[1B4] Acousto-ultrasonic modal identification for health monitoring of safety-critical engineering structures
A Gullapalli¹, A Kundu¹, C Featherston¹, R Pullin¹ and M Naaman²
¹Cardiff University, UK
²Babcock International Group, UK
Acousto-ultrasonic inspection techniques have shown great promise for monitoring progressive damage in thin-walled structures and rely on extracting essential features effectively and efficiently from the acquired signals. These signal features contain damage fingerprints and can be used for an automated framework for assessment of structural integrity. The signal features are inherently linked to guided wave physics and the behaviour of fundamental guided wave modes, the sensitivity of which to various damage types is well established. This study presents a novel ultrasonic signal reconstruction approach that combines experimental measurements with physics-informed composite waveguide dispersion characteristics and harmonic wave propagation functions to reconstruct the ultrasonic guided wave modes. The modal amplitudes and dispersion characteristics were calibrated using deterministic and Bayesian probabilistic joint parameter estimation techniques, thereby helping to quantifying uncertainties in both experimental measurements and model predictions. The reconstructed signals showed good agreement with their experimental counterparts, highlighting the ability of the proposed framework to capture the true signal characteristics along with the associated confidence estimates. These calibrated guided wave modes will serve as a foundation to establish reliable damage signatures and hence facilitate the development of a physics-informed, robust and non-intrusive structural health monitoring system with real-time decision-making capabilities.
