Sructural Analysis Of Thin-Walled Curved Box Girders Based On Stochastic Finite Element Method

This dissertation established numerical methods in the stochastic structural analysis of thin-walled curved box girders on the basis of random fields theory, stochastic variational principle, Monte Carlo method, etc. The numerical methods and the Monte Carlo method for thin-walled curved box girders are studied systematically. The one-dimensional stochastic finite element method and Monte Carlo method for curved box-girders is proposed. The main contributions of the thesis are as follows:1. On analyzing kinds of theories for thin-walled curved box girders synthetically,the finite element method with considering the shear lag,the bending and torsion for thin-walled curved box girders is presented and the corresponding structure model is established. The one dimensional discretized finite element method is combined with the theory of thin-walled curved box girders.The formula for calculating and the detailed approach of the method is established.2. Based on the local average theory of random fields, the computation formulas for discretized random fields is established. Combining thestochastic variational principle with the thin-walled curved box girders theory, the one-dimensional stochastic finite element method for thin-walled curved box girders is proposed.3. Sampling methods for common distributing random numbers and the samples of random fields are proposed. Combining the Monte-Carlo simulation technique with the one-dimensional discretization finite element method, the Monte Carlo-finite element method for stochastic analysis of thin-walled curved box girders are proposed in this thesis.