Damage Detection of Steel Moment Frames with Multiple Damages using Wavelet Transform

Regarding the possibility of lifetime reduction of the structure due to different factors, including cracking and phenomena, such as earthquakes, it is crucial to constantly monitor the structure so that the location and severity of the damage can be timely identified to correctly estimate the remaining lifetime of the structure. Wavelet transform is a robust tool for identifying local structural damages, by which the damage location is identified by observing a disturbance in the wavelet coefficients of the damaged structure, regardless of the main (undamaged) structure response. This paper addresses the application of wavelet transforms for identifying multiple cracks in steel moment resisting frames, which has received less attention in the literature. To this end, different crack modes were considered in the form of element removal for one-story and two-story frames, and the first four mode shapes of the damaged structure were derived using the ABAQUS. Finally, the derived mode shapes were subjected to continuous coif5 and sym2 wavelet transform, and the capability of the method in identifying cracks in structures accounting for the effect of mode order and damage location was evaluated. The results indicated that if the proper wavelet was selected, the proposed method would be able to identify the damage location based on the response signal using a simple, robust, and reliable way. The results also showed that the damage in the beam was identified with a higher accuracy than the column foot. It was also shown that the location of the damage in the mid-span of the beam can be identified with a higher accuracy than the damage near to the column abutments. Also, it was observed that higher modes could identify the damage location with a higher resolution.