Improved Entropy Regularization for 3-D Imaging of Defects in Multilayer Composites

An improved entropy regularization method (IERM) has been proposed in a previous work of the authors in 2-D for overcoming a wrong localization problem, while the imaging method based on microwave time reversal (MTR) is used for detecting the defects in multilayer composites backed by metal. This method is further investigated in this article for defect detection in 3-D objects composed of multilayer composites. Its effectiveness for solving the wrong localization problem in 3-D is validated in particular via detecting a small defect in a multilayer composite skin of an aircraft wing model. Its performance and noise tolerance on 3-D imaging of small defects are first studied based on planar composites, and the influence of source polarization is also discussed in detail. Afterward, 3-D imaging of small defects in a four-layer glass-fiber-reinforced composite skin for an aircraft wing model is carried out. Different levels of noise are applied to the time-reversed signals. Three-dimensional imaging results are discussed in detail. The effectiveness and stability of the proposed IERM for 3-D nondestructive testing (NDT) of defects in multilayer composites are finally validated.