Analysis Of Environmental Effects On The Performance Of Large Rigid-continuous Concrete Bridges

The performance of bridge structures are seriously deteriorated under thenegative effect of harsh environment, long-term overload operation, delayedmaintenance and reinforcement, leading to degraded safety and consequently majoraccidents. Considering the importance of bridge structures and multiple accidents, itis highly necessary to implement long-term real-time health monitoring on bridgesin operation state, carry out intensive study to grasp the effect of operatingenvironment on bridges, in order to make instant and precise assessment andprediction of structural condition.Choosing the Dongying Yellow River Bridge-a long-span prestressed concreterigid-continuous bridge as the research object, based on its long-term healthmonitoring data, this dissertation analyzed and studied the effect of operationalenvironment on the large-span rigid-continuous box-girder bridge. The maincontents are as followed:(1) Two finite element (FE) models of a large-span rigid-continuous box-girderbridge are developed using BEAM and SOLID elements respectively, then both theirmerits and drawbacks analyzed and compared. The BEAM model is valid forsimplified structural analysis; nevertheless, the SOLID one is more suitable forrefinement analysis, especially the structural local damage identification and modelupdating, etc. Based on massive monitoring data, radial basis function (RBF) neuralnetwork method and RBF response surface method are employed and developed,which transform the iterative computation from FE model to more effectiveapproximate model. The analysis and comparisons of the FE model updating resultsshow that the RBF response surface method is more compatible for rapid modelupdating of large-span bridge structures under real-time monitoring.(2) Based on the peak power spectrum method for modal frequenciesidentification under environmental excitation, the regularization power spectrummethod with low-pass filter eliminating high frequency noises, half powerband-width removing faulse modes, data averaging of multiple accelerations isdemonstrated to fast recognize the large amounts of online monotoring data andacquire real and accurate frequencies for different modes. Then, wavelet analysis isutilized to synthetically process the long-term health acceleration data to obtaincredible and stable frequency time histories. Thus, an improved peak powerspectrum is developed for rapid online identification using the continuousmonitoring data from actual bridge structures.(3) Amounts of fiber-Bragg grating (FBG) sensors for temperature and stress constitutes a large scale FBG sensor network by optimizing the sensor network ofthe structural health monitoring system of Dongying Bridge. Though analyzing thelong-term bridge monitoring data, the three dimensional temperature distributioncondition of the large rigid-continuous bridge are revealed: in longitudinal direction,the temperature differential is so small that can be neglected; for transverse, thedifferential is minor, mainly existing between the flanges in both sides and the topslab; in vertical, the differential is the largest, as different vertical areas suffer thesolar raidation with various time period and delay.(4) Based on the three-dimensional finite element model of a rigid-continousbox girder bridge, a simulation method is proposed to analyze the effect of ambienttemperature and its variety to the static and dynamic charateristics of this type ofbrdige. First, the theoretical method of thermal analysis for the bridge structures isproposed based the simulation study of a simply supported prestressed concretebeam, including material properties, stress and dynamic characteristics. Second, thevarious temperature distribution conditions on Dongying Bridge is discussed basedon the FE analysis using its SOLID model. Finally, thermal analysis on the SOLIDFE model and its static and dynamic analysis are conducted considering the solarradiation effect. Results from the analysis for the effect of the operationenvironment temperature on Dongying Bridge shows that ambient temperature loadhas significant effect to the static and dynamic characteristics. The temperaturedifferential inside the bridge structure and the consequent temperature stress ismainly caused by daily solar radiation.(5) The variation regulation of structure strain under temperature effect isobtained from the analysis of long-term monitoring data of Dongying Bridge. TheFBG strain sensor with temperature compensation can accurately monitor thesituation of structure strain changing with temperature and vehicle load in dailyperiod. Strucrural strain is heavily affected by environment temperature, and itschange caused by seasonal temperature differentials has severalfold differences.Comparatively strong linear correlation between strain and temperature is observedby analyzing of the monitoring data, yet different strain measuring point has its owncorrelation coefficient.(6) Vehicle loads in Dongying district are tatistically analysed according to thedata toll station of of Dongying Bridge monitoring system. The probability modeland maximum distribution of vehicle loads in this district are defined according tothe study and comparison of different probability models. Then, the load shapefunction method for moving vehicle load identification is proposed, which ultilizesthe shape function to approximate the moving vehicle load. This method has nolimitation of structure types. For complex large-span bridge structures, the movingforce can be accurately identified through measured acceleration if the FE model is precise enough. Its validity was certified by the numerical simulation of the FEmodel of Dongying Bridge.