Nondestructive crack inspection method with distributed acoustic field detection and an SVM

Structural health monitoring (SHM) is critical for aircraft safety and maintenance. This paper presents a DAS-inspired nondestructive crack inspection framework that identifies cracks by sensing perturbations in radiated acoustic-field patterns rather than relying on point-wise sensors. A coupled structural-acoustic model of an Al7075 aircraft-skin panel is established to generate distributed sound-pressure-level (SPL) fields, and virtual distributed sensing trajectories are used to emulate a dense readout within the simulated air domain. A region-of-interest (ROI)-guided sensing-path selection strategy is developed to highlight crack-induced field distortions. Statistical descriptors, specifically skewness and kurtosis, are employed to form a two-dimensional feature space for crack classification using a linear support vector machine (SVM). Trained exclusively on the 30 N dataset, the proposed method achieves accuracies of 93.33% and 96.00% on the 30 N and 100 N test sets, respectively. The proposed framework provides a promising technical pathway for real-time structural health monitoring and early warning of aircraft-skin risks.