Applications of Wavelet Transform in Ice-Induced Vibrations of Structures

Spectral methods such as Fourier Transform are used to analyze the response of structures to reveal characteristics of time series in frequency domain, as is typically done in the study of ice-induced vibrations. The limitations of these methods lie in that they assume the signal to be stationary and decompose the signal as a whole. Wavelet Transform has been developed for several decades to overcome these limitations. In this paper, the technique of Wavelet Transform is applied to ice force and structural response signals that provides a means to simultaneously analyze the data in both time and frequency domains. The results show that the applied method provides useful visualization of signals, allowing fast identification of vibration events of interest from long time series. A frequency locked-in event can be identified at intervals with high energy concentration at the fundamental frequency of the structure and relatively small variations in phase lag between the load and response. A random vibration is found that energy at the fundamental frequency of the structure occurs intermittently in time domain and almost all frequencies are excited showing a wide band spectrum. Simulation results obtained from a purpose-developed numerical model show that the same characteristics can be observed compared to full-scale measurements from the Norstromsgrund lighthouse.