Soil anchor pull-out testing is a critical geotechnical verification method used to confirm the bond capacity, load-transfer behavior, and serviceability of ground anchors installed in soil strata for excavation support, slope stabilization, and structural anchoring systems. This article provides a comprehensive technical overview of soil anchor pull-out testing, covering applicable standards, sample types, test items, loading procedures, equipment requirements, acceptance criteria, and practical applications in modern construction engineering.
Technical Overview
A soil anchor (also known as a ground anchor or soil nail in certain applications) is a tension-bearing element consisting of a steel tendon—typically a rebar, strand, or steel pipe—installed into a drilled borehole and bonded to the surrounding soil through cement grout or mortar. The anchor transfers tensile loads from a supported structure, such as a retaining wall or shotcrete facing, into stable soil layers at depth. Because the actual load-carrying capacity of a soil anchor cannot be reliably determined by calculation alone—owing to variability in soil strength, groundwater conditions, grouting quality, and construction workmanship—pull-out testing is mandatory before and during production anchoring work.
Soil anchor pull-out testing, sometimes referred to as a tensile test or load test, involves applying an axial tensile load to the anchor head at graded increments while simultaneously measuring anchor head displacement. The test yields the load–displacement curve, ultimate bond resistance, and creep characteristics of the anchor, which are then compared against design assumptions and code-defined acceptance criteria. Key standards governing these tests include GB 50086 (Technical Code for Engineering of Ground Anchorages and Retaining Structures), JGJ 120 (Technical Specification for Retaining and Protection of Building Foundation Excavations), ASTM D4435 (Standard Test Method for Rock Bolt Anchor Performance by Pulling), and EN 1537 (Execution of Special Geotechnical Works – Ground Anchors). These tests are typically performed on site by accredited third-party laboratories or qualified geotechnical testing organizations.
Testing Samples
Grouted Tension Anchors (Bolt-Type)
The most common test specimen is a cement-grouted anchor comprising a deformed steel bar (typically 25–40 mm diameter, grades such as HRB400 or higher-yield threaded bar) installed in a drilled hole of 100–180 mm diameter. The bonded anchor length is usually limited to 3–6 m for testing purposes to ensure bond failure occurs before the steel yields, allowing the unit bond stress along the soil–grout interface to be evaluated.
Pre-Stressed Strand Anchors
For deep excavations and permanent anchoring systems, multi-strand tendons (e.g., 3–7 strands of 15.2 mm low-relaxation prestressing steel) anchored in larger-diameter boreholes are tested. These samples often incorporate secondary (re-grouting) operations to enhance capacity, and test procedures must account for elastic elongation of the free length.
Self-Drilling Hollow Bar Anchors
Used in collapsible or fill soils where open-hole drilling is impractical, self-drilling hollow-core bars with simultaneous grouting are tested to verify capacity in difficult ground conditions.
Soil Nails
For soil nailing walls supporting excavations and slopes, shorter grouted bars (typically 3–12 m) are extracted or loaded in non-destructive proof tests to verify the design bond stress between grout and soil.
Testing Items
The pull-out test program evaluates the following principal performance parameters:
- Ultimate pull-out capacity (bond strength): the maximum tensile load sustained by the anchor before continuous displacement, bond interface failure, or structural failure of the tendon occurs.
- Load–displacement behavior: the complete curve of anchor head displacement versus applied load, used to assess load-transfer mechanisms and stiffness.
- Elastic and plastic displacement: separation of recoverable (elastic) elongation from permanent (plastic) displacement during cyclic unloading–reloading, indicating whether the anchor is functioning primarily through bond along the fixed length.
- Creep displacement characteristics: time-dependent displacement under sustained constant load, critical in soft clays, silts, and organic soils where creep may govern long-term performance.
- Load loss and relaxation: reduction in load after lock-off, relevant for pre-stressed anchors.
- Verification of design parameters: back-calculation of the characteristic unit skin friction (bond stress) at the soil–grout interface for refinement of the production anchor design.
Testing Methods
Basic (Investigative/Destructive) Test
Conducted before production anchor installation, the basic test loads trial anchors to failure using a cyclic loading schedule. Load is applied in increments of approximately 0.1–0.2 times the estimated ultimate load, with each increment held for a specified observation period during which displacement readings are taken at defined time intervals (e.g., 5, 10, 15 minutes or until displacement rates stabilize). Loading continues until one of the following failure criteria is reached: displacement continues to increase without stabilization under a constant load; total displacement exceeds the combined theoretical elastic elongation plus a specified plastic allowance; or the tendon itself fails. The results establish the ultimate bond capacity and the design value for production anchors, usually with a defined factor of safety.
Creep Test
For anchors in cohesive soils with plasticity indices above approximately 17, or where creep susceptibility is suspected, a dedicated creep test is performed. The anchor is loaded in increments (e.g., 0.4, 0.6, 0.8, 1.0, 1.2 times the design load), with each stage held for an extended observation period (commonly 30–60 minutes or longer). Displacement is logged against logarithmic time, and the creep rate—the displacement increment per log cycle of time—is computed. Acceptance generally requires the creep rate at the design load to remain below the code-specified limit (commonly less than 2.0 mm per log time cycle), confirming long-term displacement stability.
Acceptance (Proof/Verification) Test
Each production anchor, or a specified percentage of production anchors (typically no less than 5% and not fewer than 3 anchors per code requirements), undergoes a non-destructive acceptance test. The anchor is cyclically loaded to a maximum verification load of 1.3–1.5 times the design (characteristic) axial force, then unloaded, reloaded, and locked off at the design lock-off load. Acceptance criteria typically require: the total displacement at the maximum test load does not exceed the theoretical elastic elongation of the tendon free length plus a permissible plastic displacement; displacement stabilizes within the specified hold periods; and the residual creep rate falls within allowable limits. Anchors failing the criteria may be de-rated, re-tested, or supplemented with additional anchors.
Test Apparatus and Instrumentation
Standard test equipment includes:
- A calibrated hydraulic jack and pump acting against a rigid reaction frame or bearing plate centered on the anchor head.
- A calibrated load cell or pressure gauge (with calibration certificates traceable to national standards) to measure applied load, typically with an accuracy of ±1–2% of the measured value.
- Dial gauges or LVDT displacement transducers (resolution 0.01 mm) mounted on an independent, rigid reference beam that is completely isolated from the jack, reaction frame, and ground movement.
- A thermometer or correction procedure where temperature effects on readings may be significant.
Throughout all tests, the load is applied smoothly, alignment of the jack with the anchor axis is maintained, and readings of load and displacement are recorded at prescribed intervals to build the complete load–displacement–time history required for evaluation.
Applications
Soil anchor pull-out testing is indispensable wherever anchored systems resist lateral or uplift loads in soil. Typical applications include:
- Deep excavation support: verification of anchors for anchored diaphragm walls, soldier pile and lagging systems, and sheet pile walls supporting multi-level basement excavations in urban environments.
- Slope stabilization and landslide remediation: confirming anchor capacity for prestressed cable-and-frame support structures on highway cut slopes and unstable hillsides.
- Retaining structures: tie-back anchoring of gravity, cantilever, and reinforced concrete retaining walls.
- Anti-flotation (uplift) anchoring: anchoring of basements, underground tanks, and pump stations against groundwater buoyancy.
- Soil nailing walls: proof testing of soil nails for temporary and permanent excavations.
- Infrastructure anchoring: foundations of transmission towers, sign gantries, pipeline ballast blocks, and dock or wharf mooring structures where soil provides the resisting medium.
FAQ
Why must soil anchors be pull-out tested rather than verified by calculation alone?
The bond strength at the soil–grout interface depends on soil type, density, groundwater, borehole roughness, grouting pressure, and workmanship, all of which vary considerably even within a single site. Pull-out testing provides direct, measurable evidence of as-built capacity and safety, which empirical calculations cannot guarantee.
How many anchors must be tested on a project?
Basic (investigative) tests are performed on sacrificial trial anchors prior to production. For production anchors, acceptance tests are typically required for at least 5% of installed anchors (and not fewer than three) per project zone or soil condition, with additional creep tests where cohesive or creep-susceptible soils are present; the exact percentages are defined by the governing code and project specification.
What are the common failure modes observed during testing?
The typical failure modes include: pull-out of the grout body along the soil–grout interface (the intended governing mechanism for basic tests); internal failure between tendon and grout caused by inadequate grout strength or insufficient bond length; and, rarely, tensile failure of the steel tendon if the fixed length is excessive.
Can a failed anchor be reused or repaired?
Anchors that fail the acceptance test are generally not loaded to failure and may be de-rated in capacity, re-tested after supplementary grouting, or replaced. Anchors loaded to ultimate failure in basic tests are sacrificial and must not be incorporated into the permanent support system.
How is the design value derived from the test results?
The characteristic ultimate capacity obtained from basic tests is divided by a partial safety factor specified by the applicable standard (typically accounting for soil variability and structure importance) to establish the allowable design load, which then governs the lock-off and verification loads used for production anchors.