Study on the truss damage measurement based on Bayesian updating

Abstract

Bayesian updating can integrate new observations in real time. This online updating property makes the Bayesian approach particularly suitable for dynamic monitoring of structural damage. In this paper, the node displacements are obtained by the finite element software OpenSees, assuming that the bars are damaged and the stiffness is reduced. Rejection Sampling method is used to determine the stiffness drop interval. During the sampling process, different load sizes, load locations, errors and types are selected to realize the updating of the stiffness of the top chord bar, bottom chord bar, vertical web bar, and diagonal web bar. A Pratt-type truss is used as an example, combined with displacement data. The proposed stiffness damage measurement method is demonstrated by updating the stiffness of four types of bars separately. The results shows that damage in structures can be accurately identified by Bayesian updating methods. This method not only pinpoints the location of the damage, but also quantifies the extent of the damage.