Realistic fatigue assessment on historic riveted retrofitted railway bridges in Chile

Fatigue damage is a major cause of failure in riveted railway bridges, particularly at stringer-to-floor-beam connections, which studies have identified as the most susceptible to fatigue. However, limited historical traffic data and structural details mean most assessments rely on the bridge's current state and standard loading models, which do not accurately reflect its loading history, resulting in unrealistic fatigue damage predictions. This research presents a novel approach to assess the remaining fatigue life of a riveted bridge archetype defined from a taxonomy of the Chilean railway bridge portfolio. The assessment examines interventions implemented on Chilean railway bridges by analysing the bridge in different configurations that sequentially represent the history of structural modifications in Chile's steel bridges. A realistic train loading spectrum is defined by considering historic and current data on freight trains and traffic. Finite element numerical models are generated for the bridge in all its configurations, and the fatigue damage is evaluated through S-N curves from the Eurocode. The fatigue results from the loading model are compared with standard-prescribed models, showing the latter often overestimate fatigue damage and predict shorter lifespan. Moreover, the results indicate that retrofitting can substantially alter the structural response at the stringer-to-floor-beam connections. A reliability analysis is also employed to compare the remaining fatigue life with that obtained by applying the Miner-Palmgren rule. The analysis shows that the Miner-Palmgren rule yields conservative results in comparison to a wide range of reliability indexes proposed in the current literature. The proposed methodology is also applied to a real case study, a 130-year-old riveted through-truss bridge on the Chilean railway network. This application validates the approach developed for the archetype bridge and confirms the need to consider all bridge interventions to generate a realistic fatigue damage assessment, which can be significantly underestimated otherwise.