Guided wave‐based damage assessment on welded steel I‐beam under ambient temperature variations

Summary Welded steel I‐beams are widely used in civil infrastructures and industrial facilities as the major load‐carrying members. The fillet weld zone, which connects the web plate and flange plate, is vulnerable to defects and fatigue cracks during the long service life, which may result in catastrophic accidents. In this study, a guided wave‐based damage assessment method is presented to monitor the weld zone of a steel I‐beam under ambient temperature variations. The semi‐analytical finite element (SAFE) analysis was performed to investigate the characteristics of guided wave propagation in a welded I‐beam. A signal‐processing procedure that incorporates the principal component analysis (PCA) and independent component analysis (ICA) was proposed for damage assessment. The PCA was performed to identify and eliminate significant environmental effects from the guided wave signals. Reconstructed residual signals were obtained from the PCA for damage detection. The ICA was performed to improve the damage localization by identifying the wave packets that are closely related to the occurrence of damage. Lab‐scale experiments were performed on a steel I‐beam under various ambient temperatures. The results show that the inflicted cuts on the weld line have been successfully detected. The damage severity can be qualitatively evaluated based on the magnitudes of Q‐statistics through the PCA. The results of the damage localization can be effectively improved by incorporating the ICA‐based signal decomposition.