Crack detection in beams using Hilbert-Huang transform

This study presents a non-destructive method for detecting location and depth of crack in beams. The method utilizes Hilbert-Huang Transform (HHT) as a time-series analysis technique. Crack is considered to be open and has been modeled by rotational spring. First, the natural frequencies of the cracked beam are calculated using Timoshenko’s beam theory. Then by using the vibration signals corresponding to the beam, experimental natural frequencies are calculated by Fast Fourier Transform (FFT) and HHT. Finally, with the help of Artificial Bee Colony, the location and depth of the crack are predicted by minimization of an objective function. The objective function is constructed by the weighted sum of the squared errors between the theoretical and experimental natural frequencies of the cracked beams. To investigate the feasibility of proposed method, cracks in different locations and depths are introduced in steel beams and crack parameters are predicted. The results show that both the location and depth of the crack can be predicted well through the proposed method. Moreover, by using the experimental natural frequencies obtained by HHT method, cracks (especially small depth) can be detected with a better precision than FFT method.