Reliability Assessment and Reliability-Based Design Optimization of Large Scale Box-Girder Structural System

Reliability analysis and reliability-based design optimization of large scale box-girder structural system is an interesting topic in the field of structural design. Structural resistance, loading, plating thickness and cross-sectional area of beam are considered as stochastic variables. Safety margin equations of both beam and plate element are constructed. Since the safety margin equations are implicit function of random variables, the stochastic finite element method (SFEM) is adequate for sensitivity analysis. The advanced first order second moment (AFOSM) method is used to calculate the safety indices of structural components. In addition to this, branch and bound method is used to identify the dominant failure paths. Probabilistic Network Evaluation Technique (PNET) method is used to assess failure probability of structural system. The optimization problem of structure is formulated as a nonlinear programming problem that aims at minimizing structural weight with constraints on reliability and range constraints of design variables. An optimum vector algorithm is applied to solve the optimization problem. One illustrative example is carried out to elucidate present process of reliability assessment and reliability-based design optimization.