Fatigue assessment of riveted bridges

Sammanfattning: The majority of the railways used today were built in the beginning of the 20th century. Most of the bridges constructed at that time are still in service. This was achieved by designing the bridges with an over capacity, this extra reserve in the design of the bridges was done since the axle load of trains and locomotives were changed during this period. To ensure that the bridge stock could manage future axle load alterations a buffer was assigned to their resistance. The situation with an increasing number of old bridges still in service that are reaching their design service life is similar in all of Europe, and because of their quantity it is impossible to replace all bridges at the same time. To be able to make old bridges stay in service longer enhancement of the existing assessment methods has to be made. To do this the procedure of an assessment must be known to be able to recognize where improvements can be made and areas that are critical in a bridge must be identified. This thesis has focused on the material properties of steel bridges constructed before the 1940's and how to estimate the remaining fatigue life of riveted bridges. By gathering information from bridges where the material properties have been determined a data base was created. From the information in the data base a better prediction concerning the properties to expect in steel bridges constructed before the 1940's is obtained. By using information from the data base a more accurate calculation of the resistance can be achieved which opens for the possibilities for higher loads to be allowed. Concerning the fatigue life of riveted bridges this thesis has focused on two areas. The first area is the girders of the secondary structure, stringers and cross girders, transferring loads from trains or cars to the main girders of the bridge. These girders are often in focus in assessments, due to their length and position in the bridge that makes them more exposed to fatigue damage than other parts. The work has focused on determining which detail category that should be used in calculations of the fatigue life for riveted structures. A survey was performed containing information of large scale fatigue tests performed on riveted bridge girders taken out from service and tests on small scale specimens to investigate influencing factors of the fatigue life. The second area concerning fatigue and how to estimate remaining time in service is the connections between the stringers and cross girders. The connections between these girders are often exposed to unintended load exposure causing cracking in these joints. A fracture mechanic approach was used to estimate the degradation of stiffness in these connections and the propagation rates of cracks. Also investigations were carried out concerning the best way of modelling these connections by comparing the results from the FEanalyses to field measurements.