Structural Health Monitoring, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 31, 2024

Lamb-wave-based structural health monitoring (SHM) technology for damage location in plate-like structures relies on the postprocessing of captured signals after interacting with damage. Traditional methods typically leverage time flight (ToF) scattered waves from However, these are prone to reflected boundaries which mix This is a vital problem faced by most ToF-based detection methods, seriously narrows inspection area. To tackle this problem, machine learning framework, consisting multiscale spatiotemporal (MSST) fusion network, proposed facilitate accurate extraction ToF through eliminating influence boundary reflections. Experiments conducted time-domain Lamb wave recorded tactically designed piezoelectric sensor array 2-mm-thick Al-6061 plate. A pair circle magnets attached onto plate as reflectors. Through step-by-step moving predefined grids, corresponding measured construct database. An MSST subsequently minimize error between estimated and theoretical ToFs, wavelet coefficients transducer position inputs. The model trained Adam algorithm where 80% samples database used training rest evaluation. final validations scatters off grids. Results demonstrate that neural network architecture can effectively eliminate reflections enable precise allows enlargement area presents promising useful tool enhancing performance existing SHM complex structures.

Language: Английский