Damage evaluation of structures by wavelet analysis.

摘要

Over the last few decades, damage evaluation methods of civil and mechanical structures have been drawing much interest from various fields. Wavelet analysis, a relatively new mathematical and signal processing tool, is one of such methods that have been studied recently. It is a time-frequency analysis that provides more detailed information about non-stationary signals which traditional Fourier analysis miss. This rather new method has been applied to various fields including civil, mechanical and aerospace engineering, especially for damage evaluation and structural health monitoring (SHM).; In this research, Continuous Wavelet Transform (CWT) is used to identify the existence of damage. Two types of full scale concrete structures are tested to investigate the soundness of structures: (1) a Portland cement concrete pavements (PCCP) and (2) a prestressed concrete beam from a highway bridge. In addition, Wavelet Packet Transform (WPT) and Artificial Neural Network (ANN) are applied to a numerically modeled three span cantilever beam to identify the location of damage as well as its severity. This method is found to be promising to establish on-line structural health monitoring for small structures which are subjected to repetitive loads. In order to apply this method to buildings and civil infrastructures, some difficulties need to be addressed. Although many of these difficulties still remain unsolved, accomplishments achieved at this time will be a help to lead to next practical developments.