Vehicle Scanning Method for Bridge Mode Shape Extraction using Hilbert and Wavelet Transforms: Numerical and Experimental Validations

The vehicle scanning method (VSM) is an emerging approach for indirect identification of bridge modal properties. However, its robustness under realistic operational conditions has not been sufficiently validated. This study presents a systematic, multi-level validation of VSM-based mode shape reconstruction through numerical simulations, laboratory-scale experiments, and full-scale field tests. Signal processing is performed using the Hilbert transform (HT) and Wavelet transform (WT), combined with bandpass filtering, to extract mode shapes from both vehicle body and axle response. Parametric simulations identify road roughness and bridge damping as the primary constraints on reconstruction accuracy, while higher vehicle speeds and environmental noise further degrade performance. Conversely, increased vehicle damping modestly facilitates extraction. A key finding is that axle response is comparably more reliable under high-speed or noisy conditions, whereas body response is less affected by road-induced vibrations. Laboratory experiments further reveal practical sensitivities not fully captured in idealized simulations, including motor-induced vibrations and speed variation. Finally, full-scale field tests on a cable-stayed bridge demonstrate high repeatability of the extracted fundamental mode shapes across multiple test runs, supporting the potential of indirect mode shape reconstruction in real-world environments. Overall, WT-based reconstruction generally provides higher accuracy than HT.