Applications of continuous wavelet transforms on ice load signals

This paper gives an introduction to how Continuous Wavelet Transform (CWT) can be applied in analyzing ice load time series to gain more information about the distribution of power on different frequencies in the time domain. The aim of the study was to localize short intervals of intermittent crushing in initially long time series. Intermittent crushing was supposed to occur with dominant frequencies in the range of the fundamental frequencies of the structure and to give resonance-like response histories of the structure. To ensure that the localization of periods of intermittent crushing was realistic, corresponding time signals of normalized acceleration were plotted together with selected CWT coefficients. Five different time records were selected for analyses. The length of intervals was calculated from the up and down crossing of chosen threshold values for both normalized CWT coefficients and normalized acceleration signals. The presented method seems to be suitable for analyzing signals in time and frequency domains simultaneously, however the threshold values have to be tuned in the case of new structures and/or new loading situations. Together with already existing tools for analyses, CWT will give more information from ice load time series than traditional methods. More work is needed on significance testing of wavelet coefficients and development of a proper background spectrum for such testing.