Mechanics-based error analysis and testing parameter design for bridge modal identification under non-uniform vehicle speed

Existing Vehicle Scanning Methods (VSM) typically assume the test vehicle moves at a constant speed. However, in practical engineering, non-uniform motion is inevitable due to road conditions and driver operations, creating a complex time-varying multibody dynamics problem. To address this, this study theoretically derives the analytical solution for the vehicle-bridge contact point response under non-uniform motion. The validity of the proposed mechanics-based model is verified via finite element simulations (MATLAB and ABAQUS). Furthermore, a systematic sensitivity analysis is conducted to investigate the impact of different speed variation patterns (segmented, periodic, and random) on the accuracy of modal shape identification. The results reveal that random speed variations, contrary to intuition, introduce less error than segmented constant speeds. Finally, to optimize the design of field testing schemes, this study proposes a critical operational criterion: controlling speed fluctuation within ±0.3 m/s ensures identification accuracy, particularly for higher-order modes. These findings provide a theoretical foundation for designing more robust indirect bridge health monitoring systems.