Real-Time Structural Damage Detection Using Only Signals Acquired During Impact Events

Abstract

This work presents a novel, real-time, and baseline-free methodology for structural damage identification during impact events, utilizing the Hilbert-Huang Transform (HHT) to analyse acquired stress wave signals generated by the impact. The proposed technique is based on the decomposition of wave propagation modes induced by both low- and high-velocity impacts, with a focus on distinguishing between extensional (symmetric, high-frequency) and flexural (antisymmetric, low-frequency) components of the AE signals. The empirical mode decomposition (EMD) is applied to extract intrinsic mode functions (IMFs), enabling the separation and reconstruction of the fundamental wave modes. The instantaneous energy content of each mode is then calculated using HHT, and a damage index, Λ, is introduced, a dimensionless quantity that characterizes the shape and spread of the instantaneous energy spectrum obtained via HHT. Experimental validation was carried out on two sets of impact results including flat plates and complex composite structures. Impact scenarios varied from low velocity to hypervelocity regimes, using projectiles of different materials and geometries. The results demonstrate that a Λ value below a threshold corresponds to elastic impacts with no or minor deformation, while values approaching or exceeding the threshold indicate the onset of permanent damage, such as indentation, penetration, or complete perforation. For highly damaged structures, Λ values increased significantly, confirming the dominance of extensional wave components in such cases. This technique shows strong potential for real-time structural health monitoring (SHM) in aerospace and other safety-critical applications, offering high sensitivity to damage without requiring baseline signals. Furthermore, its robustness across different materials and geometries suggests wide applicability for both laboratory and operational environments.