Identification of multiple bridge frequencies using a movable test vehicle by approximating axle responses to contact-point responses: theory and experiment

Identifying bridge frequencies using a movable test vehicle is highly regarded for its mobility, cost-effectiveness, and broad monitoring coverage. Although experiments have shown effectiveness in extracting the fundamental bridge frequency, there are certain challenges in identifying higher-order frequencies, primarily owing to the impact of road surface roughness and vehicle self-vibrations. Theoretical studies have proposed mitigation strategies such as using contact-point (CP) responses and time-lagged signal subtraction. However, generating accurate CP responses in real-world measurements remains challenging owing to uncertainties in vehicle parameter measurements and susceptibility to environmental noise. This study proposed a novel vehicle scanning approach that approximates the axle response to the CP response. Initially, closed-form solutions for the axle-to-CP response deviations of a quarter-car vehicle traversing a simple beam were derived, considering scenarios with and without surface roughness. Subsequently, parametric studies were performed to analyze the condition of the test vehicle design for approximating the axle responses to the CP responses. Consequently, finite element simulations of the vehicle-bridge interaction system were established to validate the theoretical derivations. Finally, the theoretical findings were used to design a scaled test vehicle, and laboratory tests of a scaled vehicle-bridge interaction system verified the proposed approach. The theoretical and experimental findings suggest that, by designing a test vehicle with a sufficiently large axle stiffness, the axle responses can be approximated as CP responses. As a result, the measurement of the axle accelerations on a test vehicle facilitated the identification of multiple bridge frequencies.