Several Critical Problems Studies On Healthy Monitoring Method For Long-span Continuous Rigid-frame Bridge

As a mature kind of bridge, continuous rigid-frame bridge has been extensively usedin our country. However, due to the material properties, construction techniques and someother factors, during the period of operation, the PC continuous rigid frame bridge isvery prone to diseases, such as the main beam deflection, box girder cracking, reductionof effective prestress and so on, which is a threaten to the safe operation of the bridge. Toensure the safety of bridge structure operation, more and more large bridge healthmonitoring systems are established, which can timely and effectively find out diseases ofthe bridge structure to make a quick treatment.In this paper, take a three-span PC continuous rigid frame bridge as an example, afurther research is carried out about some key problems of the long-span PC continuousrigid frame bridge health monitoring, such as the construction of the health monitoringsystem, parameter identification, the finite element model updating, and the influence oftemperature on the dynamic parameters of the bridge. The main research contents andresults are as follows:(1) The design of the health monitoring scheme, the allocation of the censers and thehealth monitoring system of the PC continuous rigid frame bridge. Based on the designfiles and the actual characteristics of the Baihe Bridge, the health monitoring scheme hasbeen set up, and the allocation of the censors has been done following relativeoptimization requirements, and the system has been integrated to complete the monitoringin time;(2) The method of the modal parameter recognition of PC continuous rigid framebridge based on the environment excitation. With the established health monitoringsystem, a set of actual data has been selected, then use frequency domain decompositionmethod and stochastic subspace method combine with stabilization diagram respectivelyto identify the modal parameter of Baihe bridge with the data selected.(3) FE model updating of a continuous rigid frame bridge based on parametersensibility analysis. Based on bridge design documents, the initial finite element model ofthe full bridge was established with solid element. Comparing finite element calculationresults with the measured results, it can be observed that there is a considerable errorbetween frequencies of each order, and the maximum error is more than10%. Throughthe parameter sensitivity analysis, select high-sensitivity parameters to be modified, andtake the modal parameters as the target, update the initial FE model with the integrateduse of zero order method and the first order method and then the baseline finite elementmodel is get. The results show that updating FE model with high-sensitivity parameterscan greatly accelerate the convergence speed; the frequency calculated from the updatedfinite element model is very close to the frequency measured under ambient excitation. Errors between calculated values of the first eight order natural frequency of vibrationand measured values are within3%. Meanwhile, during the process of the optimalcorrection, structural variables are always changing within a preset range, making themodel results do not deviate from its actual correction physical meaning.(4) Research on temperature effects of the dynamic characteristics of the bridge.Based on long-term monitoring of data. According to long-term monitoring data gettingfrom the health monitoring system, statistical analysis is performed on the dynamiccharacteristics of structure, which is identified from the data at different times. The resultsshow that the temperature has great influence to the natural vibration frequency, and forcontinuous rigid frame bridge, modal frequency has a significant negative correlationwith ambient temperature, and the influence on the natural frequency of lower-ordermodes is even more significant than that of high order modes.