the specified concrete strength class.
Undercut Anchors
Undercut anchors form a mechanical interlock by cutting a geometric undercut into the base material. Testing focuses on the consistency of the undercut geometry, load-displacement behavior, and the achieved concrete breakout capacity, which should approach the theoretical value for the undercut diameter.
Bonded (Adhesive) Anchors
Bonded anchors, also called chemical anchors, use injection mortars based on epoxy, vinylester, polyester, or hybrid resins, or two-component capsule systems. Pull-out testing evaluates bond strength along the embedded length, sensitivity to hole cleaning, curing time, installation temperature, and sustained-load creep behavior.
Post-Installed Reinforcing Bars
Post-installed rebar connections use deformed reinforcing bars bonded into drilled holes with structural injection mortar, typically for structural extensions, slab-to-wall joints, and strengthening works. Site pull-out testing of bonded rebar is a mandatory acceptance item in many jurisdictions, confirming that the full anchorage length develops the required bar stress without pull-out or splitting of the cover concrete.
Testing Items
Depending on the test objective, the following quantities and behaviors are measured during post-installed anchor pull-out testing:
- Ultimate tensile (pull-out) capacity of the anchorage under monotonic loading to failure.
- Displacement (slip) at service load, at proof load, and at ultimate load, recorded as a load-displacement curve.
- Residual displacement after unloading, indicating permanent slip of the anchorage.
- Failure mode classification, compared against the mode assumed in design.
- Characteristic resistance and related reduction factors derived from statistical evaluation of replicate tests.
- Suitability in cracked concrete, including cyclic crack movement tests for qualification programs.
- Sustained-load creep displacement for adhesive anchors, which governs long-term bond performance.
- Installation sensitivity parameters such as hole cleaning condition, mortar temperature, and installation direction (horizontal, overhead, or downward).
Testing Methods
Laboratory Qualification Testing
Qualification programs per ASTM E488, ACI 355.2, ACI 355.4, and the European assessment documents test anchors in standard concrete members at defined strength classes, in both uncracked and cracked concrete. Reference tension tests may be performed unconfined, with the reaction frame bearing far from the anchor so that concrete cone breakout can develop freely, or confined, with a closely spaced bearing ring that suppresses cone failure to expose bond or pull-out behavior. Results from a statistically sufficient number of specimens establish the characteristic capacity used in design to EN 1992-4 or the corresponding national code.
On-Site Destructive Testing
In destructive site testing, representative anchors installed by the project team are loaded in increments until failure. The measured ultimate load is compared against the required design resistance, and results from multiple specimens are evaluated statistically. Tested anchors are subsequently removed or cut back and the holes repaired, so destructive testing is normally applied to dedicated sacrificial locations or where repair is acceptable.
On-Site Non-Destructive (Proof Load) Testing
Non-destructive acceptance testing loads the anchor in stages up to a specified proof load, typically related to the design load by a defined safety factor, holds that load for a specified duration to reveal creep or slip, and then unloads. Acceptance criteria generally require that the anchorage sustain the proof load without rupture, pull-out, excessive displacement, or cracking of the base material. This method preserves the usability of the tested anchors and is therefore the default approach for site acceptance of post-installed rebar and structural anchors.
Equipment and Procedure
A typical test setup consists of a calibrated hydraulic hollow cylinder and pump, a reaction frame or bearing ring positioned at an adequate clear distance from the anchor to avoid artificial confinement, a load cell or calibrated pressure measurement system, and dial gauges or displacement transducers fixed to an independent reference frame to record anchor head movement. Loading is applied at a controlled rate, generally as a stepped sequence with hold periods, and load, displacement, time, and observed distress are recorded continuously throughout the test.
Applications
Post-installed anchor pull-out testing is applied across a wide range of construction and retrofit scenarios:
- Structural strengthening and renovation of existing concrete buildings, including steel plate or FRP attachment and section enlargement using post-installed rebar.
- Structural extensions such as new floors, slab continuations, staircases, and beam-to-column connections formed with bonded rebar.
- Steel superstructure connections, canopy supports, and machinery bases anchored to hardened concrete foundations.
- Curtain wall, façade, and cladding systems where anchor capacity directly affects envelope safety.
- Mechanical, electrical, and plumbing equipment anchorage, including supports subject to vibration.
- Seismic retrofit works, where anchor ductility and cyclic performance are critical.
- Post-construction acceptance and dispute resolution when as-built anchorage quality must be objectively verified by accredited third-party laboratories.
FAQ
What is the difference between destructive and non-destructive pull-out tests?
A destructive test loads the anchorage to failure to measure its true ultimate capacity, and the tested anchor cannot remain in service. A non-destructive proof test loads the anchor only to a specified verification load and accepts it if that load is sustained within displacement limits, allowing the anchor to remain functional.
How many anchors should be tested on a project?
Most acceptance standards require testing of a minimum percentage or count of installed anchors per batch, geometry, base material, and installation crew, with expanded sampling when a nonconforming result occurs. The project specification and the governing code, such as JGJ 145 or GB 50550, define the exact sampling rules.
What load is applied during a non-destructive proof test?
The proof load is defined as the design tensile resistance multiplied by a specified verification factor set by the applicable standard, so that the anchor is challenged beyond service load but remains within its elastic and non-damaging range. The load is held for a defined period to expose creep, slip, or marginal installation quality.
What do the different failure modes indicate?
Steel failure usually confirms that the anchor or bar governs and the anchorage is efficient. Pull-out or bond failure indicates insufficient embedment, poor hole preparation, or unsuitable adhesive. Concrete cone or splitting failure points to inadequate embedment depth, edge distance, spacing, or low base material strength.
Why do adhesive anchors require creep testing?
Polymer-based mortars can deform progressively under sustained tension loads, particularly at elevated in-service temperatures. Sustained-load tests measure long-term displacement to confirm that creep remains within acceptable limits over the design service life.
What happens if an anchor fails an acceptance test?
A failed test triggers expanded sampling of the same batch as required by the standard, a review of installation records, materials, and drilling procedures, and engineering assessment of the affected connection. Depending on the findings, remedial measures such as deeper embedment, additional anchors, or complete reinstallation may be required.