Traffic-induced fatigue damage evaluation of long-span suspension bridge integrating 27-year monitoring data and multi-scale finite element analysis

Abstract Fatigue damage is a main concern to the safety of long-span suspension bridges, especially when they carry both railway and highway loads. Most fatigue assessment of bridges is based on short-term measurement data or numerical analysis only. This paper takes advantage of 27-year long-term field monitoring data of the Tsing Ma Bridge and evaluates the bridge’s fatigue condition with the aid of multi-scale finite element analysis (FEA) at the component and weld levels. First, the stresses of all components are calculated using a global multi-scale model, and their fatigue damage is estimated following the Rainflow algorithm and Miner’s law. Second, at the weld level, a refined multi-scale finite element model is developed with a focus on the deck-rib welds of the upper orthotropic deck. Five sub-models of the trough-to-deck joint are established with consideration of varying wheel locations. A sub-modeling technique is employed to calculate the effective notch stress and its fatigue damage. At both component and weld levels, railway and highway loads are applied, and the load combination factors are estimated. The results show that the railway load plays the dominant role in the fatigue damage of the main components, while the highway load governs the fatigue damage of the deck-rib welds. The railway beam is the most fatigue-critical component with a fatigue life of 536 years, and the most vulnerable weld has a fatigue life of 30 years only. In the study, the monitoring data provides input for the long-term railway and highway loads on the bridge deck and validates the stress results calculated from the FEA. The super long-term field monitoring system provides realistic load and response data, yielding more accurate and reliable fatigue damage assessment of long-span bridges.