Structural damage detection of cable-stayed bridges using changes in cable forces and model updating

A new static-based method for damage detection of cable-stayed bridges using the changes in cable forces is presented in this paper. In recognizing the fact that damage incurring in bridge girders will cause a redistribution of forces in stay cables, the proposed method uses the measured changes in cable forces to detect the damage in bridge girders. In this method, the damage identification is formulated as an optimization problem in which the cable force error between measurement results and analytical model predictions is minimized. The sensitivities of cable forces with respect to the structural parameters of bridge girders for the linear and geometrically nonlinear bridge structures are obtained by using the direct differentiation method and the finite difference method, respectively, and are used to solve the optimization problem. Regularization technique is used to alleviate ill-conditioning in the solution course of the optimization problem. The validity of the method is illustrated by numerical studies of damage detection of the cable-stayed Sutong Bridge with a main span of 1,088 m. The effect of measurement noises and the number of measured cable forces on the accuracy of damage quantification is investigated. The proposed method correctly identifies the damage locations and damage magnitudes in bridge girders when using noise-free cable forces. It is also shown that structural damage can be quantified with acceptable accuracy in the case of low-level measurement noises and the damaged members being well spaced, and the identification accuracy is greatly improved when the forces in the cables connected to the damaged members are available for damage quantification.