On the effect of spatial sampling in damage detection of cracked beams by continuous wavelet transform

Modern measurement techniques are improving in capability to capture spatial displacement fields occurring in deformed structures with high precision and in a quasi-continuous manner. This in turn has made the use of vibration-based damage identification methods more effective and reliable for real applications. However, practical measurement and data processing issues still present barriers to the application of these methods in identifying several types of structural damage. This paper deals with spatial Continuous Wavelet Transform (CWT) damage identification methods in beam structures with the aim of addressing the following key questions: (i) Can the cost of damage detection be reduced by down-sampling? (ii) What is the minimum number of sampling intervals required for optimal damage detection? The first three free vibration modes of a cantilever and a simple supported beam with an edge open crack are numerically simulated. A thorough parametric study is carried out by taking into account the key parameters governing the problem, including the wavelet mother function, level of noise, crack depth and location, mechanical and geometrical parameters of the beam and the padding method to reduce border distortions. The results are employed to assess the optimal number of sampling intervals for effective damage detection. (C) 2015 Elsevier Ltd. All rights reserved.