Probabilistic fatigue load spectra for riveted railway bridges

This paper presents a probabilistic load model for fatigue assessment of old metallic railway bridges. The traffic basis of the model consists of trains obtained from past records and present-day trains obtained from codes. The traffic is divided into four distinct time periods each associated with particular characteristics in rail traffic. Traffic evolution scenarios for the future, in terms of increase in train axle loads or train frequencies, are also considered. With the help of a finite element model of a short-span, riveted railway bridge, train loading is converted into probabilistic fatigue load spectra via Monte Carlo train simulations. The spectra are found to be described well by Weibull distributions. As an example, the application of the proposed load model for the estimation of the remaining fatigue life of a riveted bridge connection is presented. Uncertainties in terms of loading, resistance and modelling are taken into account. The results show that introduction of higher train axle weights in the future is expected to reduce the remaining life of the connection considerably faster than the accommodation of increased freight through higher train frequencies. as random. Model uncertainty is represented by a ratio of actual to calculated stresses.