A statistical index based damage identification method of a bridge using dynamic displacement under moving vehicle

Potential damage identification is essential to avoid the sudden and premature failure of aging bridges, which have attracted ample attention over the decades. This paper proposes an output only data driven technique for damage identification which will not require any physical model of the structure nor any known input excitation. Statistical moments have been considered as the key feature for the proposed method of damage identification which has been derived from the measured vehicle-induced dynamic displacement responses of a pre-stressed concrete I-girder bridge. The displacement responses data of a bridge is influenced by many factors, including Vehicle-Bridge Interaction (VBI), vehicle speed and road roughness. The interaction between vehicles and bridges includes dynamic models for bridge structure subsystem and vehicle subsystem, interaction constraints and road roughness modelling. Firstly, the coupled formulation is derived based on the interaction constraints. Then, Newmark's beta method is utilized for solving the coupled VBI problem to extract the vehicle-induced displacement responses. Later, the statistical moment is calculated from the Power Spectral Density (PSD) of the displacement response retrieved from sensors positioned on the span of the bridge for both damaged and undamaged conditions. A damage index named Curvature of Statistical Moment Difference (CSMD) is calculated to identify the damages. Artificial damage is incorporated through the reduction of structural stiffness of the bridge element. The results of the study demonstrate the method's capability to identify and locate damage at any position of the bridge span and assess the relative severity of the damage. The effect of varying pavement roughness conditions is also considered in this study.

Automated summary

This paper proposes an output-only data driven technique for damage identification in aging bridges. By utilizing statistical moments derived from measured vehicle-induced displacement responses, the method is able to identify and locate damages without the need for a physical model or known input excitation. The results demonstrate the method's capability to assess the relative severity of the damage and consider varying pavement roughness conditions.