Specific distortion-induced fatigue failure at main girders of a railway bridge – efficiency of different reinforcements based on strain measurements

At the main girders of a three span continuous girder railway bridge, built in the 1960s, through-thickness cracks were detected in the web, at the location of the transverse stiffener to main girder connections. The transverse stiffeners are not welded to the girder flange, in order to increase their fatigue resistance. Instead, fitting strips between the stiffeners and the bottom flanges were used.

First of all, the paper presents the results of local strain measurements at the original detail, without any reinforcement, for individual train passages. These results confirm a distortion-induced fatigue failure, due to an unexpected high-frequency oscillation of the web plate. Based on that, three different types of reinforcement were developed and installed at the bridge. For each type of reinforcement further strain measurements were done during regular train service. Also, a weigh-in-motion-system was installed, to specify a representative passenger and freight train each. From the local strain measurements, during the passage of these two trains, the stress-spectra were evaluated for all three types of reinforcement. Furthermore, existing fatigue test data of a similar connection detail from the literature were reassessed in order to provide an accurate fatigue resistance for the damage accumulation. The comparison of the evaluated fatigue damage of these individual stress spectra indicates the efficiency of each type of reinforcement.

Finally, also representative reduction factors of the initial stress ranges without reinforcement, were developed for each type of reinforcement. This allows a more generalized assessment, regarding the efficiency of the different types of reinforcement, not just valid for individual trains but also for the whole train traffic.