The continued and expanded use of composite materials in aerospace applications necessitates structural health monitoring and/or nondestructive evaluation techniques that can provide quantitative and detailed damage information for layered plate-like components (such as composite laminates). Guided wavefield methods are at the basis of a number of promising techniques for the detection and the characterization of damage in plate-like structures. Among the processing techniques that have been proposed for guided wavefield analysis, the estimation of instantaneous and local wavenumbers can lead to effective metrics that quantify the size and the depth of delaminations in composite laminates. This article reports the application of both instantaneous and local wavenumber damage quantification techniques to guided wavefield data for delaminated composite laminates. The techniques are applied to experimental data for a simple single delamination case and to simulated data for a more complex multi-ply delamination case. The two techniques are compared in terms of accuracy in damage characterization and computational demand. The proposed methodologies can be considered as steps toward a hybrid structural health monitoring/nondestructive evaluation approach for damage assessment in composites.

The paper presents a Compressed Sensing technique for the reconstruction of guided wavefields. Structural inspections based on the analysis of guided wavefields have proven to be effective at detecting and characterizing damage. However, wavefield detection is often a time consuming process, which limits practicality. The proposed reconstruction technique estimates the location of sources and structural features interacting with the waves from a set of sparse measurements. Such features include damage, described as a scattering source. The wavefield is reconstructed by employing information on the dispersion properties of the medium under consideration. The procedure is illustrated through a one-dimensional analytical example, and subsequently applied to the reconstruction of an experimental wavefield in a composite panel with an artificial delamination. The results confirm the ability of the technique to identify the defect, while reconstructing the wavefield with good accuracy using a significantly reduced number of measurements.

This paper presents the use of a kernel-based machine learning strategy targeting classification and regression tasks in view of automatic flaw(s) detection, localization and characterization. The studied use-case is a structural health monitoring configuration with an array of piezoelectric sensors integrated on aluminum panels affected by flaws of various positions and dimensions. The measured guided wave signals are post processed with a guided wave imaging algorithm in order to obtain an image representing the health of each specimen. These images are then used as inputs to build classification and regression models. In this paper, an extensive numerical validation campaign is conducted to validate the process. Then the inversion is applied to an experimental campaign, which demonstrate the ability to use a numerically-built model to invert experimental data.