Key points are not available for this paper at this time.
Timely detection of structural damage to orthotropic steel box girders is crucial for ensuring operational safety and preventing catastrophic failures. This study proposes a hybrid signal processing method that combines frequency response function (FRF), wavelet packet decomposition (WPD), and permutation entropy (PE) for identifying the location and severity of damage under impact excitation. Firstly, the acceleration response generated by hammering is converted into a frequency response function to standardize the structural output and enhance frequency domain sensitivity. Subsequently, the FRF signal was decomposed into multiple sub bands using WPD for multi-resolution analysis, and permutation entropy was calculated for each sub band to quantify signal complexity. A damage sensitivity index, permutation entropy difference (PED), is defined as the difference in permutation entropy between healthy and damaged states. A finite element model of a steel box girder was developed to simulate various damage scenarios, including changes in location and severity. The numerical results under white noise interference indicate that the proposed FRF-WPPE method can accurately locate and quantify damage in noisy environments. The combination of dynamic characteristic analysis of FRF, time-frequency resolution of WPD, and nonlinear sensitivity of PE enhances the anti-interference ability against excitation and measurement uncertainty factors, highlighting the potential of this method in practical structural health monitoring applications. • A hybrid FRF-WPPE framework has been developed for the damage detection of steel box girders. • FRF improves frequency sensitivity; WPPE enables damage localization and severity quantification. • Numerical and scaled experiments verify noise robustness and engineering applicability.
Zhou et al. (Mon,) studied this question.