Investigation On Theory For Quantitative Durability Design Of Concrete Structures Under Chloride Environment

Affected by chloride environment, concrete structures are often damaged on reinforcement corrosion far less than the design life. Because of the randomness of factors such as structural geometry, material properties and environmental conditions, the complexity of durability analysis and design for concrete structures are exacerbated. Therefore, durability analysis methods are established for time-dependent diffusion coefficient and constant diffusion coefficient by Fick’s second law and lumped concentration matrix. And then the Krylov stochastic diffusion finite element method (referred KSFEM) for the large variation system is established by introducing the Krylov stochastic finite element method of homogeneous basis vector. The durability governing domains of concrete structures are proposed by the compensation theory for chloride diffusion field. A three-parameter model of durability design is determined quantitatively by calculational model of governing domain. According to the reliability theory, the partial factors of random variables are determined. Thus the design values of durability parameters are gotten by probabilistic limit state expression for the three-parameter model. Previous methods are abe to improve the corrosion resistance and service life of concrete structures. It is important for engineering significance and academic value. This thesis includes the following conclusions:(1) Acoording to Fick’s second law and lumped concentration matrix, the durability analysis methods are established under constant diffusion coefficient and time-dependent diffusion coefficient. Which overcome the unreasonable phenomenon by coordination concentration matrix, such as oscillation or negative value. Numerical examples show that the methods are with high accuracy, computational efficiency and algorithm stability. The results of durability analysis for concrete structures will be conservative when the time-dependent of chloride diffusion coefficient is ignored.(2) In stochastic systems with large variation, the Krylov stochastic finite element method of homogeneous basis vectors is introduced. Base on the zero-mean random diffusion equation, the basis vectors of stochastic diffusion are derived. And then the KSFEM of durability probabilistic analysis is established by Galerkin weighted residual method, the diagonal elements of stiffness equation and the dominant expressions of mean and covariance are derived. The results show that, the KSFEM is suitable for durability probabilistic analysis of concrete structures with high accuracy. It avoids problems of other methods, such as the expansion problems of spectral stochastic finite element method, small variation problems of perturbation stochastic finite element method and low efficiency of reduction stochastic finite element method.(3) According to compensation theory of chloride diffusion field, the governing domains of durability analysis are established, and the three-parameter models of durability quantitative design are proposed. Engineering applications show, the durability governing domains with higher accuracy and efficiency, can analyse the process and distribution of chloride concentration in concrete structures quickly and accurately. The durability design parameters could be quantitatively determined by the durability quantitative design approach. It effectively overcomes the disadvantages for lack of quantitative design model in existing durability design methods.(4) Under the reliability and sensitivity analysis of concrete structures, the partial factors of random variables are determined by target reliability index. Combined with the three-parameter model of durability quantitative design, the calculation model and design method of durability quantitative probabilistic design are established. Results show that, the service life of concrete structures will be overestimated with randomness of parameters ignored. The durability quantitative probabilistic design method can reflect the influence of randomness for structural geometry, environmental effect, material parameters to the durability and service life of concrete structures. It has a better advantage than traditional methods.