Investigation of the Natural Frequency Change of the Suspension Bridge Under Operating Conditions

This study addresses the challenge of accurately correlating the bridge natural frequency with influencing factors during ambient vibration by analysing on-site monitored data. This knowledge gap arises from the combined uncertainties of environmental factors and monitoring equipment noise. To tackle this challenge, the Fourier synchrosqueezed transform technique is employed and validated first by the simulated signal, as well as the Welch method. Then the instantaneous frequency of recorded acceleration at the real bridge is tracked, and a distinct diurnal pattern in the natural frequency is revealed. Then the two-stage strategy is adopted for the regression analysis. Firstly, the regression models between the normalised vibration intensity and the normalised frequency change of the vertical mode are established. Building upon these results, the additional factor, namely the effective wind speed, is considered in the second stage. The multiple linear regression model is established between the natural frequency change, the vibration intensity, and the effective wind speed. A thorough comparison of the results from both regression models reveals in-depth statistical insights. This study confirms that vibration intensity has a negative effect on the bridge natural frequency, i.e., higher vibration intensity leads to a decrease in natural frequency. Besides, the study also shows that while the effective wind speed has a statistically significant impact on the frequency change of the vertical modes, vibration intensity (caused by traffic loads) appears to be a more dominant factor.