Weidong Zhu / University of Maryland; Baltimore County
Damage detection approaches relying on flexural vibration shapes (VSs) have been widely developed. However, these approaches are inapplicable to some structures governed mainly by tension and not bending, such as cables in a cable-stayed bridge, because bending effects on their flexural vibrations are much smaller compared to their tension effects. In contrast, longitudinal VSs can be useful for detecting such damages, although they are much more difficult to measure than flexural VSs. This study reports a novel methodology for damage detection of longitudinally vibrating structures relying on their longitudinal VSs. For such structures, damage-caused axial stiffness changes can induce discontinuities in derivatives of their longitudinal VSs, by which the damages can be detected. The capability of the methodology is verified using analytical models of damaged bars. The applicability of the methodology is experimentally validated by identifying a two-sided notch in an aluminum bar whose longitudinal VSs are acquired through 3D laser scanning.