A request to pull test the anchors to 10kN may sound clear on a construction site, but under BS 8539 it is not enough information to define a technically valid anchor test.
The first question is not how many kilonewtons the testing equipment can apply. It is what the test is intended to prove. BS 8539 separates site testing into fundamentally different objectives: testing used to establish whether an anchor is suitable for a particular base material and testing used to check the quality of anchors already installed for the works.
Those objectives lead to different test anchors, different sample sizes, different loads, different calculations and different engineering conclusions. Confusing them can result in a production anchor being unnecessarily overstressed, an unsuitable anchor being accepted on the basis of a proof test, or a test report being asked to answer a structural-design question that the test was never capable of answering.
Illustrative anchor proof-testing arrangement. Under BS 8539, proof testing is used to validate installation quality and should not be confused with testing to determine allowable resistance.
The key message: a BS 8539 proof test is principally an installation-quality check on working anchors. Testing for allowable resistance is a different process used where suitable performance data for the actual base material is unavailable. Passing a proof test does not establish the ultimate capacity or design resistance of the anchor.
Jump to: Two different questions | Terminology | Current standard | Proof testing | Sample rates | Failures | Allowable resistance | ETA decision route | Preliminary tests | Reaction bridge | Responsibilities | Test request | Test report | Common mistakes | Checklist | FAQ
BS 8539 Starts With Two Different Engineering Questions
Clause 9 of BS 8539 distinguishes between testing carried out before anchor selection is finalised to establish suitability and allowable resistance, and testing carried out after installation to validate installation quality.
| Question | Proof Testing | Allowable Resistance Testing |
|---|---|---|
| Main purpose | Validate the quality of installation of anchors intended to remain in service. | Establish suitable site-specific resistance where appropriate published performance data is unavailable. |
| Typical anchors tested | A sample of installed working anchors. | Anchors installed specifically for testing and not intended for use in the works. |
| Typical objective | Check that installation workmanship has produced satisfactory anchorages. | Determine whether the proposed anchor can provide adequate resistance in the actual substrate. |
| Does it establish ultimate capacity? | No. The proof load is deliberately below a destructive capacity investigation. | Depending on the regime, ultimate/failure results may form part of deriving an allowable or characteristic resistance. |
| Relevant BS 8539 route | Clause 9.3 and Annex B.3. | Clause 9.2 and Annex B.2. |
This distinction is the foundation of a technically correct test instruction. A contractor cannot turn a proof test into a capacity test simply by increasing the load, and a tester cannot derive a reliable allowable resistance from a single working anchor that happens to survive an arbitrary load.
STRUCTinspect has previously covered the broader requirements in BS 8539 Anchor Testing: What Contractors Need to Know and the distinction between terminology in Anchor Proof Testing vs Pull-Out Testing: What Is the Difference?.
The Terminology Engineers Need to Keep Separate
A recurring problem in anchor testing is the mixing of actions, resistances and test loads. These are not interchangeable values.
| Term | Notation | Practical Meaning |
|---|---|---|
| Characteristic action | NEk | The characteristic tensile action applied to the anchor by the fixture. Often described in practice as the unfactored applied load. |
| Design action | NEd | The action after the appropriate action factors have been applied for limit-state design. |
| Characteristic resistance | NRk | A statistically derived characteristic capacity associated with the relevant failure mechanism. |
| Design resistance | NRd | Characteristic resistance reduced by the applicable material partial safety factor. |
| Recommended resistance | Nrec | Maximum working resistance recommended by the manufacturer for the relevant anchor and base-material condition. |
| Allowable resistance | NR,all | A working resistance derived from site testing where the appropriate published resistance is unavailable for the actual substrate. |
| Proof load | Np | The tensile load applied during a proof test of an installed anchor. |
| Preliminary test load | Ntest | The calculated load applied to specially installed test anchors under the preliminary-test regime. |
Action is demand. Resistance is capacity.
A proof test load is neither the anchor's ultimate resistance nor its design resistance. It is a test load derived for a defined quality-control purpose.
A proof test load is neither the anchor's ultimate resistance nor its design resistance. It is a test load derived for a defined quality-control purpose.
BS 8539:2012+A1:2021 in 2026
At the time of writing, BSI identifies BS 8539:2012+A1:2021 — Code of practice for the selection and installation of post-installed anchors in concrete and masonry as the current release.
A BSI revision project is also under way. This is important because the standard itself still normatively references an earlier CFA site-testing guidance document, while the Construction Fixings Association has since issued its Procedure for Site Testing Construction Fixings 2024. The CFA states that the updated guidance works alongside BS 8539 and includes revised detail particularly relevant to masonry testing.
Engineers preparing current specifications should therefore check the latest status of BS 8539, the current CFA guidance and any relevant ETA, EAD, project specification or sector-specific requirement before issuing a test instruction.
What Does a BS 8539 Proof Test Actually Prove?
Proof testing under Clause 9.3 is intended to check the quality of installation of anchors that are to remain in service. It is therefore fundamentally a quality-assurance test, not a destructive investigation into the maximum capacity of the fastening.
A satisfactory proof test demonstrates that the tested anchorage sustained the specified proof load without movement, damage or deformation under the test conditions.
What it does not independently establish includes:
- the ultimate tensile resistance of the anchor;
- the characteristic resistance of the anchorage;
- the design resistance under BS EN 1992-4;
- the long-term creep behaviour of a bonded anchor;
- performance under every environmental condition;
- the adequacy of an anchor outside the base-material scope of its performance data;
- the adequacy of every untested anchor merely because a sample passed.
BS 8539 specifically describes proof tests as providing a modest safety margin without risking the integrity of the working anchor. That is very different from loading a sacrificial test anchor towards failure in order to obtain information about substrate resistance.
How Is the BS 8539 Proof Load Calculated?
Annex B.3 gives the basic relationship:
Np = NEk × Î½P,test
Where:
Np = proof test load
NEk = characteristic tensile action
νP,test = proof-test load factor
Where:
Np = proof test load
NEk = characteristic tensile action
νP,test = proof-test load factor
| Proof-Test Sample | νP,test | Minimum Requirement |
|---|---|---|
| 2.5% of installed anchors | 1.5 | At least three tests, subject to discrete-area requirements. |
| 5% of installed anchors | 1.25 | The higher sampling rate permits the lower proof-load factor. |
BS 8539 states that the proof-test factor should never exceed 1.5. The current CFA guidance also states that the calculated proof load should not exceed 1.5 times the manufacturer's recommended resistance for the relevant anchor and substrate, taking account of applicable edge-distance and spacing reductions.
Simple Proof-Test Example
Assume the engineer has established a characteristic tensile action of:
NEk = 8kN
At a 2.5% sampling rate:
Np = 8 × 1.5 = 12kN
At a 5% sampling rate:
Np = 8 × 1.25 = 10kN
At a 2.5% sampling rate:
Np = 8 × 1.5 = 12kN
At a 5% sampling rate:
Np = 8 × 1.25 = 10kN
Those numbers are not interchangeable with anchor design resistance. They are proof-test loads derived from the engineer's characteristic action and the selected sampling regime.
Important: giving the tester only NEd, the factored design action, is not the same as providing NEk. The proof-load equation is based on the characteristic action.
Why Does BS 8539 Talk About “Discrete Areas”?
The minimum of three proof tests is not necessarily three tests across an entire large project. Annex B.3 applies the minimum within discrete areas where the installation population may differ.
Factors that can create separate test populations include:
- different anchor types;
- different base materials;
- different exposure or weather conditions;
- different elevations;
- different installation teams.
This is an important practical point. Three satisfactory tests in dry internal concrete do not automatically provide evidence about anchors installed by another crew into external masonry on a different elevation.
What Happens If a Proof Test Fails?
A proof-test failure is not simply a failed individual fixing to be replaced and forgotten. It may indicate a problem with selection, substrate condition or installation quality across a wider population.
| Result | BS 8539 / CFA Response |
|---|---|
| No movement, damage or deformation | The tested anchor has satisfied the specified proof-test objective. |
| One failure in a discrete area | Investigate the reason and increase the testing population. Current CFA guidance states 5% and at least six anchors. Annex B.3 also refers to 5% or 10% depending on the proof-load level. |
| More than one failure | Test 100% of the anchors in the affected population, determine the reasons for failure and reconsider the specification. |
| Failed fixing | Normally requires replacement, with the replacement specification confirmed by the specifier. |
Current CFA guidance defines a satisfactory proof test by the anchor holding the required load without movement, damage or deformation to the fixing or base material. Any anchor exhibiting movement or damage should be recorded as a failure.
Do not automatically apply a generic 60-second hold rule. The current CFA 2024 procedure requires load to be applied slowly and progressively and, where movement is not being recorded in detail, to reach the proof load over approximately 30 seconds. Project or sector-specific requirements may add other holding requirements, but they should be stated explicitly rather than assumed.
What Is Testing to Determine Allowable Resistance?
Testing for allowable resistance addresses a different engineering problem: what resistance can reasonably be assigned to the proposed anchor in the actual base material when suitable published performance data is not available?
BS 8539 identifies this most commonly with variable masonry such as brickwork, blockwork or stonework where the actual substrate may not be adequately represented by published manufacturer values. The test anchors used for this purpose are installed specifically for the testing programme. They are not working anchors that will subsequently support the fixture.
This distinction matters because some allowable-resistance regimes deliberately load anchors to failure, while the preliminary simplified approach may escalate to failure testing if the required preliminary test load is not reached.
How the ETA Changes the Testing Decision
| Situation | Typical BS 8539 Position |
|---|---|
| ETA-covered anchor in concrete and site conditions fall within the applicable approval/design basis | Testing to determine allowable resistance is generally not required. Published design data is used. |
| ETA covers the category of masonry and the actual masonry meets the relevant strength and dimensional qualifications | Site testing to establish resistance may not be required; the applicable ETA data can be used. |
| ETA covers the general masonry category but the actual units do not meet the required dimensions or strength | BS 8539 directs the testing regime to Annex B.2.2 or the relevant EAD procedure. |
| No suitable EAD/ETA performance route for the actual substrate, but manufacturer approves use in that general material type | Annex B.2.3 provides routes for establishing allowable resistance. |
| Concrete condition gives reason to doubt whether published performance can be achieved | The responsible engineer may need a site-specific investigation/testing approach rather than blindly applying published capacity. |
The presence of an ETA therefore does not mean that every piece of masonry encountered on a project automatically matches the assessed substrate. Equally, the absence of a matching published resistance does not mean that an arbitrary proof test can be substituted for the Annex B procedure.
Preliminary Tests: The Simplified Allowable-Resistance Route
For anchors without the relevant EAD route, Annex B.2.3.1 provides a simplified method traditionally known in the fixings industry as preliminary testing.
The concept is to test at least five specially installed anchors to a calculated test load. If all five achieve the required load satisfactorily, the required characteristic action can be taken as the allowable resistance for that application under the stated procedure.
Preliminary testing is not the same as proof testing.
The five anchors are installed for the test programme. They are not five production anchors randomly selected from the finished installation.
The five anchors are installed for the test programme. They are not five production anchors randomly selected from the finished installation.
The current CFA 2024 guidance expands the field calculation to take account of factors including the test-load factor, reaction-bridge geometry and site/application adjustment factors. Depending on the application these can address conditions such as long-term tensile loading, wet substrate, elevated service temperature, mortar-joint uncertainty and cracked-concrete assumptions.
If One Preliminary Test Anchor Does Not Reach Ntest
A failure to reach the calculated test load does not mean that the tester simply lowers the target until all five anchors pass.
The situation returns to the specifier for review. Options may include changing the anchor type, diameter or embedment, increasing the number of working anchors where structurally appropriate, or progressing to further failure testing and calculation under the relevant Annex B procedure.
Where the same anchor is retained following a failure close to the required test load, Annex B provides a route based on failure results, mean and lowest failure loads and the applicable safety factors. That process is an engineering assessment of the test series, not a simple “lowest pull wins” rule.
Statistical Testing to Failure
Annex B also provides statistical routes in which multiple specially installed anchors are loaded to failure. These regimes record parameters including the load at first movement, ultimate resistance and failure mode.
The result is then treated statistically. Sample size matters because the standard uses a statistical factor that reduces as more tests are undertaken. The variation between results also matters: a high average failure load combined with very inconsistent test results does not justify treating the average as a reliable resistance.
| Example Sample Size | K Factor in BS 8539 |
|---|---|
| 5 tests | 3.40 |
| 10 tests | 2.57 |
| 15 tests | 2.33 |
The important site lesson is that an allowable resistance is the result of a defined test regime and calculation. It is not simply the load displayed on a hydraulic gauge when an anchor first begins to move.
Why the Test Rig Can Change the Result
The reaction arrangement is not merely a practical way of holding the hydraulic tester against the structure. Its geometry can influence the measured result.
During a tensile test, the equipment pulls the anchor away from the base material while the bridge supports react against the surrounding material. If those supports are too close to the anchor, they can restrain the natural development of a concrete cone or local masonry failure mechanism.
The result can therefore be artificially enhanced because the testing equipment is effectively confining the same material that is trying to fail.
| Test Type | Current CFA Guidance |
|---|---|
| Ultimate / failure testing in concrete | Closest bridge support should ideally be at least 1.5 × effective embedment depth from the anchor centreline. |
| Preliminary test before possible failure loading | At test-load stage, CFA guidance permits closer geometry, but spacing may need increasing if anchors are subsequently taken to failure. |
| Proof testing | Closer support positions can be acceptable because the objective is not to develop an ultimate concrete cone failure. |
A high failure load obtained with an incorrectly positioned reaction bridge may be less useful than a lower result obtained from a test arrangement that allows the real failure mechanism to develop.
Testing Does Not Prove the Structure Can Carry the Load
There is another important limitation that is easily missed. Before anchor selection, the designer has to establish that the structure itself can sustain the design actions transferred through the anchorage.
A pull-testing rig reacts its test load locally back into the structure around the test anchor. That test arrangement does not prove that the entire slab, wall, beam or other structural element can safely transfer the eventual design action through the real structural load path.
Anchor testing therefore cannot replace structural assessment of the supporting element.
Who Is Responsible for What?
BS 8539 places different responsibilities on the designer, specifier, contractor, installer, supervisor and tester. Site testing does not transfer the design responsibility to the technician operating the hydraulic equipment.
| Role | Relevant Responsibility |
|---|---|
| Designer | Responsible for the overall structural design and for providing the design information needed for anchor selection. |
| Specifier | Selects and specifies the anchor, identifies whether testing is required and defines the objective and relevant test information. |
| Contractor | Ensures the specified anchor is procured, competent installation/supervision is provided and required testing is arranged. |
| Installer / supervisor | Installs and supervises the anchors in accordance with the specification and manufacturer's instructions. |
| Tester | Carries out the specified tests using the appropriate procedure and records and communicates the test results. |
BS 8539 requires testers to be competent and states that knowledge of anchor behaviour, installation and likely failure mechanisms is desirable. The CFA operates an Approved Tester scheme, but the British Standard does not say that possession of a particular commercial accreditation is the only possible route to competence.
Likewise, a testing company should not automatically be asked to redesign the fixing after a failure. Current CFA guidance says the report should primarily be a factual statement of the test and results. Unless suitably qualified to make wider anchorage-engineering judgements, the tester should not diagnose structural causes or prescribe remedial design.
What the Engineer Should Give the Tester Before Attendance
BS 8539 Clause 6.6 is particularly useful because it sets out information that should be passed from the specifier or responsible engineer to the tester.
| Information | Why It Matters |
|---|---|
| Test objective | Defines whether the exercise is proof testing or determination of allowable resistance. |
| Anchor designation | Confirms exactly what product and size are being tested. |
| Installation details | Allows relevant embedment, torque, drilling, cleaning and curing requirements to be checked. |
| Base material and strength if known | Anchor performance is substrate-dependent. |
| Ntest or Np | The tester should not be expected to invent the engineering test load on arrival. |
| Direction of loading | Tension and shear represent different structural behaviours. |
| Number of tests | Controls the required sampling regime. |
| Locations / edge requirements | Ensures the tested population and rig arrangement represent the engineering requirement. |
| Access constraints | Can determine equipment, bridge arrangement and safe working method. |
“Please attend site and pull test the anchors to 10kN” is not a complete BS 8539 test instruction.
What Should a BS 8539 Anchor Test Report Contain?
A useful report should allow the engineer to understand exactly what was tested, where it was tested, how it was tested and what happened.
| Report Item | Typical Detail |
|---|---|
| Administration | Date, project, reason for test, client, requestor, tester and witnesses. |
| Anchor details | Manufacturer, type, size, finish and intended application. |
| Test objective | Proof testing or testing to determine allowable resistance. |
| Test location | Unique location, drawing/sketch, edge distance, spacing and structural thickness where relevant. |
| Base material | Material type, strength where known and whether solid or hollow. |
| Installation information | Hole diameter/depth, embedment, cleaning, torque and bonded-anchor curing conditions where applicable. |
| Equipment | Tester/ram details, capacity, calibration information and reaction-frame geometry. |
| Results | Maximum applied load, movement where required, condition of anchor/base material and failure mode where relevant. |
The current CFA guidance recommends keeping the report as a factual record and limiting the conclusion to whether the tested fixings met the stated test objective. A statement such as “all anchors are structurally safe for the proposed application” goes beyond the test data unless the person making that statement also has the information, competence and responsibility required to make the engineering assessment.
Common BS 8539 Testing Mistakes
| Mistake | Why It Matters | Better Approach |
|---|---|---|
| “Pull test to 10kN.” | Does not define the test objective, action, sample rate or basis for the load. | Issue the test type, calculated test load, population, locations and anchor details. |
| Using one anchor to establish a “safe working load”. | One result does not provide the required site-test regime or statistical confidence. | Use the relevant Annex B allowable-resistance procedure. |
| Treating a proof-test pass as proof of ultimate capacity. | The anchor was not taken to its ultimate failure condition. | Use approved design data or the correct allowable-resistance test regime. |
| Using NEd as though it were NEk. | Can produce the wrong proof-test load. | The responsible engineer should provide the correct characteristic action and proof load. |
| Using production anchors for destructive allowable-resistance tests. | The anchor may have been permanently affected by the test. | Install dedicated test anchors in representative substrate. |
| Reaction bridge too close to the anchor during failure testing. | The rig may restrain the failure cone and artificially increase the result. | Use reaction geometry appropriate to the test objective and effective embedment. |
| Ignoring one proof-test failure after replacing the individual anchor. | The failure may indicate a wider installation-quality problem. | Investigate and escalate the sampling regime. |
| Mixing different substrates or crews into one sample population. | Local installation problems can be hidden by unrelated successful tests. | Define discrete test areas. |
| Calling every tensile site test a “pull-out test”. | The term says nothing about the actual test objective and can also describe a failure mode. | State proof test, preliminary test or allowable-resistance test explicitly. |
| Asking the tester to declare the structural connection safe. | Test data is only part of the design and acceptance process. | Tester reports the evidence; responsible engineer/specifier reviews it in context. |
Checklist Before Ordering BS 8539 Anchor Testing
- Purpose: proof test or determination of allowable resistance?
- Anchor: confirm manufacturer, designation, diameter, length and anchor type.
- ETA/EAD: establish whether the product and actual substrate fall within the applicable performance scope.
- Base material: identify concrete, brick, block, stone or other approved category and record strength where known.
- Action: provide the characteristic tensile action NEk where relevant.
- Proof load: provide Np rather than expecting the site tester to derive it without design information.
- Preliminary test load: provide the calculated Ntest for the applicable regime.
- Population: define how many anchors are installed and the required test percentage.
- Discrete areas: separate different anchors, substrates, exposures and installation crews where necessary.
- Locations: mark working anchors or dedicated test locations clearly.
- Edge distance and spacing: provide project-specific geometry where it affects resistance.
- Embedment: confirm effective embedment depth.
- Installation: confirm hole preparation, cleaning, setting torque and resin curing requirements.
- Reaction bridge: ensure the test rig is appropriate to the test objective and substrate.
- Calibration: verify current calibration of the test equipment.
- Acceptance: identify who will review the test report and make the engineering decision.
Before Anchor Testing Is Booked
The testing instruction should define the anchor, substrate, objective, test load, sample size and locations before mobilisation. Where these have not been established, they should be confirmed by the responsible engineer or anchor specifier rather than improvised at the test location.
Who Decides Whether the Anchors Pass?
There are two levels to this question.
At the individual test level, the tester can record whether the anchor satisfied the defined test objective. For example, a proof-tested fixing can be reported as having sustained the specified proof load without observed movement or damage.
The wider engineering decision remains different. Whether the completed anchorage system is acceptable for the intended structural application depends on the specification, design actions, resistance calculations, substrate, installation records, ETA/design data and any required remedial action.
The test equipment measures what happened. It does not replace the design.
Evidence-Based Summary
BS 8539 recognises two fundamentally different reasons for anchor site testing.
Proof testing checks the quality of installation of anchors intended to remain in service.
Testing for allowable resistance establishes a site-specific resistance where appropriate published data for the actual base material is unavailable.
A proof-test pass does not establish ultimate, characteristic or design resistance.
The normal Annex B.3 proof load is Np = NEk × Î½P,test, with factors linked to the selected sampling rate.
The minimum proof-test sample is normally 2.5% of installed anchors and at least three within each relevant discrete area.
One failure triggers investigation and increased testing; more than one failure can trigger 100% testing and reconsideration of the specification.
Allowable-resistance testing normally uses specially installed test anchors rather than working anchors.
Reaction-frame geometry can materially influence failure-test results and must suit the objective of the test.
The tester produces measured evidence. The responsible designer/specifier retains the wider engineering decision.
FAQ: BS 8539 Proof Testing and Allowable Resistance
What is a BS 8539 proof test?
It is a test normally carried out on a sample of installed working anchors to validate installation quality. It is not intended to determine the ultimate capacity of the anchor.
Does passing a proof test prove the anchor's design capacity?
No. A proof test confirms performance only against the specified proof-test objective. Anchor design resistance comes from the relevant design/assessment framework or an appropriate allowable-resistance test regime where published data is unavailable.
What percentage of anchors should be proof tested?
BS 8539 Annex B.3 gives a normal minimum sample of 2.5% of the installed anchors, equivalent to one in 40, with at least three tests. The minimum applies separately to relevant discrete areas.
What proof-load factor is used?
Annex B.3 gives νP,test = 1.5 when 2.5% of anchors are tested and 1.25 when 5% are tested. The proof-test factor should not exceed 1.5.
What happens if one proof-tested anchor fails?
The reason should be investigated and the sample increased. Current CFA guidance states that testing should be doubled to 5% and at least six anchors in the affected discrete area. Annex B.3 also references 5% or 10% depending on the proof-load level.
What happens if more than one anchor fails?
BS 8539 states that 100% of the anchors in the relevant population should be tested, the reasons determined and the specification reconsidered.
Can a single pull test establish the safe working load of an anchor?
Not under the normal BS 8539 allowable-resistance procedures. The applicable Annex B regime requires a defined number of dedicated test anchors and an assessment method appropriate to the substrate and anchor.
Are proof-test anchors sacrificial?
Normally no. Proof tests are carried out on working anchors and are intended not to compromise their integrity. Anchors installed specifically to determine allowable resistance are different and should not subsequently be used for the project.
Does an ETA mean anchors never need site testing?
No. It depends on the objective and the actual conditions. Suitability testing is normally unnecessary where applicable published performance covers the substrate and conditions, while proof testing may still be required where installation-quality assurance is needed. Conversely, masonry outside the dimensional or strength qualifications of the relevant assessment may require site testing.
Who calculates the proof-test load?
The proof-test level should be defined through the fixing specification by the responsible engineer/specifier. BS 8539 requires the proof-test load to be provided to the tester.
Can the testing technician say the anchors are structurally safe?
The technician can report whether the tested anchors met the defined test objective. Wider structural acceptance requires review by the person responsible for the anchorage specification and design.
Primary Technical References
The principal UK reference used for this article is BS 8539:2012+A1:2021 — Code of practice for the selection and installation of post-installed anchors in concrete and masonry, particularly Clauses 4, 6.6 and 9 and normative Annex B.
Site-testing methodology has also been reviewed against the Construction Fixings Association Procedure for Site Testing Construction Fixings 2024, which provides updated practical guidance on allowable-resistance testing, proof testing, test equipment, reaction spacing, reporting and tester competence.
For fastening design in concrete, the wider design framework includes BS EN 1992-4:2018 — Eurocode 2: Design of fastenings for use in concrete and its UK National Annex, together with the applicable ETA/EAD and manufacturer data for the anchor system.
Source Context and Technical Note
This article is a STRUCTinspect technical explainer intended to help structural engineers, temporary works engineers, contractors, project managers and testing teams distinguish between anchor proof testing and testing used to determine allowable resistance.
BS 8539 is a code of practice and should be applied alongside the actual anchor specification, applicable ETA/EAD, current manufacturer instructions, relevant structural design standards and any sector-specific requirements. The correct testing regime depends on the purpose of the test, the anchor system, the substrate, installation conditions and the engineering design.
This article does not provide structural design, fixing design, legal or contractual advice. Test loads, sample sizes, test locations, anchor selection, acceptance criteria and any remedial action should be confirmed by the competent engineer or specifier responsible for the anchorage.
