Bridge abnormality detection utilizing acceleration of a moving vehicle

Abstract

Drive-by bridge health monitoring (BHM) method inspecting bridge health condition utilizing measured data on the test vehicle has been investigated. While its great advantage is economy, the challenge of the drive-by BHM is pursuing enough accuracy for real-world application. Many drive-by methods of identifying modal parameters of the bridge are proposed. However, it is challenging to obtain high accuracy in the experiment in many cases especially for damage detection, because the measured vehicle acceleration is the coupled vibrations of the vehicle, bridge and road roughness profile interaction. Meanwhile, a method of detecting abnormalities in the bridge without referring any detail of the damage type or specific modal parameter values are less difficult to realize. Therefore, this study aims to propose a bridge abnormality detection method utilizing measured vehicle acceleration.

The proposed method can determine existence of any abnormality utilizing difference of the measured vehicle acceleration at the inspection time and the reference acceleration when the bridge is considered healthy. In specific, the subtraction of the test and reference acceleration is calculated in time domain, then the root mean square (RMS) of the subtracted vehicle force is calculated, and this RMS becomes an indicator of the damage level. The band pass filter is applied to the acceleration before the subtraction to eliminate any undesirable vibration component other than the vibration derived from the bridge. The advantage of the proposed method is (1) No information of the vehicle properties is needed as input. (2) The method is theoretically simple and easy to understand. (3) By utilizing hypothesis testing with multiple measured data, it is possible to estimate the level of damage as well as the reliability of the results.

The feasibility of the proposed method is investigated by a laboratory experiment using a test vehicle equipped with accelerometers. The results showed that the method successfully classifies the bridge damage conditions in acceptable accuracy.