Modal strain energy based enhanced approaches for damage detection and severity estimation

Abstract

Structural health monitoring (SHM) has been utilized to assess structural deficiency for preventing the invisible failure turning into collapses. Damage localization and severity assessment based on vibration-based characteristics have received considerable attention in civil and engineering fields during recent decades. This study presents a modal strain energy (MSE) based damage detection of a beam-like system. Particularly, based on the performances of existing MSE based damage indices (DIs), a robust and fast technique, called averaging scheme is developed. Observing from the two original DIs that one of them leads to underestimations while overestimations are witnessed using the other one when estimating damage severity, the average of their severity estimations absolutely is supposed to result in better anticipations. Furthermore, another MSE-based updating procedure that was also established based on the original method to reach more accurate predictions is also improved in this study. However, the updating technique considers only the fundamental mode while higher modes are ignored. It is assumed that damage can be caused by higher modes and therefore accounting for more modes possibly lead to better failure identification. Hence, in this study, a numerical investigation is carried out to examine the feasibility of the averaging scheme's deployment on a cantilever beam. The contribution of higher modes to the performance of the proposed technique as well as the updating procedure is also evaluated. The numerical validation shows that the proposed averaging scheme is outstanding since it results in damage identification results comparable to that of the updating procedure but more promptly. Furthermore, compared to using only the fundamental mode, the cumulative contribution of higher modes, particularly, the lowest four modes, tremendously leads to more accurate damage identification, especially under noisy conditions.