Post-expansion undercut anchors are among the most reliable post-installed fastening systems for concrete, and tensile testing is the core procedure used to verify their pull-out resistance, stiffness, and failure behavior. This article explains how undercut anchor tensile tests are performed, which international standards apply, what samples and equipment are required, and how the results are evaluated for structural design and product qualification.
Technical Overview
A post-expansion undercut anchor is a mechanical fastener installed into a pre-drilled hole in hardened concrete. A dedicated undercut drill first creates a cylindrical borehole and then enlarges (undercuts) the base of the hole. After the anchor is inserted, the expansion sleeve is deformed into this undercut by applying torque or percussion, so that tension loads are transferred through direct mechanical interlock (keying) with the concrete rather than through friction or expansion pressure alone.
Because load transfer relies on geometric locking, undercut anchors achieve high and predictable tensile capacity, perform well in cracked concrete, and exhibit relatively small slip under service loads. Tensile (pull-out) testing is therefore essential to quantify the ultimate resistance, to classify failure modes, to confirm design values, and to support product qualification and factory production control.
Testing is typically carried out by accredited third-party laboratories in accordance with the following reference documents:
- EN 1992-4:2018 – Design of fastenings for use in concrete (design verification of test results)
- EAD 330232-01-0601 – European Assessment Document for undercut anchors (formerly covered by ETAG 001)
- ASTM E488/E488M – Standard test methods for strength of anchors in concrete elements
- ACI 355.2 – Qualification of post-installed mechanical anchors in concrete, referenced by ACI 318 Chapter 17
- JGJ 145 – Chinese technical specification for post-installed fastenings in concrete structures
- ISO/IEC 17025 – General requirements for the competence of testing and calibration laboratories
Testing Samples
Anchor Specimens
Test specimens are production-run anchors taken from a representative manufacturing batch, complete with the supplied expansion sleeves, threaded rods or internally threaded bodies, washers, and nuts. Common sizes range from 8 mm to 30 mm in diameter (M8 to M30), with effective embedment depths typically between 60 mm and 200 mm. Material grades such as carbon steel grade 5.8 or 8.8 (with galvanized or zinc-flake coating) and stainless steel grades A2 and A4 are documented through mill certificates, and dimensional inspection precedes testing.
Concrete Test Members
Anchors are installed in unreinforced or lightly reinforced concrete slabs or walls with a defined compressive strength. Qualification programs normally use at least two strength levels, for example C20/25 to C25/30 for low-strength concrete and C50/60 for high-strength concrete. Members are at least 28 days old, thick enough to prevent splitting (typically no less than 1.5 times the embedment depth, or as specified by the standard), and the actual strength is verified on cubes or cylinders in parallel, following EN 12390-3 or ASTM C39.
Cracked Concrete Specimens
For qualification in cracked concrete, specimens are pre-cracked so that a controlled crack of defined width passes through the anchor axis. Crack widths of 0.3 mm are commonly used for serviceability conditions, while widths up to 0.5 mm are applied in simulated seismic qualification. Cracks are induced by hydraulic jacks or post-tensioned rods and are monitored with crack gauges or clip-on displacement sensors during the test.
Testing Items
The tensile test program determines the complete axial load-bearing behavior of the undercut anchor system. Typical testing items include:
- Ultimate tensile (pull-out) resistance, expressed as the mean value N_ru,m of the test series
- Characteristic tensile resistance N_Rk, calculated as a 5% fractile with a defined confidence level
- Load–displacement (load–slip) curve, including anchor stiffness and displacement at service load levels
- Failure mode classification: steel fracture, concrete cone breakout, pull-out, splitting, or side-face blow-out
- Verification of installation parameters, including drilling tolerance, hole cleaning, and expansion (installation) torque
- Influence of concrete strength, crack width, and embedment depth on tensile capacity
- Complementary investigations such as sustained (creep) tension tests, crack-movement cycling, and seismic tension tests
Testing Methods
Specimen Preparation and Installation
The hole is drilled with a rotary hammer drill fitted with the manufacturer's undercut bit to the nominal diameter and depth. After drilling, the hole is cleaned according to the manufacturer's printed installation instructions, typically using blow–pump–brush cycles. The anchor is then inserted and the sleeve is expanded into the undercut using a calibrated torque wrench or the prescribed percussion setting, and the applied installation torque is recorded.
Test Setup
The standard configuration is an unconfined tension test: a rigid reaction frame transfers the test load to the concrete surface through supports placed at a clear distance of at least twice the effective embedment depth (2 h_ef) from the anchor, so that an unrestricted concrete cone can form. Confined setups, where the reaction bears directly on the concrete around the anchor, are only used for specific reference or comparison tests because they suppress cone breakout. Load is applied through a calibrated load cell (accuracy Class 1 per ISO 7500-1) and a servo-hydraulic actuator or loading frame. Anchor displacement is measured relative to the undisturbed concrete surface using two or three displacement transducers (LVDTs) with a resolution of 0.01 mm, mounted on an independent reference frame and averaged.
Loading Procedure
For static tension tests to failure, load is increased either in steps of approximately 10% to 20% of the anticipated ultimate load, each held constant for one to three minutes, or continuously at a controlled rate such that failure occurs within the time window prescribed by the standard (typically two to five minutes). Load, displacement, and crack development are recorded continuously. At failure, the maximum load, the failure mode, and the shape and dimensions of any concrete cone are documented photographically and in the test report.
Data Evaluation and Acceptance
For each test series of at least five valid specimens, the mean ultimate load and coefficient of variation are computed, and the characteristic resistance is derived as the 5% fractile with 90% confidence using statistical coefficients that depend on the number of tests. Results are normalized to the reference concrete strength where required, and are compared with the design resistance models of EN 1992-4 or ACI 318, for example the concrete cone breakout value N_Rk,c = k1 · sqrt(f_ck) · h_ef^1.5 and the steel strength N_Rk,s = A_s · f_uk. Based on reliability and sensitivity test results, qualification standards such as ACI 355.2 assign anchors to categories with corresponding capacity reduction factors (for example 1.00, 0.80, and 0.65), which are then applied in design.
Applications
Verified tensile performance makes post-expansion undercut anchors suitable for demanding structural and non-structural fixings in new construction and renovation, particularly where cracked concrete, dynamic loads, or seismic action must be considered. Typical applications include:
- Structural steel connections, beam seats, and column base plates anchored to concrete
- Curtain wall and facade bracketry subjected to wind tension loads
- Anchoring of machinery, pumps, compressors, and steel base plates subject to vibration
- Pipe supports, cable trays, and suspended systems in industrial plants
- Post-installed connections in strengthening and retrofit projects, including guardrails and noise barriers
- Nuclear, tunnel, and bridge applications requiring demonstrated reliability under sustained and seismic tension
In all cases, the published characteristic tensile resistance is only valid when the anchor is installed with the qualified drilling system, hole cleaning procedure, and installation torque, and when the design checks for concrete cone breakout, pull-out, splitting, and steel strength according to EN 1992-4 or ACI 318 are satisfied.
FAQ
What is the difference between an undercut anchor and an ordinary expansion anchor?
A torque-controlled expansion anchor transfers tension mainly through friction generated by expansion pressure against the hole wall, and its capacity can degrade if the expansion relaxes. A post-expansion undercut anchor locks mechanically into a machined undercut at the base of the hole, so load transfer is by direct keying. This gives higher and more stable tensile capacity, especially in cracked concrete, but requires a special undercut drill and stricter installation control.
Which failure modes can occur in a tensile test?
The principal modes are steel fracture of the rod or body, concrete cone breakout, pull-out of the anchor without a full cone, splitting of the concrete member, and side-face blow-out near edges. Steel failure is the most ductile mode; a correctly functioning undercut anchor in well-proportioned concrete typically fails either by steel fracture or by a well-developed concrete cone close to the theoretical value.
Why must tests be performed in both uncracked and cracked concrete?
Cracks in concrete reduce the confinement available at the anchor base and may intersect the undercut. Testing with a defined crack width (commonly 0.3 mm, or 0.5 mm for seismic qualification) demonstrates that the anchorage retains sufficient tensile capacity and acceptable displacement even in the cracked state expected in real structures.
What is the difference between confined and unconfined tension tests?
In an unconfined test, the reaction frame bears on the concrete at least 2 h_ef away from the anchor, allowing a free concrete cone to develop; this is the reference condition for design. In a confined test, the reaction bears directly around the anchor, suppresses cone breakout, and produces higher pull-out values; confined results are used only for specific comparison purposes and not for design resistance.
How many specimens are required for a valid test series?
Standards such as ASTM E488 and the qualification protocols based on ACI 355.2 and EAD 330232 require a minimum of five valid results per test series and condition, while five to ten specimens are customary. Invalid results, for example due to incorrect installation or interference from reinforcement, must be replaced.
What tensile capacity can be expected?
Capacity depends on anchor size, embedment depth, concrete strength, and crack condition. As an illustration, a 12 mm undercut anchor embedded approximately 100 to 125 mm in normal-weight concrete typically achieves characteristic tensile resistances in the range of roughly 20 kN to 60 kN, governed either by steel strength or by concrete cone breakout. Design values are obtained by applying the partial factors of EN 1992-4 or the strength-reduction factors of ACI 318.
Who should perform the testing?
Qualification and certification tests should be conducted by accredited third-party laboratories operating in accordance with ISO/IEC 17025, using calibrated load and displacement instrumentation and concrete members whose strength is verified in parallel. Routine factory production control tests may be performed internally, but periodic independent verification is recommended.