A fast damage locating approach using digital damage fingerprints extracted from Lamb wave signals

A fast damage locating approach using digital damage fingerprint data, extracted from raw Lamb wave signals and accommodated in a damage parameters database (DPD), was developed in this study. A new multilayer feedforward artificial neural network was designed and trained with the DPD under the supervision of an error-backpropagation algorithm. Assisted by an active system for online structural health monitoring, the proposed method was validated by locating actual delamination and through-thickness holes in quasi-isotropic CF/EP (T650/F584) composite laminates. Compared with a quantitative methodology for evaluating full damage parameters developed in an earlier study (Su and Ye 2003 J. Intell. Mater. Syst. Struct. 16 97–111), the present approach performs damage evaluation much more quickly and cost-effectively by determining damage location only.