Non-destructive evaluation of manufacturing defects in GFRP using cross-correlation method and enhanced terahertz imaging

Composite materials are increasingly used in the aerospace and automotive industries due to their high strength-to-weight ratio and fatigue resistance. These structures may develop internal defects during manufacturing or service, compromising their integrity and safety. Thus, non-destructive defect detection is critical during production to ensure safety and reliability. This study explores the application of terahertz time-domain spectroscopy (THz-TDS) for detecting damage in different glass fiber-reinforced polymer (GFRP) manufacturing processes. A cross-correlation-based impulse response function extraction algorithm and image enhancement method are proposed. The method targets both deep and minor defects in GFRP, leveraging the correlation between the terahertz reference signal and the time-domain detection signal to extract the terahertz impulse response function via a one-dimensional iterative deconvolution algorithm. Testing on samples with pre-fabricated delamination, three-point bending damage, and drilled damage demonstrated the method’s efficacy in extracting impulse response functions of delamination defects, which aids in defect localization and improved image representation. In the THz imaging analysis of GFRP manufacturing process damage, signal alignment, windowing, and morphology positioning techniques were applied based on the extracted impulse response functions. The proposed method significantly improved the quality of THz images for defect detection in GFRP, as demonstrated by objective evaluations and comparisons. These advancements provide a robust and effective tool for the non-destructive evaluation of composite materials.