When Do You Need a Pull-Out Test for Anchors?

Pull-out tests for anchors — when are they essential, and when are they redundant? The answer depends on three factors: the anchor, the installer, and the substrate. Here is how to decide when to require a test, which type of test suits, and how to make sure the test is reliable.

Pull-out testAnchorsETAInstall qualitySubstrateChemical anchorsTesting

Formally, the European standard does not require a pull-out test when an anchor has a European standard approval (ETA) and is installed in accordance with the specification instructions in the certificate. In Israel there is no local standard on the matter. That is, formally, an anchor with an ETA certificate does not require testing — while an anchor without a European approval must undergo a pull-out test.

However, in many cases, pull-out tests carried out on anchors with a European standard revealed serious findings that prevented disasters. This raises the question: are pull-out tests essential, or can you rely on the standard alone? We will examine three factors that affect the decision: the anchor, the installer, and the substrate.

1. The anchor

When an anchor has a European standard approval (ETA), an orderly technical specification is attached to it, based on a rigorous testing process and a controlled production process. In this case, there is usually no need for an additional pull-out test.

When an anchor has no European approval, and its technical data is based solely on internal tests by an unrecognised manufacturer — pull-out tests must be carried out.

When the anchor is made by a known manufacturer and comes with orderly load tables, or with tests from a recognised laboratory under conditions similar to the planned application, you must examine the ratio between the estimated capacity and the load that will actually be applied. If the ratio is higher than 5, there is no need for a pull-out test. If the ratio is lower than 5, a test is recommended, since a single test cannot determine a general capacity.

2. The installer

Most failures in the anchoring field stem from poor execution and non-compliance with the manufacturer’s instructions — sometimes also from a lack of clear instructions. This is especially prominent in chemical anchors, where the rate of poor installations is relatively high. Professional training could help greatly, but few suppliers provide such a service.

In most cases, the purpose of the test is to examine the installation quality. But often, the supplier’s representative or the contractor installs the anchor intended for testing in a different, more meticulous way than the way the workers will install it in the field — thereby defeating the purpose of the test. For the test to be reliable, there must be no difference between the way the tested anchor is installed and the rest of the anchors in the project.

The recommended solution for examining installation quality

The inspector will accompany the installing worker for a few minutes, examine their working method, and ask them to install a further anchor in the same way — at a distance from the first anchor. This way you can verify that the tested anchor was installed by the same method as the rest of the anchors, and allow a real test to destruction.

A further problem arises when testing anchors to destruction: the test damages the adjacent area and requires repairs and additional work. On the other hand, a test to a partial load only (usually the design load) may not reveal a failure — since a poorly installed anchor may still withstand part of the required load.

This test is especially vital for chemical anchors (in mechanical anchors, install quality can usually be assessed by a visual inspection), and especially when the required loads are close to the maximum permitted values, or in facilities with large quantities of anchors.

3. The substrate

When the substrate is standard concrete in sound condition, the data is known and allows a direct calculation. But when the concrete is in a defective condition — wide cracks, corrosion attack, or concrete that does not meet the standard — it is highly recommended to carry out a pull-out test to examine the substrate’s behaviour. Important: the test must be carried out with the legs at a large distance from the tested anchor, to also allow examination of a failure of the substrate itself. Using a jack installed directly on the wall will not allow a correct test.

When there is no orderly technical data on the anchor’s performance on the specific substrate — and particularly in Israel, where the blocks differ from those common in Europe — you must not rely on foreign manufacturers’ tables. If the anchor has many tests on a certain substrate and the capacity-to-applied-load ratio is high, a pull-out test can be avoided. On the other hand, when there are few tests for the anchor on a specific substrate and you intend to use a significant part of the estimated capacity — a test is recommended.

When the substrate is not uniform, several tests must be carried out in different areas to get a full picture of the substrate’s behaviour under different conditions.

Types of pull-out tests

1. A test with displacement per the European standard’s guidelines

This is a test that few laboratories in the country are certified to perform. Its great advantage: after the initial tests are performed, you can test anchors installed in the project with displacement measurement up to the design load and compare the displacement-load curve to the initial tests — thereby verifying that the installation was carried out correctly without damaging the anchor.

2. A recognised-lab test without displacement, with direct pressure on the wall (no table)

This is the most common test in Israel. It has an important limitation: the pressure created around the anchor by the jack’s environment slightly improves the test results in the anchor’s favour and can give a more optimistic picture than the real situation. In addition, in a problematic substrate, the test does not allow examination of a concrete-cone failure.

A central question in this test is the method of load application: do you raise the load continuously until failure, or in stages with waiting times for stabilisation? The second solution is professionally preferable, but anchors usually require a long stabilisation time at high loads — because of the system’s elasticity (the device, the legs, the steel) — and in Israel it is often not waited long enough, which leads to an incorrect determination of failure.

Recommendation: apply the load in stages with a wait of about half a minute between one stage and the next, while ignoring slight load drops. After a drop, try to increase the load again to distinguish between elastic movement and a final failure.

3. A recognised-lab test without displacement, with a table (which prevents direct pressure around the anchor)

This method offers two main advantages over the previous test:

• The tested anchor’s capacity is not affected by external pressure around the anchor.
• You can examine and detect a failure in the substrate itself.

In the other parameters, everything said about the previous test method applies here too.

4. A test by the anchor supplier

Everything said above applies also when the tester is an employee of the anchor supplier — however, there is a built-in conflict of interest. An example: a well-known and respected supplier in the country submitted a pull-out test report for a 12 mm-diameter rebar with its adhesive, in which the tested load exceeded the maximum capacity of a rebar of that diameter by 50%. The contractor did not notice this until it was brought to their attention — since a rebar connected with chemical adhesive cannot hold more than the capacity of the steel itself. In short: do not rely on a supplier to test its own product without close supervision.

During a supplier’s test, verify the following points:

• The anchor installation method: was it installed with the same tools and accessories the workers use in the field? Were extra cleaning operations performed that will not be done in practice? Was it installed to the required depth? Was adhesive added that is not part of a regular installation?
• Contact between the device parts: supplier devices are usually built from a grip part for the anchor head, a small table, and a hydraulic mechanism. If the grip part touches the table during the test, the measured load will be artificially high. A gap must be kept between the parts throughout the test.
• Video documentation: it is highly recommended to require video documentation of the test. Many testers object to this — but whoever claims transparency should not fear full documentation. A supplier, who naturally raises suspicion, is obliged to prove extra transparency.

Summary — principles for deciding on a pull-out test

You must take into account the anchor’s quality, the installer’s identity and the substrate’s conditions — and by a combination of these factors decide when to require a pull-out test and which type of test suits.

Demand full transparency from the tester: what is the test method, how is failure defined, and how was the tested anchor installed. Remember: to examine the anchor’s behaviour reliably, it must be tested under conditions identical to those in which it will actually operate — in every respect. The inspector’s role also includes examining the full match between the sample-test conditions and the real anchor in the field.

Sources

• ADIT engineering guidance on anchor pull-out testing (when required, test types, reliability).
• European anchor approval (ETA) standards and Israeli-substrate considerations.