Damage Detection of Bridge from Response of a Passing Vehicle Based on Wavelet Transform Considering Surface Roughness

Abstract

This paper presents a damage detection method using a mobile exciter-sensor system. This system contains a vehicle carrying an sinusoidal exciter to better actuate the vehicle-bridge interaction system and a sensor to acquire dynamic response of the vehicle. When system runs over the bridge with a vertical sinusoidal force applied to the vehicle, the vertical accelerations of the vehicle and bridge are excited and the vibration of the vehicle can be acquired by the sensor. The vertical vibration of the vehicle contains the dynamic properties of the bridge besides those of the vehicle itself. Continuous wavelet transform is applied to analyze the acquired signal and a series of coefficients can be obtained. Although the inflicted damage is not obvious by directly observing these coefficients when the surface roughness of the bridge and measurement noise is considered, a damage indicator is found to reflect the appearance and location of the damage. This method is very attractive because it does not need any reference properties of the bridge but it allows direct identification of the damage from the current response. Numerical studies are conducted to verify the proposed method. The bridge is simply supported and the vehicle is modelled as a mass-spring-damper systems. The damage is inflicted on the bridge by reducing stiffness of an element. The results show that the damage can be correctly identified even if damage is inflicted on two elements.