Bridge identification based on the double-pass mass-addition method considering surface roughness

Abstract

Recent years have witnessed a rapid development in the use of a passing vehicle for bridge identification. The concept of contact point response is introduced in this thesis, which is superior to the normally adopted chassis or axle responses for identifying the bridge modal properties. The presence of surface roughness, however, severely affects the vehicle-bridge interaction and hence the vehicle vibration responses, thereby rendering many vehicle-response-based bridge identification methods ineffective. Proper identification of the surface roughness will help minimize the associated uncertainties and improve the credibility of numerical simulation. A double-pass mass-addition method is proposed to estimate the bridge surface roughness profile, identify the bridge frequencies, construct the mode shapes and locate any damage using the contact point response of a test vehicle passing on the bridge twice with extra mass added for the second time. Theoretical derivation, numerical simulation and experimental verification are presented in the thesis. Firstly, the method to estimate the bridge surface roughness is elaborated. No baseline data is required. Numerical simulation with simply supported and continuous bridges shows that the roughness profiles can be estimated with satisfactory accuracy. Two error indicators, namely modified standard error (MSTD) and relative area difference (RAD), are suggested to evaluate the effectiveness of the method. A parametric study is conducted to investigate the influence of various factors on the identification performance. Next, the application of contact point response difference obtained by the proposed method to bridge modal identification is demonstrated. It is revealed theoretically that it contains bridge information and is relatively immune to the contamination of surface roughness. The bridge frequencies and mode shapes can then be extracted from the contact point acceleration difference (CPAD) using fast Fourier transform and Hilbert transform, respectively. Numerical analysis verifies the feasibility of the proposed method. Then, damage detection can be conducted based on the mode shapes obtained. Two damage indicators, namely wavelet transform coefficients and coordinate modal assurance criterion (COMAC), are adopted to estimate possible damage location(s). Simulations confirm that the damage can be located for both single and multiple damage cases. Finally, experiment is conducted on a simply supported aluminum channel beam with a specially designed model vehicle, and the surface roughness is formed using thin wooden strips of different lengths. Artificial damage can be inflicted as pairs of slots cut from the flanges of beam with mass compensation for damage detection. The surface roughness estimation performance in terms of MSTD is similar to that in numerical simulation. The first three frequencies and the first two mode shapes can be constructed from CPAD. The damage locations can be roughly estimated. In summary, this thesis contributes primarily to the indirect approach for bridge identification. An economical and convenient method is developed to estimate bridge roughness profile, identify modal parameters and locate possible damage using the response of a moving vehicle. The novelty lies in the application of contact point response to obtain the bridge information and the development of double-pass mass-addition method to resolve the adverse effect of bridge surface roughness.