Damage Detection, Localization and Size Estimation Using Broadband Correlation-Based Imaging
Abstract
Imaging approaches based on guided waves aim at detecting, locating and estimating the damage severity on a structure. The actual limitations of imaging approaches are that (1) sensitivity do damage depends on frequency used, and (2) the damage growth can be observed but not yet quantified. This paper presents a robust methodology for damage imaging and size estimation using reconstructed broadband signals, where measurements of the transfer function between each emitter and receiver are made using a sub-band decomposition strategy. Pristine transfer function is subtracted from damage transfer function signature. Imaging is conducted using a correlation-based approach (Excitelet), and dispersion compensation with reconstructed broadband signals. The approach is validated experimentally on a 1.54 mm thick aluminium plate, where only three piezoceramic transducers are bonded on the structure. Measurements are taken for two artificial damage of 13 mm and accurate detection and dimensioning is achieved. The analysis of the transfer functions using the A0 mode shows that the wavelengths reflecting most of the energy are associated to the damage size within an accuracy of one millimetre.
Origin : Files produced by the author(s)
Loading...