Damage identification of long-span bridges based on the correlation of monitored global dynamic responses in high dimensional space

Abstract

Structural damage identification and condition assessment is the most direct goal in structural health monitoring (SHM). Vibration-based methods (VBM) using the global dynamic responses (displacement or acceleration), including modal-based methods and statistical methods, have been comprehensively investigated in the last few decades in the SHM community. These two types of methods face challenges in practice due to the indicators being sensitive to the environment and insensitive to local damage. It is curious whether new features of data in high-dimension space can be found and the new features can be used as damage indicators. This study proposes a method based on the correlation of the global dynamic responses in high dimensional space for damage identification. It is proved that the ratio of mean square (RoMS) between any two responses (two-dimensional space) is dependent on the modal parameters of the structure only and changes only with the damage of the structures. Therefore, it can be used as a damage indicator. The proposed method is validated through a finite element numerical analysis on a cable-stayed bridge and the monitored acceleration of a real-world suspension bridge, and is further compared with the modal analysis methods. The results show that the proposed method can accurately identify the damage and is sensitive to local damage compared to modal parameters.