Study On Time-varying Reliability Of Existing Buildings Considering Material Performance Degradation And Random Earthquake Action

Engineering structures are not only affected by different environmental effects during service,but also may suffer from earthquakes.Considering the time variation of material mechanical properties under environmental actions,as well as the randomness of earthquake action,is a basic problem to evaluate the anti-seismic reliability of existing structures during the whole life cycle.This paper takes existing high-rise buildings as the engineering object,and focuses on seismic action.Based on theoretical derivation,numerical simulation and measured data,the dimension reduction modeling method of random ground motion is studied.Then a unified dimension-reduction theory of random dynamic action modeling based on frequency-domain and time-domain analysis is developed.Meanwhile,considering the time variation and randomness of the mechanical properties of steel and concrete materials,a probabilistic model of reinforced concrete material degradation in general atmospheric environment is established.Furthermore,integrating the probability density evolution method,the fine analysis of the nonlinear random seismic response and seismic reliability of existing high-rise buildings are conducted,which provides theoretical support for the dynamic evaluation of seismic performance of complex engineering structures in the whole life cycle.The main research contents and results are listed as below:(1)The dimension-reduction methods for simulating non-stationary stochastic processes and spatio-temporal stochastic field are proposed.Specifically,the spectral representation and filtered-white-noise representation of non-stationary random process and spatio-temporal stochastic field are systematically derived,and the differences and connections between the two methods are clarified.On the basis of the original spectral representation,the simulation formula of the traditional Monte Carlo method is derived,which reveals the relationship between the traditional and the original spectral representation,and fully expounds the necessity of dimension reduction of orthogonal random variables.Further,the unified theoretical basis of dimension-reduction simulation based on spectral representation and filtered-white-noise representation is established.Through the construction of random function of orthogonal random variable set,the efficient dimension-reduction representations of non-stationary random process and spatio-temporal stochastic field are realized,which effectively breaks through the challenge of high-dimensional random variables faced by traditional Monte Carlo method.(2)The verification and parameter modeling of non-stationary ground motion dimensionality reduction method are realized.Specifically,using the proposed dimension reduction method,combined with the classical second-order statistical model of non-stationary ground motion and empirical parameters,the dimension reduction simulations of one-dimensional and three-dimensional ground motion process and ground motion continuous field are realized.By comparing the accuracy and efficiency with the traditional Monte Carlo method,the accuracy and superiority of the dimension reduction method are verified.Meanwhile,the parameter identification method of the dimension reduction model of non-stationary ground motion is proposed.Based on the measured strong motion records,the probability distributions of the basic parameters of the dimension reduction model of the non-stationary ground motion are identified,and the dimension reduction model of non-stationary ground motion with random parameters is established.The research shows that the representative time history set of non-stationary ground motion generated by the ground motion model in this paper has good consistency with the measured strong motion records,and can fully reflect the natural variability of ground motion.(3)The probabilistic model of mechanical properties degradation of reinforced concrete material in general atmospheric environment is established.The mechanism of concrete carbonation and steel corrosion is expounded,and the prediction models of concrete strength and carbonation depth,steel corrosion and yield strength suitable for existing building structures in general atmospheric environment are given.On this basis,the existing constitutive models of concrete,steel bar and steel bar slip are modified,and the time-varying constitutive model of reinforced concrete materials in general atmospheric environment is obtained,which lays a foundation for the seismic reliability evaluation of existing reinforced concrete structures in general atmospheric environment.(4)The application of probability density evolution theory in the seismic time-varying reliability evaluation of existing structures is expanded,and thus the seismic time-varying reliability of the structure during the whole life cycle can be obtained by using less deterministic analysis.At the same time,taking the randomness and time-varying of material mechanical properties degradation and the randomness and natural variability of earthquake action into account,combined with the probability density evolution and change of probability measure method,the nonlinear random seismic response and seismic reliability of new and existing high-rise building structures are analyzed in detail.The time-varying law of seismic response evolution and seismic reliability of existing structures during the whole life cycle under general atmospheric environment is revealed.