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

Area of Science:

• Aerospace Engineering
• Materials Science
• Acoustics

Background:

• Structural health monitoring (SHM) is essential for aircraft safety and proactive maintenance.
• Traditional methods often rely on point-wise sensors, which can be limited in coverage and sensitivity.
• Detecting cracks early is crucial to prevent catastrophic failures.

Purpose of the Study:

• To develop a distributed acoustic sensing (DAS)-inspired framework for nondestructive crack inspection in aircraft structures.
• To identify cracks by analyzing perturbations in radiated acoustic-field patterns.
• To enhance real-time structural health monitoring and early warning systems for aircraft skin.

Main Methods:

• A coupled structural-acoustic model of an Al7075 aircraft-skin panel was established to simulate sound pressure level (SPL) fields.
• Virtual distributed sensing trajectories were employed to emulate dense acoustic data acquisition.
• A region-of-interest (ROI)-guided strategy was used for sensing-path selection to detect crack-induced field distortions.
• Statistical descriptors (skewness, kurtosis) and a linear support vector machine (SVM) were used for crack classification.

Main Results:

• The framework achieved high classification accuracies: 93.33% on the 30 N test set and 96.00% on the 100 N test set.
• The method effectively identified crack-induced distortions in the acoustic field.
• The trained model demonstrated robust performance on unseen data.

Conclusions:

• The proposed DAS-inspired framework offers a promising approach for real-time structural health monitoring of aircraft.
• This method provides a novel pathway for early detection of aircraft skin risks through acoustic field analysis.
• The technique moves beyond point-wise sensing, offering a more comprehensive inspection capability.