| Nowadays,bridge health monitoring is a research focus of structural safety state assessment.Two steps are generally required when the mass health monitoring data is applied: At first,the mass monitoring data preprocessing step,the monitoring raw data in the health monitoring system inevitably contain a lot of distorted data.Hence,it is necessary to process,analyze and evaluate these data systematically and quickly;And finally,the engineering application of preprocessing health monitoring data,which is mostly explicit data,and it is necessary to deeply dig into the multi-dimensional recessiveness of these data in order to use for structural safety performance evaluation and reliability prediction,which is one of the current research focuses in the field of bridge health monitoring.Therefore,basing on the theory of structural reliability and long-term monitoring of long-span suspension bridge health monitoring data,this paper conducts a systematic study on the prediction of the time-varying reliability of the steel truss main girder.The main results include:(1)Based on the mass data of the health monitoring system of long-span suspension bridge,traffic load effect and temperature action in the strain data of steel truss girder are separated by using wavelet transform,and the results shows that the wavelet transform works well.The data of temperature and humidity through statistical analysis,found that the location of bridge has high relative humidity and large temperature difference between day and night,which can easily cause corrosion to the steel truss beam member;the resistance attenuation of steel members is approximate linear decay under the optimum curing condition,which still has high resistance when serviced for one hundred years,and the resistance attenuation of steel members is approximate exponential decay under the wrong curing condition,so the component in service will be in a state of low resistance at the middle and later periods;(2)Steel truss beam in the design reference period,the statistical analysis and hypothesis test for the separated data,found that the temperature action obeyed normal probability distribution,the extreme traffic load effect obeyed extreme value type I probability distribution,and the equivalent stress amplitude of fatigue loading effect conformed to the lognormal probability distribution as same as the cyclic number of fatigue loading effect.Obtained the extreme strength and fatigue reliability indexes of the component in the design reference period by using Monte-Carlo method,which both under the target reliability indexes,and the comparative analysis shows that the fatigue failure probability of the component is larger than the strength failure probability in the design reference period.When the structure is in service to the 111 th year,the extreme strength failure probability of the component is the large one.Considering the influence of traffic growth on the fatigue reliability of the steel truss girder component,the results shows that traffic growth can significantly reduce the fatigue reliability index;(3)The parameters sensitivity analysis on component reliability shows that the main influence factors of reliability assessment are the strength of the resistance attenuation of the component,the main influence factors of fatigue reliability assessment are the equivalent stress amplitude of fatigue load effect.Finally considering the corrosion and fatigue damage coupling effect on the resistance attenuation of the component,the material performance degradation of the corrosive component is increasingly significant due to the fatigue damage of the fatigue loading effect as the structure service time growing.Therefore,the resistance will attenuate rapidly in the mid-late service period,and the reliable index of the component will be lower than the target reliability index when it services for ninety years,the component failure probability will be increased remarkably. |
PDF Full Text Request