Shape monitoring and damage identification in stiffened plates using inverse finite element method and Bayesian learning

Abstract

Stiffened plate structures are common structural forms that are widely used in various fields. Structural Health Monitoring (SHM) is an important tool to maintain their safe operation. The inverse finite element method (iFEM) is a state-of-the-art methodology for SHM that can precisely reconstruct full-field displacements through limited strain sensors. In this paper, we study the stiffened cantilever plate and propose a complete damage localization and quantification system based on iFEM. By solving the damage location index, we can recognize the damage locations in real time, independent of loads and load combinations. Moreover, the strain modes at strain sensors are obtained by modal tests, which can be transformed into full degrees of freedom (DOFs) mode shapes by the iFEM. The damage severities can be further determined using the Bayesian learning (BL) algorithm. In this way, the mode shapes required for damage identification are transformed into strain modes which are easier to obtain, especially for structures with rotational DOFs such as plates and shells. Considering different damage scenarios, the proposed method can detect true locations and severities of damage even with a limited number of strain sensors and under measurement noise.