CROSS Safety Report
Fall of bridge deck support due to bolt over-tightening
This report is over 2 years old
Overview
A 7t steel frame designed to temporarily support part of a bridge deck during bearing refurbishments work fell from height as it was about to be lowered some 21m to the ground.
Key Learning Outcomes
For the construction team:
-
Consider introducing a quality control procedure for the inspection of safety critical connections
-
Connections can often be the weak link in structures and attention to detail is required
Full Report
Find out more about the Full Report
The Full Report below has been submitted to CROSS and describes the reporter’s experience. The text has been edited for clarity and to ensure anonymity and confidentiality by removing any identifiable details. If you would like to know more about our secure reporting process or submit a report yourself, please visit the reporting to CROSS-UK page.
A 7t steel frame designed to temporarily support part of a bridge deck during bearing refurbishments work fell from height as it was about to be lowered some 21m to the ground. An operative who was standing on the frame at the time of the dangerous occurrence suffered minor injuries, inflicted by his safety harness as it arrested his fall.
The frame was supported on two air hoists, each of which was supported from a stub cantilever. The stub cantilevers were bolted to the web of a steel beam. It was the failure of this bolted connection that the investigation by the Health and Safety Executive (HSE) and others identified as the principal cause of the accident.
The detailed investigation concluded that the principal cause of failure was the accidental over-tightening of the bolts using an impact wrench which damaged the threads. In testing carried out on behalf of the HSE, the damaging of the threads and eventual stripping of an M16 Gr 8.8 bolt was repeated, using an impact wrench. It is noteworthy that the impact wrench in question has a quoted bolt range of M16 to M22.
The detailed investigation concluded that the principal cause of failure was the accidental over-tightening of the bolts using an impact wrench which damaged the threads.
The torque graph from the impact wrench manual indicates that the appropriate torque for an M16 bolt can be achieved in less than a second. It was also noted that although the design considered the bolt to be a non-preloaded bolt, operator experience was solely relied upon to achieve the correct torque.
For design, the correct torque settings for safety critical connections should be stated clearly on the drawings. Furthermore, the HSE says that critical connection designs should be checked by an external consultant and that critical connections must use high strength friction grip (HSFG) bolts with load indicating washers or tension control bolts (TCBs).
Submit a report
Your report will make a difference. It will help to create positive change and improve safety.
Our secure and confidential safety reporting system gives professionals the opportunity to share their experiences to help others.
Feedback
This report includes a commentary on the failure of a group of M16 grade 8.8 bolts that connected a stub bracket to the web of a steel beam. The bolts had been excessively tightened by the erector, using an air-powered torque wrench. The fact that the grip length (and hence the thread length within the grip) was short was cited as one reason for the over-tightening of the bolts. The recommendation follows that 'critical connections should be checked by an external consultant and that critical connections must use HSFG bolts with load indicating washers or TCBs' (attributed to HSE). While the external checking of critical connections is seen as being a good recommendation, the effective banning of the use of 'untorqued' bolts in all critical connections is seen as misguided, particularly if (as appears to be the case here) the bracket had a single load application and thus fluctuating stress at the base of the thread was not a design consideration. The logic behind this recommendation appears to be that an HSFG bolt will not be broken by a air-powered torque wrench. In practice this is known to be wrong. If the wrench used is too big, then an HSFG bolt can be broken and it is down to the skill of the erector to ensure that this does not happen. Erectors have been using air wrenches to tighten (not to torque) for many years and, if they are experienced, this method has been seen to be safe and very cost-effective. SCI Advisory Desk Note AD302 is topical, which in turn refers to the NSSS 4th edition stating 'Bolts may be assembled using power tools or shall be fully tightened by hand' (this referring to non-preloaded bolts). The AD302 document, however, is concerned more with under-tightening bolts than over-tightening, though it does contain the caveat 'Ordinary bolts particularly those specified to BS 4190, should not be torqued to the values used for preloaded (HSFG) bolts because they have thinner nuts than preloaded bolts'. Other European countries, and the builders of mechanical installations in the UK use grade 8.8 bolts with (usually) grade 10 (but not HSFG) nuts under conditions of controlled torque to create bolted connections that work in the same way as those which the bridge industry uses HSFG bolts. The fact that the nuts used in such connections are significantly smaller than the HSFG variety means they use lower torque values and achieve similarly lower axial preload in the shank of the bolt. The lower torque values are demanded because of the danger of stripping the nut' thread due to excess hoop strain in the nut if the same torque is used as would be the case for an HSFG bolt. It has also sometimes been the practice to apply a grade 10 nut to grade 8.8 bolts when loaded in tension. Experienced erectors will, when tightening 'untorqued' bolts, either reduce the torque limit on the wrench, or throttle back on the air supply. The use of an air-powered wrench is not something for untrained erectors. The comments in Report 216 that experience and site control are necessary are therefore welcome. However, the effective banning of non-preloaded bolts from critical connections due to a single incident of improper site action is less appropriate - good guidance on the safe operation of air-powered wrenches would be a more useful tool.
I am somewhat surprised to read that the HSE recommends use of load-indicating washers or tension control bolts in place of calibrated torque wrenches. All three methods are prone to the risk that they can apply the incorrect bolt strain, and as a result the part-turn method has been traditionally preferred in the UK. SCI guidance SCI 7-05 gives more details of the problems with the other methods, but essentially they are all prone to creating an incorrect strain in the bolts, due to uncertainties in the thread friction, and the load indicating washers have the additional flaw of introducing a corrosion-prone crevice into the steelwork.
Unfortunately the comments on this report do not reflect the current bolt standards. HSFG bolts used to be specified to BS 4395. That is obsolete and has been superseded by BS EN 14399 (High Strength structural bolting assemblies for preloading). There is a grade 8.8 bolt to this standard. Therefore it is not so much the grade of bolt that was wrong but the fact that they were to the wrong standard.
Expert Panel Comments
Expert Panels comment on the reports we receive. They use their experience to help you understand what can be learned from the reports. If you would like to know more, please visit the CROSS-UK Expert Panels page.
The report states that this was a grade 8.8 bolt that failed. Such bolts should not be torqued because control is difficult and indeed they are not designed for this action. If pre-loading is required, the correct bolt to use is an HSFG bolt which has a heavier head and appropriate thread with a heavy nut to assure against thread stripping failure.
The report also states that the connection was through a beam web. That suggests the bolt grip length was very short and as such, any over torquing would have imposed a high plastic strain on very short length. Additionally, the torque tension relationship in bolts is very unreliable since the amount of friction within threads is uncertain.
If there is too much friction, then over torquing can cause serious damage. It appears that there was a failure to follow recommended procedures and there was a lack of experience. Good site control together with the proper use of HSFG bolts should prevent such occurrences.