Detecting Changes in the Behavior of the Indian River Inlet Bridge Through Cross-Correlation Analysis of Truck-Induced Strains

Abstract

The Indian River Inlet Bridge (IRIB) is a cable-stayed bridge located in southern Delaware. A comprehensive structural health monitoring (SHM) system was installed on the bridge to collect the dynamic and static responses of the bridge. In this study, truck-induced strains were analyzed using a cross-correlation methodology to evaluate the IRIB’s condition. The reason for using cross-correlation methodology is due to its simplicity, speed, and economy. Meanwhile, because this methodology builds correlations of a network of sensors, it is effective to eliminate false alarms caused by a malfuntioning sensor. The strain data was continuously collected at 12HZ from 2015 to 2019 by the SHM system. The long-term thermal trend in the strain data was filtered using a moving average technique to obtain the live load-induced strains. Then, the truck-induced strain data was identified by comparing the live load strains with strains from six load tests. Consistent correlation matrices were produced and monitored continuously by using truck-induced strains. Statistical analyses were then performed on the correlations of each pair of sensors, and threshold values were established for use in detecting significant changes in the correlations. By applying the cross-correlation methodology to truck-induced strain, future minor abnormal behavior of the IRIB can be detected as well as when and where the abnormal behavior occurred.