In this paper, the analysis of errors in the determination of modal curvatures and their propagation in the experimental evaluation of damage is carried out with reference to a simple structure, i.e., a slender beam. The beam is suspended with springs to reproduce conditions similar to floating structures like ships, and damage is represented by a local reduction of stiffness. The mode shapes, from which curvatures are calculated with a finite difference scheme, are experimentally identified using operational modal analysis on the recorded acceleration data. The damage identification is based on multiple hypothesis testing employing four different damage indexes, three related to modal strain energy and one to modal curvature integral. Truncation error and bias error affecting the modal strain energy computation are modelled and their sensitivity on sensor spacing is analysed. After damage is localized, the uncertainty in damage severity is calculated using Monte Carlo method adding the experimentally identified measurement noise on the analytical modes.
Hybrid Uncertainty Analysis of Damage Indexes Based on Modal Strain Energy / Dessi, Daniele; Passacantilli, Fabio; Venturi, Andrea. - 514 LNCE:(2024), pp. 561-572. (Intervento presentato al convegno 10th International Operational Modal Analysis Conference (IOMAC 2024) tenutosi a Naples; Italy) [10.1007/978-3-031-61421-7_54].
Hybrid Uncertainty Analysis of Damage Indexes Based on Modal Strain Energy
Daniele Dessi
Primo
;Andrea VenturiUltimo
2024
Abstract
In this paper, the analysis of errors in the determination of modal curvatures and their propagation in the experimental evaluation of damage is carried out with reference to a simple structure, i.e., a slender beam. The beam is suspended with springs to reproduce conditions similar to floating structures like ships, and damage is represented by a local reduction of stiffness. The mode shapes, from which curvatures are calculated with a finite difference scheme, are experimentally identified using operational modal analysis on the recorded acceleration data. The damage identification is based on multiple hypothesis testing employing four different damage indexes, three related to modal strain energy and one to modal curvature integral. Truncation error and bias error affecting the modal strain energy computation are modelled and their sensitivity on sensor spacing is analysed. After damage is localized, the uncertainty in damage severity is calculated using Monte Carlo method adding the experimentally identified measurement noise on the analytical modes.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.