Tao Zhang

Multi-material structures such as carbon fiber reinforced polymer (CFRP) and epoxy resin are increasingly used in modern power equipment. However, significant differences in their thermophysical properties result in distinct defect thermal responses, which can reduce the reliability of infrared thermography inspections. To address this issue, this study investigates the thermal response mechanisms and quantitative analysis of defects in multi-material power equipment through finite element simulation and experimental validation. Three-dimensional transient heat transfer models containing air voids and heterogeneous insert defects were established using COMSOL Multiphysics for both carbon fiber reinforced polymer and epoxy resin matrices. Pulsed infrared thermography experiments were subsequently conducted to verify the simulation results. The effects of material properties, defect geometry, and cover-layer thickness on thermal response characteristics were systematically analyzed. The results show that thermal diffusivity is the key factor governing defect signal evolution. Carbon fiber reinforced polymer exhibits rapid thermal propagation and early transient responses, whereas epoxy resin produces delayed and slowly increasing thermal signals. Greater defect depth weakens thermal contrast and delays peak response time, while larger defect diameters enhance defect detectability. Increasing cover-layer thickness significantly attenuates defect signals and reduces imaging contrast. Experimental results are in good agreement with simulation predictions, confirming the validity of the proposed models. This work provides a quantitative analysis of defect thermal behavior in multi-material systems and offers a theoretical basis for adaptive infrared thermography inspection and condition assessment of power equipment. It should be noted that this study focuses on mechanistic understanding and parametric analysis rather than on proposing a dedicated quantitative defect-sizing or inversion method.

Multi-material structures such as carbon fiber reinforced polymer (CFRP) and epoxy resin are increasingly used in modern power equipment. However, significant differences in their thermophysical properties result in distinct defect thermal responses, which can reduce the reliability of infrared thermography inspections. To address this issue, this study investigates the thermal response mechanisms and quantitative characterization of defects in multi-material power equipment through finite element simulation and experimental validation. Three-dimensional transient heat transfer models containing air-void and heterogeneous insert defects were established using COMSOL Multiphysics for both CFRP and epoxy resin matrices. Pulsed infrared thermography experiments were subsequently conducted to verify the simulation results. The effects of material properties, defect geometry, and cover-layer thickness on thermal response characteristics were systematically analyzed. The results show that thermal diffusivity is the key factor governing defect signal evolution. CFRP exhibits rapid thermal propagation and early transient responses, whereas epoxy resin produces delayed and slowly increasing thermal signals. Greater defect depth weakens thermal contrast and delays peak response time, while larger defect diameters enhance defect detectability. In-creasing cover-layer thickness significantly attenuates defect signals and reduces imaging contrast. Experimental results are in good agreement with simulation predictions, confirming the validity of the proposed models. This work provides a quantitative understanding of defect thermal behavior in multi-material systems and offers a theoretical basis for adaptive infrared thermography inspection and condition assessment of power equipment.

In wind energy generation, monitoring the health of wind turbine blades is essential for ensuring efficient energy conversion and system safety. However, the complex operating environment and diverse failure modes of wind turbine blades pose challenges for traditional monitoring methods, which often struggle with data scarcity and feature extraction, leading to the poor fault classification performance. Consequently, this study introduces an innovative approach that combines a variational autoencoder (VAE) model with adaptive dimensionality selection and perceptual loss. By employing cumulative explained variance ratio (CVR) based on dimensionality reduction techniques and rate-distortion theory from information theory, the model’s latent space dimension is dynamically optimized using a threshold $v_{\text {th}}$ . Furthermore, the model incorporates perceptual loss based on multilayer features of the VGG network to enhance the semantic consistency of defect edges and textures. Experimental results demonstrate that the proposed framework achieves a remarkable accuracy of 98.23% $\pm ~0.35$ % in fault classification tasks while maintaining a lightweight architecture and producing high-quality reconstructions, and compared with the initial model, PSNR and SSIM were improved by 36.24% and 30.32%. Even in simulated complex scenarios, the model’s classification accuracy exceeds 90%. Ablation studies highlight the critical role of perceptual loss in improving the feature extraction. This work offers an efficient and cost-effective solution for wind turbine blade maintenance and supports the intelligent transformation of wind energy operations.