ABSTRACT Guided wave structural health monitoring (SHM) has been extensively investigated to evaluate and assess the damage in pipelines, plates, and welded joints due to guided wave capabilities that have the capacity to travel over large distances and are sensitive to surface and subsurface flaws. However, guided‐wave signals can be complex due to dispersion, mode conversion, boundary reflections, and variation in the environment, and thus, they are challenging to interpret, which complicates the process and requires effective feature extraction methods. This review focuses on three commonly used features in SHM, which are amplitude change, time of arrival (ToA), and spectral shift. From the previous studies, these features have been used in a variety of damage assessments, although their diagnostic accuracy may be compromised due to temperature variations, multimode interference, and sensor coupling variations. The direct cross‐feature comparison is given to summarize the reported damage types, indicators of their representative sensitivity, and key limitations to consider for feature combination to enhance the reliability and accuracy of signal processing and interpretation in field application.
Manap et al. (Wed,) studied this question.