Life-Cycle Performance Monitoring And Analysis Of Long Span Rigid Frame-Continuous Beam Bridges

Construction monitoring, loading test and operation health monitoring are the indispensable components of large-scale bridge structure from construction to operation period, and they respectively provide performance, status information and scientific basis for bridge design, construction, use and maintenance strategies. They link up in order of time sequence and they has not only the similarity in monitoring content, monitoring method and technical route, but also the sharing content in monitoring facilities and monitoring information. Comprehensively considering the three stages, conducting the trinitarian design and then setting up the life-cycle monitoring system, is most reasonable whether from the view of economy or the integrity of monitoring system, the coupling of monitoring data and other aspects. This dissertation systematically summarizes the life-cycle monitoring systems of domestic and foreign large-scale bridges and analyzes the research state of their performance analysis and safety evaluation. Based on the project of Dongying Yellow River Bridge, this dissertation presents, designs and realizes the triune monitoring system integrating construction monitoring, loading test and operation health monitoring, deeply and systematacially analyzes the monitoring data and structural performance of all phases and soleves some concerned key scientific problems. The main research are as follows:First, this dissertation presents, plans and designs the life-cycle monitoring system of long span rigid frame-continuous beam bridges. Through the cost-investment analysis for the monitoring system scheme, purpose, function requirements and design criteria of life-cycle monitoring system for bridge are determined. Based on this, the monitoring stems and contents of life-cycle monitoring system of long span rigid frame-continuous beam bridges are introduced detailedly in terms of the three sub-stages of construction monitoring, loading test and operation health monitoring. The selection principles and criterions of sensors, especially the strain and temperature sensors, are illuminated and the problem of optimal sensor location is analyzed. Based on sensor sharing and information sharing, this research plans, analyzes, designs and realizes the life-cycle monitoring construction system.Second, aiming at the key control objectives about internal force and geometry, parameter sensitivity analysis of mechanical state during bridge construction is carried out. 15 state parameters are selected from the aspects of material characteristics of concrete, prestress and its losses, geometric property of structural members, construction temporary loads and environment condition, and the variability of the 15 parameters is analyzed. According to the general requirement of geometry being smooth and internal force being uniform and rational, 4 control objectives are selected in the states of construction and completion. By adjusting the value of state parameters and conducting a great amount of calculation, the control objective values are gained. Through the normalization of values of control objectives, the sensitivities of these state parameters to control objectives are analyzed and then the key state parameters subjected to different control objectives are determined. Based on the experiments of material performance of concrete, structural model of bridge structure is updated using the experimental values of key state parameters and validity of this model updating method is verified through static loading test and modal test.Third, grey system theory, Kalman filtering method and ANN method are applied to the geometric control of cantilever casting construction of girders and the geometric controlling effect of 3 kinds of theory is comparatively analyzed. The ratio and difference between the value of theoretical and measured deflection of each girder segment are regarded as the state variables to set up grey system models. Differentiating in the manner of data preprocessing and feedback compensation, this dissertation set up 8 kinds of grey system model to predict deflection and control geometry respectively. The model predictive values are contrasted to the measured values recorded in construction monitoring and then the controlling effect of 8 kinds of model is comparatively analyzed. This dissertation regards the deflection of girder segment as the state variables and filters the monitoring data with self-adaptive Kalman filtering method, and then carries out the geometric control by one-step prediction. According to the operational experience gained from construction site and analysis of the measured data, one ANN model is set up to predict the deviation of deflection. The controlling effect of difference-data preprocessing- metabolism GM(1,1) model, which is the best among all the 8 kinds of grey model, is compared with that of the Kalman filtering method and ANN model.Fourth, temperature and temperature effect of large-scale rigid frame-continuous beam bridges are analyzed comprehensively. During construction, typical sections are selected to carry out the temperature field test and the elevation of girder of T structure is measured synchronously. The changing character of temperature gradient and its 2 key parameters are generalized and the deflexion of"T"structure is analyzed in view of temperature field. The seasonal temperature and temperature gradient data recorded by the operation monitoring system are carried out the probability statistic analysis and their probability distribution functions are fitted and gained. The structural stress effect caused by the normal value of seasonal temperature and temperature gradient is compared with that of temperature gradient specified by code. The combined effect of seasonal temperature and temperature gradient is contrasted to that of the designing vehicle load. Taking temperature as input elements and mode frequency as output elements, this paper designs two kinds of BP neutral network model, which are respectively based on single section and multi section temperature distribution, to contrastively analyze their fitting and prediction effect about frequency. The meteorological temperature data are analyzed and the statistical parameters derived from the data of meteorological temperature and structure temperature are contrasted.Fifth, based on the multi-category monitoring information and Monte-Carlo method, the structural time-variant reliability for operation bridge in life-cycle period is quantitatively analyzed. About structure resistance,the strength of normal concrete specimens is statistically analyzed and conducted K-S hypothesis test, and then its probability distribution function is determined. Based on stress-time curves, this paper analyzes the structural fatigue damage, and then updates the available time-dependant model for concrete strength. About dead load effects, the relative deviation between calculated and measured strain is conducted statistical analysis. About live load, the number and weight of vehicles loaded on the structure are analyzed and carried out hypothesis test, and then the regularity of their probability distribution is determined. Integrated the statistical analysis result of temperature data and based on two load modes: monitoring load and design load, this paper computes the structural failure probability with the method of Monte-Carlo numerical simulation. The main failure fields and failure probability under two load modes are compared and analyzed. This paper also explains the failure cause and clarifies the time-variant characteristic of bridge structure.