Detection of Barely Visible Indentation Damage in Composite Sandwich Structures Using Nonlinear Guided-Wave Features

Abstract

Composite sandwich structures (CSSs) serve a multitude of applications in the transportation, construction, and aerospace fields. Their structural integrity is essential to prevent any catastrophic failures, especially in large massive structures. CSSs are prone to low-velocity impacts which may create barely visible indentation damage (BVID) that can lead to serious defects over time resulting from irregular distribution of loads.

The main purpose of this study is to develop a robust framework for the selection of an actuation method and the use of the nonlinearly-generated features from its interaction with BVID to accurately localize damage. The frequency content of ultrasound signals actuated and received by a network of piezoelectric transducers are scrutinized to study the nonlinear effects of different sizes of BVID on Lamb-wave propagation in CSSs. Different actuation approaches and nonlinear parameters are explored to find the optimal method of damage detection and localization. The proposed novel technique involves the use of cross-correlation factors computed by considering different frequency ranges corresponding to fundamental-actuated frequencies and nonlinearly-generated frequencies. An imaging technique using the computed-cross-correlation factors is applied to generate a heatmap to localize damage.

The novel technique provided accurate localization of about 45 mm from the actual location, while the resolution of the conventional linear technique is much lower reaching 80 mm for a 2-mm-indentation depth.

For future work, the developed technique can be automated to give accurate results without having to collect a lot of data, furthermore, machine learning models can be developed to assist in the feature extraction of non-linear behavior.