Safety Evaluation Of Large-span High-speed Railway Bridges Based On Vertical Deflection Monitoring

With the rapid development of high-speed railways in our country,the operating safety of high-speed railway bridges has been gradually concerned by domestic and foreign scholars.The dynamic response of high-speed railway bridges is far greater than normal railway bridges on account of the high-speed train load,which puts forward higher requirements for stationarity,stability and security of high-speed railway bridges.Girder deflection is a comprehensive response of bridge structures under different factors.It is one of the most important parameters for the evaluation of bridge applicability and safety since it contains characteristic information of structural damages.Therefore,it is necessary to carry out long-term monitoring and safety assessment for the girder deflections of a special bridge type like Nanjing Dashengguan Bridge.Based on the girder vertical deflections of Nanjing Dashengguan Bridge,this paper investigates the characteristics of the girder deflections under different factors and proposes a new method for security early warning of vertical stiffness degradation of the girder.The main contents and achievements in this paper are as follows.1.Based on the long-term monitoring data of girder deflections,the characteristics of deflections under different factors are analyzed.The results indicate that:(1)Deflection changes are basically caused by environment temperature and train load.There is a strong correlation between girder deflections and long-term temperature differences as well as short-term temperature differences.(2)The separation of different effects on girder deflections can be made through a multi-scale wavelet decomposition method.(3)Under the effect of long-term temperature differences,a strong linear correlation is found between daily average thermal-induced deflections and girder temperatures.With the increase of temperature,all the measuring points on the girder show the trend of upward.Furthermore,the thermal-induced deflections at different points along the girder have obvious spatial linear correlation with their position.(4)The characteristics of girder deflections in different train load conditions are different.2.Based on the deflection data after separation of load effects,the influencing factors of peak values and the statistical characteristics of probability density of the train-induced deflections in different train load conditions are analyzed.The results indicate that:(1)Strong correlations exist between peak values of the train-induced deflections and the girder temperatures.When the temperature increases,the peak value of the train-induced deflections whose measuring points are on the main span decreases while the the peak value of the train-induced deflections whose measuring points are on the side span increases.(2)The peak values of the train-induced deflections and the speeds of the trains have no obvious correlation.(3)The amounts of train carriages and driving directions have influences on the peak values of the train-induced deflections.However,the factor of upstream and downstream makes no difference.(4)The discrete degree of the data is supreme in the ninth train load condition.(5)The statistical characteristics of probability density of peak values of the train-induced down and up deflections can be fitted by different probability density functions.For instance,the probability density of peak values of the train-induced deflections is fitted well by t Location-Scale distribution or Normal distribution when the train carriage number is eight.Extreme Value distribution and Generalized Extreme Value distribution are needed if the probability density is asymmetric.3.Based on the monitoring deflection data of the measuring points on the middle of the main span close to Beijing,time series analysis is performed.On this basis,a new method combined with mean control chart is proposed for security early warning of vertical stiffness degradation of the girder.The results indicate that:(1)Deflection data of Dashengguan Bridge is a set of time series which can be studied by time series analysis for the prediction of its development laws.(2)Using ARMA-GARCH model can achieve high accuracy due to the correction of conditional heteroskedasticity in actual measuring deflection data.(3)The short-term prediction accuracy and fitting degree of the model is high enough while it decreases as time goes on.It is viable using the time series model to forecast the deflection changes.(4)The mean control chart method can be used to identify the abnormal change of the deflection data.Whether the vertical stiffness of the girder appears degradation can be proactively determined by combing the time series analysis and the mean control chart method,which means security early warning system of the girder vertical stiffness,can be established based on the monitoring deflection data.