Adaptive Multi-category Train Scheduling Validation Based on Fatigue Reliability of a Long-Span Suspension Bridge

Fatigue is an important issue in the durability of long-span cable-supported bridges. These structures are submitted repetitively to loads such as strong winds, varying temperatures, and increasing roadway or railway traffic. The trainload accounts for a relatively significant portion of the fatigue of critical members. Usually, trains consist of different numbers of carriages, and they cause different levels of stress. Consequently, the induced fatigue damage varies accordingly. This paper proposes an adaptive multi-category train schedule validation approach based on fatigue reliability using bridge monitoring data. The method involves three steps. Firstly, the typical train categories are identified based on long-term train weigh-in-motion (WIM) data, and their percentages are obtained. Secondly, the stress responses under each type of train are collected to establish a database, and fatigue damage is calculated to generate the probability density distribution. Thirdly, the number of different types of trains can be obtained for a typical train daily schedule, and the fatigue damage is calculated and checked with the design fatigue life. The case study of a suspension bridge is used to verify the approach, and conclusions are drawn from the perspective of the bridge management.