Speaker
Description
Vibration-based structural diagnosis is considered as an essential technique for assessing the condition of composite-steel structures in marine environments, where durability and corrosion resistance are main considerations. This research proposed a framework that uses the application of continuous wavelet transform on curvature-derived debonding information. Three levels of debonding (4%, 6%, and 9%) in terms of percentage areas positioned at two different locations and propagating to the length were considered. A finite element dynamic implicit model was used to replace the real experimental vibration based modal analysis to compute the frequency response function. Based on the results, an individual modes and wavelet scale can result in misleading pattern when trying to accurately determine the location of the debonded region. However, the wavelet coefficients across multiple scales and modes provides a more accurate indication of structural irregularities. Moreover, the suggested debonding index effectively determines when a defect is centrally located rather than at the edge. This is because the closeness of debonding to the structural boundary causes even minor boundary-induced changes to be significantly amplified, resulting in the debonding-induced singularity overlapping with the boundary response. The research indicated that the proposed final index that combined wavelet-curvature based diagnostic approach enhances sensitivity in detecting debonding, especially for smaller sizes, which are challenging to identify with conventional global methods. The suggested framework is an effective method with good sensitivity and resolution for debonding defect in composite-steel marine structures of all sizes.
| Topic Area | Fatigue, Corrosion and Structural Integrity of Engineering Structures and Mechanical Systems |
|---|