Damage detection of bridges based on responses of a moving vehicle under impact excitation

Abstract

Damage detection of bridges based on mode shapes constructed from signals extracted from a moving vehicle has been a fascinating idea especially for those bridges without permanent structural health monitoring systems. As mode shapes are not only important modal properties but also sensitive damage indicators, it’s desirable to seek more convenient and economical ways to construct accurate mode shapes for identification damage so as to obviate the need for traditional methods of installing sensors on bridges. If an instrumental vehicle runs over a bridge, thereby acting as an exciter and a sensor to collect signals, the problem will be solved perfectly. An additional impact excitation generated by a drop mass on the instrumental vehicle may help reduce the adverse effect of road surface roughness, which may otherwise lead to erroneous results. The accelerations of a moving vehicle can be measured more accurately and conveniently than other signals. Moreover, it contains modal information such as natural frequencies and mode shapes of the bridge. The mode shapes associated with the bridge natural frequencies can hence be constructed from the acceleration responses. A suitable vehicle-bridge interaction system is developed to simulate the process of a vehicle moving over a bridge and generate the acceleration responses for further analysis. The fast Fourier transform and suitable bandpass filters can be used to extract the component responses associated with the frequencies for constructing mode shapes. The additional impact excitation applied on the vehicle helps to excite the bridge further to overcome the adverse effect of various uncertainties. The proposed method can be used to construct the mode shapes of simply supported bridges as well as multi-span continuous bridges taken into account road surface roughness and measurement noise. The feasibility is verified not only numerically but also experimentally by a simply supported aluminium model bridge. Continuous wavelet transform (CWT) can help identify local changes in the constructed mode shapes which are caused by the reduction of stiffness. Co-ordinate modal assurance criteria (COMAC) can help identify the difference between the constructed and theoretical mode shapes and hence find out the locations of damage. A numerical study verifies the effectiveness of the two damage detection methods by considering constant-section and variable-section continuous bridges. However, as the information of damage is often masked by the road surface roughness and measurement noise, an additional impact excitation helps to boost the accuracy of identification results. These methods are also verified in the laboratory by a simply supported aluminium model bridge. The proposed methods have also been extended to construct the flexural and torsional mode shapes for damage detection taking into account the transverse eccentricity of traverse path of vehicle. In summary, a convenient and economical methodology of damage detection has been developed based on the constructed mode shapes obtained from a moving instrumental vehicle under additional impact excitation. The proposed methodology has been verified both numerically and experimentally. It can provide a convenient way to identify the locations of damage for further examination