Chemical Anchor Pull-Out Testing: Standards, Procedures, and Acceptance Criteria

Chemical anchor pull-out testing is the principal method for verifying the tensile capacity and bond performance of adhesive anchoring systems installed in concrete and masonry. By loading bonded threaded rods or reinforcing bars to a specified proof load or to ultimate failure, engineers can confirm both installation quality and design assumptions in accordance with international standards such as ASTM E488, ACI 355.2, EN 1992-4, and JGJ 145. This article explains specimen preparation, test items, testing procedures, evaluation criteria, typical failure modes, and practical answers to frequently asked questions.

Technical Overview

A chemical anchor (also called a bonded anchor, adhesive anchor, or injected anchor) transfers load into the base material through a two-component thermosetting resin—typically epoxy acrylate (vinylester), pure epoxy, polyester, or hybrid systems—rather than through mechanical expansion. The resin is injected into a drilled hole, a threaded rod or deformed reinforcing bar is inserted, and after curing the load is carried by adhesion and micro-interlock between the resin, the steel element, and the walls of the drilled hole.

Pull-out testing applies an axial tensile force to the installed anchor to verify that the bond strength, steel capacity, and concrete capacity meet the specified design requirements. Testing is performed at two levels:

• Laboratory qualification (type) testing, which establishes the characteristic bond stress and suitability factors of an anchoring product, typically under the framework of ACI 355.2 in North America, or the European Assessment Document EAD 330499 (formerly ETAG 001) published by the European Organisation for Technical Assessment (EOTA), with design rules given in ACI 318 Chapter 17 and EN 1992-4 (Eurocode 2, Part 4).
• Jobsite verification testing, in which a sample of production anchors installed in the actual structure is proof-loaded to confirm workmanship, hole preparation, and in-place resistance, as required by project specifications and national acceptance codes such as GB 50367, GB 50550, and JGJ 145.

Because chemical anchors depend strongly on installation quality—hole cleaning, mixing ratio, curing time, and temperature—pull-out testing is widely regarded as the single most important quality-control measure for adhesive anchoring works, and it is routinely executed by accredited third-party laboratories.

Testing Samples

Laboratory Concrete Specimens

Qualification tests use dedicated concrete members (slabs or blocks) with documented compressive strength, commonly in the range of C20/25 to C50/60. Both uncracked and cracked members are used, with induced cracks of defined width (for example, 0.3 mm to 0.5 mm) running through the anchor position. Specimen dimensions must be large enough to avoid uncontrolled splitting during testing.

Steel Anchor Elements

Typical test specimens include zinc-plated or hot-dip galvanized threaded rods from M8 to M30, stainless steel rods, and deformed reinforcing bars from 12 mm to 32 mm in diameter. Embedment depths (h_ef) usually range from 4 to 12 times the anchor diameter, and every specimen must be installed strictly in accordance with the manufacturer's written installation instructions.

Field (In-Place) Installations

For site acceptance testing, anchors are selected from the completed installation population and grouped into inspection lots by anchor diameter, embedment depth, base material, and installer. Sampling plans defined by the project specification or by GB 50550 / JGJ 145 commonly require testing roughly 0.1% to 0.5% of anchors per lot, with a minimum of three to five anchors for each diameter and embedment combination. Tested anchors must have reached full cure before loading.

Documented Installation Conditions

For each sample, the following parameters should be recorded: base material temperature during installation, hole drilling method, hole cleaning procedure, hole moisture condition, resin batch number, installation date, and elapsed curing time. These records are essential for interpreting test results and for traceability.

Testing Items

Depending on the purpose of the investigation, a pull-out test program may include the following items:

• Ultimate tensile (pull-out) resistance, N_u, in kN.
• Bond stress at failure, calculated as tau = N_u / (pi x d x h_ef), where d is the anchor diameter and h_ef is the effective embedment depth.
• Load–displacement behavior, including anchor stiffness, displacement at the design service load, and displacement at the proof load.
• Failure mode classification: steel rupture, bond failure (pull-out), concrete cone breakout, combined cone-and-bond failure, splitting, or pull-through.
• Residual (permanent) displacement after unloading, as an indicator of elastic recovery in proof tests.
• Sustained-load (creep) displacement for long-term qualification, where anchors carry a constant tension for an extended duration, often several months, at elevated temperature.
• Reduced-capacity conditions, such as cracked concrete, water-filled (wet) holes, overhead installation direction, elevated installation temperature, and short curing times.
• Seismic suitability testing, involving cyclic crack movement combined with tensile loading, as covered by ACI 355.2 and the seismic categories of EAD 330499.

Testing Methods

Laboratory Reference Tests

Reference pull-out tests establish the basic bond capacity. Confined tension tests, in which the loading equipment bears directly against the concrete surface around the anchor, suppress concrete cone failure and isolate the bond interface, producing pull-out failures. Unconfined tension tests, with the reaction frame positioned well away from the anchor, allow the natural governing failure mode to develop. Loading is applied continuously at a controlled rate such that failure occurs within approximately one to three minutes, or under displacement control at rates of roughly 0.5 mm/min to 1.5 mm/min.

Field Proof (Non-Destructive) Testing

In proof testing, the anchor is loaded in increments up to a specified test load—commonly 1.5 times the design tensile service load, or a value defined by the project specification—and the load is held constant for one to two minutes. Acceptance generally requires that the anchor sustain the proof load without failure, that displacement during the hold period remains within a specified limit (frequently 0.1 mm of additional movement), and that no relaxation, slipping, or visible cracking occurs. After a successful test the load is released and the anchor can normally remain in service.

Field Destructive Testing

In destructive testing, the anchor is loaded to ultimate failure to obtain statistical evidence of in-place resistance. This approach is used for critical applications or when non-destructive results are inconclusive. Because the anchor is damaged, tested anchors must be replaced or excluded from the load-bearing layout, and adjacent positions must be checked for concrete damage.

Equipment and Setup

A typical test setup consists of a calibrated hollow-piston hydraulic jack, a load cell or calibrated pressure gauge, and a displacement transducer (LVDT) or dial gauge mounted on an independent reference frame. Key setup requirements include:

• Coaxial alignment of the loading axis with the anchor to avoid bending effects, using a threaded coupler or gripper.
• Reaction supports placed at least two times the embedment depth (and not less than four anchor diameters) from the anchor center for unconfined conditions; closer support converts the test into a confined configuration and must be reported.
• Selection of test anchors located away from member edges whenever possible, or explicit accounting for edge effects.
• Continuous recording of load and displacement, plus photographic documentation of the failure mode.

Data Evaluation and Acceptance Criteria

For laboratory programs, individual ultimate loads are converted to bond stresses, and statistical reduction to the 5% fractile at 90% confidence yields the characteristic bond stress used in design per ACI 318 and EN 1992-4. For field acceptance, every tested anchor must reach the specified proof or failure criterion; if any anchor fails, the usual protocol is to suspend acceptance, investigate the cause, test an increased number of additional anchors in the same lot, and replace all failed anchors. Test reports issued by accredited third-party laboratories should record specimen data, installation records, calibration certificates of the loading system, load–displacement curves, failure modes, and a clear statement of conformity.

Applications

Chemical anchor pull-out testing supports a wide range of construction and engineering activities:

• Structural strengthening and retrofitting, including post-installed rebar for section enlargements, beam-column joints, and new openings, verified under GB 50367 and GB 50550 acceptance procedures.
• Steel-to-concrete connections such as column base plates, steel beam seats, machinery plinths, and equipment subject to vibration or dynamic loading.
• Curtain wall and facade systems, where bracket anchors must resist sustained tension and wind-induced cyclic loads.
• Infrastructure works, including bridge parapets, noise barriers, guardrails, railway fastening systems, and safety-critical attachments.
• Building services installations such as pipe supports, cable trays, handrails, and lift guide-rail anchorages.
• Research and product development, where systematic pull-out campaigns support European Technical Assessments, code-listing programs, and design manual data.

FAQ

What standards govern chemical anchor pull-out testing?

Product qualification is typically performed to ACI 355.2 or EAD 330499, test methods follow ASTM E488 (and equivalent EN test methods), design verification follows ACI 318 and EN 1992-4, and site acceptance in China is governed by GB 50367, GB 50550, and JGJ 145. Testing should be carried out by accredited third-party laboratories with calibrated equipment.

What is the difference between non-destructive and destructive pull-out tests?

A non-destructive proof test loads the anchor to a predefined inspection load and holds it for a set time; if the criteria are met, the anchor remains in service. A destructive test loads the anchor to failure to measure true ultimate capacity, and the anchor must afterwards be replaced or excluded from the structural layout.

How many anchors must be tested on a construction site?

The number depends on the project specification and local code. Common practice requires approximately 0.1% to 0.5% of the anchors in each inspection lot, with a minimum of three to five anchors per diameter and embedment depth combination, plus additional retesting whenever a failure occurs.

How long must anchors cure before testing?

Curing time depends on the resin chemistry and the base material temperature, and the manufacturer's technical data sheet governs. As an indication, vinylester systems may reach handling strength in about 45 to 90 minutes at 20 °C, whereas pure epoxies and low-temperature installations may require several hours to several days. Pull-out testing should only begin after full cure.

What do the different failure modes indicate?

Steel rupture indicates that the bond and concrete are stronger than the anchor rod, which is usually the desired ductile mechanism. Bond failure (the rod pulling out with the resin) often points to inadequate hole cleaning, insufficient curing, or incorrect installation. Concrete cone breakout indicates that the base material governs and depends on embedment depth and concrete strength.

What happens if an anchor fails a proof test?

Acceptance for that lot is suspended, the installation procedure and materials are investigated, double the original number of anchors in the same lot is normally retested, and all failed anchors are replaced. If repeated failures occur, the anchoring system or installation method may require full requalification.

Does passing a short-duration pull-out test guarantee long-term performance?

No. Short-term pull-out capacity does not by itself demonstrate resistance to creep under sustained load. Long-term performance is covered by product qualification, which includes sustained-load (creep) testing and temperature sensitivity assessment in accordance with ACI 355.2 and EAD 330499, so only fully qualified systems should be used under permanent tension.