Prediction and Evaluation of Bridge Extreme Value Effect Under Heavy Traffic Loads

The long-term heavy traffic loading conditions monitored by the Wuhan Smart Bridge Management System are drawn on to predict and evaluate the bridge extreme value effect in different regression periods under random heavy traffic loads, aiming to assess the current traffic loading levels of the city and provide guides for the structural design of urban bridges. First, the vehicles weighing more than 3. 8 t were extracted from the monitoring system of bridge clusters to serve as basic data, by which the proportions of different types of vehicles, share of different axle weights, vehicle weights and distances between vehicles were counted. At the same time, random heavy vehicle loads models of bridge clusters were built up, considering the normal and congested vehicle travelling conditions. A prestressed concrete continuous girder bridge with six spans of 20 m was used for a case study. The two random heavy vehicle loads models were applied to the bridge to calculate the related bending moments? deflections and shear forces? afterwards? the daily extreme value effects were extracted as samples to evaluate the extreme value distribution parameters? and the daily extreme value effect of the bridge agreed with the Weibull distribution. Furthermore? the bridge extreme value effects in different regression periods are predicted in accordance with the principle of extreme value extrapolation? and the load effect ratio and average load effect ratio were defined to evaluate the levels of traffic loads that would affect the safety of the bridge structure. The results demonstrate that the absolute values of extreme value effects are in linear relation with the increase of regression periods? growing faster at the early age and tending to be stable later. Under normal travelling conditions ? the vehicle load levels are far lower than the design vehicle load levels at the service stage? while under congested travelling conditions? the pier-top hogging moments are prone to exceed the suggested values in the code? which should be given special concern. © 2023 Wuhan Bridge Research Institute. All rights reserved.