Defect detection in CFRP cables via improved ART and dimension reduction with electromagnetic tomography

Defect imaging of carbon fibre reinforced polymer cables is a high-dimensional inverse problem. High-quality reconstruction of defect images is constrained by limited induced voltage measurements, leading to severe underdeterminationand ill-posedness. To address this issue, an image reconstruction method combining improved algebraic reconstruction technique and dimension reduction is proposed. First, a weighted truncated singular value decomposition strategy is applied to reduce the dimensionality of the sensitivity matrix, which compresses the solution space for defect imaging. Second, momentum and regularisation terms are introduced into the algebraic reconstruction technique iteration to accelerate convergence and suppress noise amplification. Finally, a non-negative pixel constraint is implemented for the purpose of mitigating edge artefacts in the reconstructed image. Simulation and experimental studies on carbon fiber reinforced polymer cable specimens with single and double defects are conducted to reconstruct images and evaluate the proposed method. It is shown that compared with Tikhonov regularisation, Landweber iteration, Iterative Shrinkage-Thresholding Algorithm, and algebraic reconstruction technique, the proposed method demonstrates superior performance in both Structural Similarity Index Measure and Peak Signal-to-Noise Ratio, while significantly reducing reconstruction time. This study proposes a novel non-destructive approach for rapid, high-quality defect imaging in carbon fiber reinforced polymer cables.