The Research On Time-dependent Reliability Of Concrete Cable-stayed Bridges During Construction

The life cycle of concrete bridges consists of three periods, i.e., construction period, service life and degradation process. The failure probability of cable-stayed bridges during construction is higher than that during service life. However, the existing codes provide no uniform safety requirement for bridges during construction and nearly all of the existing literature is not particularly surprising since very little is known about the structural performance of a bridge during its construction based on structural reliability theory.This dissertation studies the probabilistic model of structure member resistance for concrete cable-stayed bridges during construction, the probabilistic analysis method of load during construction, time-dependent reliability analysis, and evaluation of system probability of failure based on time-dependent reliability. The main work is as follows.1. On the basis of in-site experimental results for the concrete cable-stayed bridges, the early mechanical properties of high-strength concrete are studied. Probability distributions of the early strength and elastic modulus of concrete are analyzed. Time-dependent models of mean and variation coefficient of early strength and elastic modulus are developed respectively. Also the relationship between the early strength and elastic modulus of concrete is analyzed.2. Considered of the randomness and the time-dependence of resistance during construction, a unbalanced random process is introduced to describe member' resistance. The mean and variance function as well as self-correlation coefficient of resistance for flexural member of concrete bridge during construction are determined by taking the independent increment process as the probabilistic model of resistance. And the two methods of calculating the standard auto-correlation function are introduced: the method of discreteness data and the method of continuum function assuming resistance as independent increment random process. Additionally, the influence of parameters on structural resistance is studied.3. Based on the in-site investigation and statistical analysis of live loads during construction, the changing law of construction live load of main beam with the changing cantilever length is studied, and the probabilistic model of live loads on cable-stayed bridge is established. It is found that the mean value of live loadssubmits to exponential distribution and standard variance of live loads submits to logarithm distribution.4. The two parameters, density and elasticity module, are identified used ANN. The two new parameters which represent load affects, i.e., equivalent density and equivalent elasticity module, are suggested to evaluate the workmanship during construction. And the unitary quality control method and the duality quality control method are used to evaluate the stability of workmanship.5. An modified virtual intervening variable method is introduced to deal with the reliability analysis of cable-stayed bridges, the performance functions of which are implicit. The performance function approximation is implemented in the standard normal space of random variables. The efficiency and accuracy of the proposed algorithm are demonstrated by several examples.6. Combined with the Construction Process Analysis Program for Bridges, an efficient algorithm is proposed to calculate the time-variant reliability index of cable-stayed bridge during construction. Based on principle of first order and second moment method, the formulas are introduced. The parameters' sensitivity is studied for time-variant reliability of cable-stayed bridges during construction. The results show that important emphases can be paid to the key parameters in advance which results in better construction quality.7. A modified/?—Unzipping Method is suggested to identify major failure paths. A practical approach for system failure probability evaluation is proposed by considering the effect of correlation between modes of failure. A cable-stayed bridge consisting of double-tower pylons is selected to show the efficiency and accuracy of the proposed algorithm.