A Signal-Based Data Fusion Approach in Non-DestructiveTesting of Composite Materials in the Aerospace Industry

Non-Destructive Testing (NDT) plays a crucial role in the aerospace industry by ensuring aircraft structural integrity and safety, while also supporting operational efficiency. Aircraft components are subject to extreme conditions and stresses, making routine inspections vital to prevent failures. Traditional NDT techniques—such as ultrasonic testing, radiography, and eddy current testing—have long been relied upon in aerospace maintenance, providing valuable insights into material integrity. However, these methods have limitations. Ultrasonic testing effectively detects internal flaws but depends on material acoustic properties and requires coupling media, limiting its applications. Radiography offers high-resolution imaging yet involves ionizing radiation and is constrained by material thickness. Eddy current testing is efficient for surface flaws but is limited by the material's conductivity and permeability. Recent advancements in NDT, including thermography and shearography, have introduced enhanced capabilities. Thermography leverages infrared imaging to identify surface temperature variations indicative of subsurface defects like delamination and corrosion, allowing for rapid, non-contact inspection of large areas in real-time. However, it is sensitive to environmental conditions and material emissivity. Shearography, an optical technique that assesses surface deformation under stress, reveals defects by analyzing interference patterns from laser light, particularly useful for detecting cracks and voids. Despite its sensitivity to stress-induced anomalies in composites, it requires a stable environment due to susceptibility to vibrations. This study explores signal-based data fusion, integrating traditional and advanced NDT techniques to enhance defect detection and characterization in aerospace materials. By combining multiple signals, data fusion and deep learning improve NDT reliability, offering a comprehensive understanding of material integrity. This approach enables detection of various defect types, assists inspectors by reducing noise and false detections, and enhances precision through combined signals. Our findings highlight that data fusion markedly improves NDT effectiveness in complex, multi-layered composite aerospace materials, raising industry standards in both efficiency and accuracy.