Terahertz Spectral Characteristics of Debonding Defects in Composite Insulators and Research on Imaging Detection Methods
Zhenyong Chen, Yanyan Bao, Pei Wang, Yushuo Wu, Bodong Chen, Xingyuan Chen, Tao Geng, Shuaibing LiIntroduction:
Internal debonding defects at the silicone rubber–GFRE core rod interface of composite insulators pose significant risks to power transmission systems. However, existing detection methods remain insufficient for high-sensitivity, non-contact, and quantitative characterization of sub-millimeter to millimeter-scale interfacial defects. To address this limitation, this paper develops a multimodal detection framework integrating Terahertz Time-Domain Spectroscopy (THz-TDS), terahertz imaging, and Finite-Difference Time-Domain (FDTD) numerical simulation.
Methods:
Composite insulator specimens with controlled air gap defects (0.2–1.0 mm) were fabricated and examined using THz-TDS to extract time-domain and frequency-domain characteristic parameters. Terahertz imaging experiments were conducted to obtain two-dimensional spatial maps of defect regions. The complex permittivities of silicone rubber and GFRE were experimentally measured and fitted using a fourth-order Debye dispersion model via Particle Swarm Optimization (PSO), then incorporated into an FDTD simulation model to analyze electromagnetic wave propagation within defective structures
Results:
The 0.3–1.0 THz band was identified as a sensitive characteristic range for debonding defect detection. Defective regions exhibit earlier negative peak arrival times (~17.50 ps ~23.60 ps for normal regions) and increased pulse amplitudes. Terahertz imaging at 0.421 and 0.458 THz provides intuitive visualization of defect spatial distributions with high consistency to actual defect locations. Statistical analysis across multiple measurement points confirms the repeatability of the extracted parameters.
Discussion:
The observed time-domain and frequency-domain differences between defective and normal samples can be physically attributed to the significant refractive index contrast at the “silicone rubber–air gap–GFRE” debonding interface (n_air≈1.0 vs. n_silicone rubber≈1.5–1.6), which reduces the optical path length and interfacial impedance, leading to earlier negative peak arrival times and increased pulse amplitudes in defective regions. The measurement error analysis demonstrates that the parameter differences between defective and normal regions (Δt≈6.05ps; ΔA≈0.134a.u.) substantially exceed the intra-- group measurement deviations (CV<5%), confirming the statistical robustness of the extracted features. Compared with conventional detection methods such as ultrasonic testing and infrared thermography, the proposed terahertz-based multimodal framework offers distinct advantages in terms of non-contact operation, sub-millimeter sensitivity, and simultaneous spatial visualization capability. However, the current study employs a simplified planar layered model and controlled rectangular defects, which may not fully represent the complexity of real-world debonding morphologies involving irregular geometries, material degradation, and non-uniform air gap distributions. These limitations should be addressed in future investigations to enhance the practical applicability of the proposed approach.
Conclusion:
FDTD simulation results show good qualitative agreement with experimental observations, confirming that air gap defects induce localized electric field enhancement and increased reflection. The proposed multimodal approach achieves both qualitative identification and quantitative characterization of millimeter-scale debonding defects, providing an effective non-destructive testing methodology for composite insulator condition assessment.