This website uses non-intrusive cookies to improve your user experience. You can visit our cookie privacy page for more information.

RR643 - Assessing and modelling the uncertainty in fatigue crack growth in structural steels

The prediction of the fatigue life of steel structures can be carried out in a number of different ways. A common method at the design stage is to use the so-called S-N approach using design data from a standard such as BS 7910. However, for existing structures containing defects of a known or postulated size, a fatigue life assessment is generally carried out using fracture mechanics. In keeping with normal engineering practice, it is usual to calculate conservative (safe) estimates of fatigue life, although occasionally best estimates may also be of interest. For more advanced structural assessments, reliability-based methods can be used to calculate the remaining life corresponding to a number of different probabilities of failure (eg 10-4, 10-6). The motivation for the research reported here was the need to improve the current methods of reliability assessment for structures, and in particular steel offshore structures approaching the end of their design lives. As part of this research, work was carried out to investigate the variability in the fatigue crack growth of steels and the way in which the corresponding uncertainties could best be incorporated into the assessment process. This included the fatigue testing of specimens of BS 4360: 1990 Grade 50DD steel with the explicit aim of studying the variability in crack growth under different conditions. The results of these tests are presented in this research report. The relatively large uncertainties associated with fatigue crack growth behaviour, even within relatively homogenous sets of specimens, means that the variance in the predicted fatigue life is relatively large. It has been shown, however, that the use of fatigue crack growth rate data from relatively early in the life of a particular structure can significantly reduce this uncertainty and improve the reliability predictions through the process of reliability updating.

This report and the work it describes were funded by the Health and Safety Executive (HSE). Its contents, including any opinions and/or conclusions expressed, are those of the authors alone and do not necessarily reflect HSE policy.

Assistance in the use of Adobe Acrobat PDF files is available on our FAQs page.

Updated 2010-03-19