Skip to main content

CROSS Safety Report

Adequacy of termination connectors for tensile bars

Report ID: 256 Published: 1 July 2012 Region: CROSS-UK

This report is over 2 years old

Please be aware that it might contain information that is no longer up to date. We keep all reports available for historic reference and as learning aids.

Overview

The following report is intended to raise concern regarding the structural adequacy, quality control and inspection techniques used in the manufacture of proprietary bar systems, in particular the cast end terminations of the bar systems.

Key Learning Outcomes

For civil and structural design engineers:

  • When specifying proprietary tension bars, it is good practice to specify the required properties of the end connections and use expert advice as necessary

  • It is also good practice to specify the tests that will be carried out on completed components and state the required acceptance criteria

Full Report

Find out more about the Full Report

Our secure and confidential safety reporting system gives professionals the opportunity to share their experiences to help others. If you would like to know more, please visit the reporting to CROSS-UK page.

The following report is intended to raise concern regarding the structural adequacy, quality control and inspection techniques used in the manufacture of proprietary bar systems, in particular the cast end terminations of the bar systems.

History

At a pre-manufacture meeting with the bar supplier, the reporter requested that 10% of the cast end terminations of the bar system be non-destructively tested by radiographic methods. A request was made by the reporter to the bar supplier for the submission of all quality assurance (QA) and inspection material related to the manufacture of the bar system. The Project Specification documents required this information to be provided before manufacture of the bar systems commenced.

It would include details of the proposed techniques for radiographic testing of the cast end terminations as well as copies of the radiographic film test results. The technique sheets were requested in order to verify that full inspection coverage of the casting was achieved in the radiographic test procedure, a key stage in ensuring that any defects would be accurately captured in the inspection test. The assessment criteria was to be based on ASTM E446, an industry standard specification for castings up to 2" (50mm) thick using radiographic techniques and reference radiographs.

ASTM E446 provides five levels for the severity of any discontinuities detected, 1 being the least severe, 5 being the most severe. Defects range in category, from gas porosity (Category A) to sand and slag inclusions, shrinkage and tearing (Categories B, C and D respectively). Initial radiographic test results indicated that shrinkage defects to level CB4 ASTM E446 were present in the 10% of castings tested. This level of defect, in the reporter's opinion, was outside the generally accepted industry standard for the particular application although no specific classification had been defined in the specification. Rather, the specification stated that the castings shall be manufactured to the level stated in the supplier's own Quality Assurance System (something which in the reporter's opinion is not uncommon for proprietary systems such as bar/stay systems).

Independent testing

The radiographic test report deemed these defect levels acceptable in accordance with this. However, the report did not provide specific details of the tests so the reporter raised concerns regarding the validity of the test in terms of inspection coverage of the castings. Because the issues were not resolved all the cast end terminations, which were on site by this time, were radiographed by an independent testing house.

It was alleged that the results showed that all tested castings included level 5 defects (the most severe defect level class in ASTM E446) and were thus outside of the supplier's internal acceptance criteria. Hence, without the persistence of the reporter, which included strong client support, for all castings to be tested a second time, the defective end terminations would have been installed in the permanent works. A new set of castings were procured by the supplier and were successfully installed on site.

Lessons learnt (from the reporter)

This incident has highlighted the importance of a set of well-defined acceptance criteria for cast components, which should be set out clearly in the contract documents (i.e. specification). Furthermore, an industry standard specification should be developed for proprietary bar type tension systems, which should include the end terminations. The reporter proposes developing such a standard in consultation with others and consideration will be given as to how this can be circulated through the industry.

This incident has highlighted the importance of a set of well-defined acceptance criteria for cast components, which should be set out clearly in the contract documents (i.e. specification)

In the interim period, designers and clients should exercise additional precautions when specifying such components. For a relatively small additional cost, the role of the designer in an inspection capacity during construction has been shown to be a key factor in ensuring that the as-constructed structure is in accordance with the specification. In this instance, the possible catastrophic consequences of failure of defective cast components installed on a structure, including the likely claims that would have resulted, far out-weighed any delays and costs associated with the additional inspection and manufacture of the castings.

Expert Panel Comments

Find out more about the Expert Panels

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.

Tension systems normally work at high stress and this is only justified if the system is ductile because the true elastic stress in any part of it might be quite different from that predicted by design. This can come about from a number of causes, for example cable structures are not entirely elastic; so introducing some element of unpredictability in stress calculations. This situation is tolerable provided the system as a whole can deform and permit load redistribution. The need for this capability is particularly relevant because any single hanger failure on any part of the system raises the possibility of a progressive collapse as hangers adjacent to a failed one might get severely overloaded.

Modes of failure

To assure cable ductility as a whole it is normally desirable that the end terminations should be stronger than the tension members themselves since that avoids excessive yield demand on short lengths and permits the favourable long displacement failure mode of hanger stretch itself. But in turn that demand leads to a need to have high confidence in end connector capacity. It is particularly important that the mode of failure of the connector is yield not fracture since a fracture is instantaneous and prevents there being warning of failure by prior excessive deformation. The global protection against failure is absorption of excess stress by ductility before fracture.  Anything that undermines the ability to stretch, or the ability to cope with local overstress, undermines this fundamental philosophy.

The risk of fracture

In some of the reports received, the hangers have been made from thick steel. With thick cast steel, a potential mode of failure is fracture which is extremely serious as a fracture can be instantaneous. The risk of fracture increases if the steel lacks toughness (Charpy) and lacks ductility. Achieving adequate toughness in thick steel can be problematical and heat treatment may be required. There is always tensile stress on hangers but within castings there may additionally be high residual stress as a consequence of differential cooling.

Stress concentrations and defects (which can give rise to acute stress concentrations) raise the risk further. Furthermore, fracture risk is higher in cold weather. Within connectors there are always going to be stress concentrations. For example, pin misalignment is a potential source. The stress concentrations should be controlled by elimination of casting defects as much as possible. It is essential to ensure the properties of ductility and toughness which permit plastic stress redistribution of stress concentration without facture.

Specification and testing

Overall, there must be a proper specification for the essential properties (not just concentrating on strength) and proper testing and checking systems to ensure that the components are fit for purpose. The engineering link between defect tolerance and service acceptability is extremely complex and expert advice may be needed. This report emphasises the need to ensure that what has been provided is what was required. In addition to radiography, ultrasonic examination may be used to good effect for defect discovery.
In summary the key points to are to:

  • Identify what are safety critical tension structures

  • Take great care with analysis and design

  • Specify the required properties of the end connections and use expert advice as necessary

  • Specify the tests that will be carried out on completed components and state the required acceptance criteria

Share your knowledge

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.