Practical issues with on-site testing of anchors 27 March 2018

By John Muir, technical manager, Construction Fixings Association

In our recent article in Fastener + Fixing Magazine (published in Issue 109, pages 130 – 131) we discussed site testing of anchors and more importantly when a test is required and when not.

Having established that an anchor test regime is required, we need to consider the practical issues that are involved in such tests.

As covered in our previous article, there are two key reasons to tests anchors:

  • Proof testing of working anchors on a project to verify the quality of the installation.
  • Tests to establish the allowable load for the specific application where no performance data is available. These tests can involve testing to failure to enable calculation of the allowable load.

There are, however, a number of practical issues that need to be taken into account to enable testing of anchors on-site. For allowable load testing, the anchors used are sacrificial and used only for the test. As such all the practical issues can be taken into account in the planning of the testing exercise.

Practical issues with on-site testing

Proof testing involves working anchors that are part of the completed project, and will involve many of the issues we will examine here:

  1. The fixture is in place at the time of the test: This is likely to be a steel plate and section, or bracket, and with this in place it may not allow the test meter to be located in position axially over the test anchor, which is essential for an accurate test result. It is also vital that the support legs of the test meter bridge are not sitting on the fixture, and are in fact located so that the bridge is spanning across the fixture. This is required to enable the tensile load on the anchor to be applied with no restraint from the plate, to restrict failure, particularly in the case of concrete cone failure.
  2. An anchor type that has a projecting stud with nut and washer: With this type of anchor, it is important that the test adaptor has sufficient thread engagement on the anchor stud. Therefore when it is likely that proof testing on a random sample of installed anchors is required, it is necessary that the following actions should be taken at the installation stage. In the case of torque controlled expansion anchors, a sufficient section of the threaded stud projects above the nut after setting. This enables connection of the test adaptor. However, if there is only a small number of threads projecting (generally less than four) then the test load may fail through thread stripping of the anchor. It is advisable therefore to insert the anchor with one or two threads exposed above the nut at installation stage, whilst ensuring the required embedment depth is still achieved. For resin bonded stud anchors it is important to ensure there are sufficient threads exposed after insertion of the anchor rod into the resin mortar, to accommodate the fixture and packing, nut and washer, and the test adaptor. This should be achieved whilst ensuring the required embedment depth is still achieved.
  3. The anchor is a bolt head type: Many anchors are configured with a bolt head, which does not directly allow for the connection of test apparatus. Proof testing of this type of anchor should be considered very carefully, as it involves considerable disturbance of the anchor that can lead to a negative influence on the anchor’s performance. One method to test such a bolt head, is to slacken off the bolt to a dimension suitable to allow a test piece to be inserted below the head of the bolt, to facilitate the test. However, if the length of the bolt does not allow the required number of threads to be engaged within the cone, to sustain the test load after unscrewing the bolt from the cone, then in this case it is not possible to test the anchor. An alternative method is to remove the existing anchor bolt and replace with a longer test bolt to ensure sufficient thread engagement. This method runs the risk of dislodging the cone and any collapsible section that exist in some anchors. Should this be the case, it may not be possible to replace the bolt and thus make a working anchor unusable. For this reason, this method is not recommended.

There will be occasions where the difficulties outlined above will effectively prevent anchor testing being a viable option. Where this is the situation, we can look at alternative methods of providing a quality control check. 

Torque checking of anchors 

For all anchorages, the anchors require a specific torque to be applied to induce a clamping force, which ensures the fixture is held tightly to the base material with no movement (for torque controlled expansion anchors, the torque applied also sets the anchor). These torque values are determined by the manufacturer through calculation and testing to induce a pre-stressing force on the anchor. This pre-stressing generates the clamping force in the anchorage, which is balanced by a tension force in the anchor bolt. The magnitude of this induced tension force is well in excess of the applied (characteristic) action.

To understand how this force is induced from the torque setting, we can look at the relationship between torque and tensile pre-stressing (or clamping) force as shown in the equation below.

It may be easier to illustrate this with a simplified equation:

In this equation k is a constant, which is more or less independent of the thread diameter for the same form and surface roughness bolt head and thread.

Example An M12 anchor electrogalvanised dry torqued to 50Nm (using k = 0.2 for worst case):

The recommended resistance of a 12mm bolt anchor in C30/37 non-cracked concrete is circa 13kN. A test load would therefore be typically 13 x 1.5 = 19.5kN. Values in cracked concrete would be even less.

By applying 50Nm torque you have already induced an approximate 20.83kN pre-stressing force in the bolt.

This represents a minimum factor of 1.6 over recommended load, which is likely to be closer to 2 over applied load. Since the calculation of test load uses a factor of 1.5 (BS8539:2012), the tension load applied to the bolt through torque setting will be in excess of the test load.

Whilst torque checking can never replace the accuracy of tension load testing, it can provide an indication of load applied to the anchor through pre-stress force. Since this load is in excess of the test load, torque checking can be used as a practical alternative.

When using the torque checking method there is one other issue that needs to be considered. After installation, the pre-stressing force (and thus clamping force) will decrease as a result of levelling irregularities in the stressed thread pitches, as well as creep and relaxation of the stressed concrete.

Over time, the pre-stress force can reduce to 40% – 50% of the original value. However, the greatest reduction takes place over the first 48 hours, so this must be taken into consideration when torque checking is carried out.

So, where we have issues as illustrated in sections 1, 2 and 3 the above use of torque checking can provide a good indication of installation quality.


Will Lowry Editor t: +44 (0) 1727 814 509


Will joined Fastener + Fixing Magazine in 2007 and over the last 12 years has experienced every facet of the fastener sector - interviewing key figures within the industry and visiting leading companies and exhibitions around the globe.

Will manages the content strategy across all platforms and is the guardian for the high editorial standards that the Magazine is renowned.