What is an anchor pull test and do I really need one for my anchor installation?

Did you know that anchor testing isn’t always required in M&E installations? Are you aware of the instances where you would need to carry out 'pull testing'?

If you usually request ‘pull tests’ regardless, keep reading. It’s great that you are going the extra mile to be super compliant, but sometimes, there’s no need to waste your time or money if it’s redundant. If you don’t usually carry out any testing at all, also keep reading! Because of course, you do need to carry out the necessary testing when it is required. And either way, it provides peace of mind by helping the specifier select the suitable anchor, as well as acting as proof of a compliant installation. 

This article is intended to clear everything about pull tests up for you. It will cover:  

• What is a ‘pull test’? 

• What type of ‘pull tests’ are available? 

• Allowable load testing

• When and how is allowable load testing carried out?

• Proof testing

• When and how are proof tests carried out?
  
  
• What does BS 8539 tell us about pull testing? 

• When is carrying out a pull test not necessary? 

• Conclusion

What is a pull test?

Anchor testing, known colloquially as ‘pull testing’, is the process whereby an anchor/fixing is pulled testing of anchors may be done either before or after the installation depending on the type of testing required.  

What types of pull tests are available? 

There are two types of anchor testing. Sit tight, there’s a lot to get through here! The first of which is… 

Allowable load testing  

An allowable load test is carried out where the application involved is not covered by a relevant ETA Guideline, the strength/condition of the base material is unknown with no published data, or the manufacturer has no recommended resistance for the specifier to complete the selection process.  

When and how is allowable load testing carried out? 

It involves attaching a hydraulic pull tester to an anchor and pulling it to a designated test load, continuing to turn the handle and increasing the load until failure. A further calculation is carried out to work out the new recommended load. The anchor should also be tested five times and the average recorded. It’s considered ‘destructive’ as it is tested to the point of failure.  

 These should only be carried out on anchors that are approved by the manufacturer for use in the category of base material involved (e.g. solid, hollow, etc.), and before the anchor selection is finalised to ensure suitability and adequate resistance. They also need to be installed specifically for the purpose of the tests only, installed in the base material for the project but away from the anchors that will be used for the project. 

There are two principal circumstances when you should conduct an allowable load test: 

1. When an ETA-approved anchor is available but doesn’t meet the qualifications for dimensions and strength. For example, if the project is for an old building and you don’t have the design strength data of the concrete, then the ETA cannot apply because it’s missing an element that would tell us how the anchor would perform in that substrate. In this case, a test would be carried out by the ETA Guidelines. 

2. When there is no suitable ETA-approved anchor and another anchor is recommended by the manufacturer but there is no published tensile or design resistance.

Testing the substrate is simple for lightweight materials, but it can be more difficult for blocks of concrete beam sections since there is no clear way of showing the concrete's capacity. It can be difficult to precisely determine the allowable working load if you don't know the substrate's capacity. 

This way of testing requires more tests and more skill from the tester (advanced tester) as the loads are higher than its counterpart… 

Not sure about the design strength of the substrate or nervous about that anchor installation?

Proof testing

The purpose of proof tests is to test the quality of installation and these are the most frequently requested tests (despite not always being required). This is where, if we’ve used an ETA-approved anchor and it’s fixed into a known substrate, we can apply a test load to the anchor to prove that it’s been installed correctly. These are carried out on a sample of working anchors on every safety-critical application. 

When and how are proof tests carried out? 

The hydraulic pull tester is attached to the anchor and pulled to a specific test load. What it doesn’t tell us, though, is if it’s been specified correctly, or if it’s fit for purpose. 

This can be prove more difficult if the anchor is flush-fixed (bolt head anchor or a frame fixing), as it's harder for the tester to connect to this type of anchor.

Proof testing typically happens after the specification and installation stages and is conducted by a ‘general tester’. The contractor is responsible for organising it, not the installer. 

What does the BS8539 say about anchor testing? 

Everything mentioned up to this point is in line with BS8539. However, it includes other points that are worth mentioning. 

For starters, BS8539 states that tests need to be conducted by someone who’s not only competent but who also has a vast knowledge of anchors, how they’re installed, and how they’re likely to fail as this will greatly improve the quality of the tests. 

It defines the responsibility of the chosen tester as:   

• To ensure tests are carried out to the correct procedures.  

• Where relevant, anchors are installed by the manufacturers’ instructions, in the correct locations as required by the specifier.   

• Results are recorded in a full and comprehensive manner and forwarded to whoever requested the tests.  

• He must hold the relevant CFA qualification. 

Moreover, the CFA guidance highlights the following points from BS8539 that should be considered. 

• How many fixings should you proof test? 
• What load should you test to? 
• What is the percentage safety factor and what should it be going to? 

If you are testing 1 in 40 fixings, you are only allowed a factor of 1.5x the applied load. The actual applied load must not exceed the manufacturer's recommended load. If you test 1 in 25 fixings, the guidelines advise the applied load can be 1.25x the manufacturer’s specified load. 

To add, the code of practice indicates that once the testing has been done using the correct procedures, and they’ve fulfilled the objectives, the results are passed on to the person who requested the tests. This report should include: 

• Administration details: date of test, reason for test, name of person requesting test, unique reference test, client's company name, site location, contact name and position. 

• Anchor/application details: name of manufacturer, anchor type, size and finish, proposed application of anchor, design resistance and manufacturer's recommended resistance in the base material concerned (for proof tests) 

• Test objectives, test location, base material, installation details, test equipment details, test results, method statement, gauge calibration certificate. 

• Specific procedures for applying test loads, monitoring movement, etc. 

This is all in accordance with the CFA Guidance Note Procedure for site testing construction fixings. 

Lastly, the standard also highlights when testing isn’t required, which brings us to our next point… 

When is and when isn’t testing necessary? 

As mentioned, while in certain situations testing is required, a lot of the time requests are made for unnecessary testing to be carried out. I’m all for the enthusiasm and overachievement (especially when it comes to safety and compliance!), but you could really save yourself time, effort, and money by double-checking the requirements before going through with a test.  

We’ve already established that ETA-approved anchors should always be used whenever applicable and available. In the case that one is used and installed into a known substrate by a trained installer with competent supervision, the BS8539 states that testing actually isn’t required. That’s because for an anchor to be ETA-approved in the first place, it needs to undergo rigorous testing (it costs €1m!) that proves its performance.  

There are exceptions, though. Even with an ETA-approved anchor, on-site testing proving its suitability for the specific application may be required if, for example, it’s been approved for use in the category of the masonry used in the project, but doesn’t match the strength and/or dimensions of the approval. And of course, if the supervision isn’t satisfactory or operatives are not adequately trained, it defeats the purpose of using an ETA-approved anchor – in which case, testing can be conducted. 

The standard states the two key areas where anchor testing is recommended: 

1. To determine the allowable resistance when there is no manufacturer data and proof tests to validate the quality of the installation  

2. To understand the performance characteristics of both ETA-approved and non-ETA-approved anchors 

Conclusion 

Anchors are one of those things that are often neglected and seen as a commodity product, so people often underestimate their importance, as well as the importance of testing their performance. In reality, this is vital to ensure a safe and successful installation. But it might not always be your responsibility to undergo the testing. Sometimes the work has already been done for your – other times, your project, in particular, will prompt a request for testing. 

In any case, I hope this article has provided you with enough insight to feel more comfortable with the concept of anchor testing, familiarised you with when you’d be required to conduct it, and the processes behind the two types. Happy testing!