Comparative study on steel beams damage detection based on continuous and discrete wavelet transforms of static and dynamic responses

Engineering structures experience different conditions during their life time that may result in damages in some structural elements under these conditions. Beams are considered as the main components of building structures, bridges, and the most important parts of the industrial machinery therefore, it is important to identify the various situations of occurred local damages. In this paper the steel beam was modeled by the plain-ends of support conditions in the healthy and damaged states in the finite element software of ABAQUS and the static analysis was performed by the influence of constant static load. The frequency analysis of the steel beam was conducted without the influence of loading. Changes of fitting curve coefficients of static displacements resulting from the polynomial regression and also changes in the frequency values of the different modes of healthy and damaged states confirm the damage of the beam. In order to detect the various failure situations, the difference of static displacements and also the difference of the interpolated modes of healthy and damaged states analyzed using the continuous and discrete wavelet transformations. The interpolated details resulted from analyzing the continuous and discrete wavelet transformations at the site of damages show the irregularities and perturbations in the wavelet coefficients, such that the relative minimums and maximums of jump in wavelet coefficients happened in all investigated states in the site of damages. Also, results show the wavelet coefficients sensitivity at the site of each damage independent of the wavelet coefficients sensitivity in other damaged sites with different intensities. Also, the occurrence place of minimums and maximums of wavelet coefficients coincide in a damaged situation with different intensities (with approximately zero error).