Self-reference broadband local wavenumber estimation (SRB-LWE) for defect assessment in composites

Local wavenumber estimation (LWE) applied to a full wavefield response is a powerful approach for detecting and characterizing defects in a composite structure. However, the narrowband nature of the traditional LWE techniques brings several challenges for application on actual test cases. This study proposes a self-reference broadband version of the LWE technique. The broadband vibrations are injected using low-power piezoelectric actuators (sine sweep signal) or using pulsed laser excitation in the thermoelastic regime. The out-of-plane velocity response of the surface is recorded using an infrared scanning laser Doppler vibrometer. The dispersive Lamb wave behavior, corresponding to the damage-free base material, is identified from the broadband vibrational response. Using the identified dispersion curves (i.e. self-reference approach), a Lamb mode passband filter bank in the wavenumber-frequency domain is constructed. Searching for the maximum bandpower density in function of the assumed material thickness provides a robust estimate of the effective local thickness of the tested component, and as such yields a detection and evaluation of damage. The performance of the self-reference broadband LWE algorithm is demonstrated on aluminum plates with various flat bottom holes, as well as on cross-ply CFRP aircraft components with a stiffener disbond and barely visible impact damage. Compared to the traditional narrowband LWE approaches, the proposed self-reference broadband LWE method allows a higher level of automation, removes the need for a priori knowledge on the material and/or defect properties, and results in an improved characterization of defects.

Automated summary

A self-reference broadband version of the local wavenumber estimation (LWE) technique is proposed in this study to detect and characterize defects in composite structures. By injecting broadband vibrations using different methods and analyzing the surface response, this approach allows for automation and improved defect characterization.