Fatigue Reliability Assessment For Orthotropic Steel Bridge Decks Subjected To Stochastic Traffic Loads

With the rapid increase of population in urban areas and the growth of the economy,in rapidly developing countries like China,an increasing number of suspension and cable-stayed bridges are constructed and built in order to meet the requirements of significantly increased traffic loads on long-span bridges.Orthotropic decks are extensively used in steel box girders of long-span bridges due to their favorable material and mechanical properties.However,it is inevitable that bridge structures sustain the increasing dynamic loads,which will cause serious fatigue damage to orthotropic decks.In order to assess fatigue damage and fatigue life of orthotropic decks,this research study presents an assessment of fatigue reliability of long span steel bridges utilizing a numerical simulation approach.Due to the stochastic nature of traffic flow and the increasing trend of overloading,traditional models for fatigue due to truck loads are no longer suitable for the assessment of fatigue reliability of long span steel bridges.Thus,the main objective of this thesis is to assess the fatigue reliability for orthotropic decks subjected to stochastic traffic loads.The main work carried out and reported herein,and the achievements of this research are listed as follows.(1)The statistical analysis of the stochastic traffic flow based on weigh-in-motion(WIM)data is carried out.Utilizing the monitoring data collected by the WIM system,distribution parameters of vehicle category,lane,vehicle weight,as well as other important parameters are calculated and fitted.Due to the increase in overloading,the assumption of unimodal distribution for vehicle weight can no longer be justified.Therefore,as an attempt to develop more realistic distributions,a Gaussian mixture distribution based on the expectation maximum algorithm is used for fitting vehicle weight.Based on the statistics of traffic flow in one year,a linear function is determined to predict traffic flow in the future.After obtaining parameters of traffic flow,a simulation program for stochastic traffic flow is prepared to lay the foundation for the subsequent numerical analyses.(2)According to design parameters from a real physical bridge,a finite element model of an orthotropic deck is constructed.Based on the specific location where the main cracks may occur,welded details of the deck are classified and subsequently,the main research objective is determined according to the physical characteristics of the details and loading cases.A rainflow-counting algorithm is applied to extract stress cycles and S-N curve in Eurocode 3 is selected to calculate equivalent stress range after comparing different national fatigue specifications.The relationship between gross weight and equivalent stress range is fitted.The equivalent stress range due to stochastic traffic loads is obtained by using the simulation program of stochastic traffic flow.After obtaining the number of cycles and equivalent stress range,fatigue damage of orthotropic decks is calculated.The calculation results indicate that fatigue damage of the slow lane is 16 times as serious as the fast lane.(3)Monte Carlo algorithm is used to calculate fatigue reliability index for orthotropic decks.Due to the growth of traffic flow and loads,reliability indices are calculated considering these factors.Later,reliability index under predicted traffic flow is also calculated.The computed results indicate that reliability index of slow lane is much lower than the one of fast lane.Given the continued growth of traffic flow and loads,the slow lane is no longer safe enough and thus,it is necessary and essential to control traffic flow and loads.Target reliability index is determined and fatigue life is calculated under target reliability index.