CT and image post-processing for fiber composites: Defect analysis, deep learning, digital volume correlation, and FE simulation – A review

X-ray computed tomography has become an indispensable technique in the multi-scale study of damage mechanisms in fiber-reinforced composites. Industrial CT and synchrotron radiation CT (SR-CT) offer complementary high-resolution imaging solutions. This review summarized the latest advancements in CT-based post-processing methods. It assesses the key factors in selecting appropriate CT equipment, weighing the nanoscale resolution of SR-CT against the practical cost-effectiveness of industrial CT. Advanced deep learning models have been proven to address the shortcomings of traditional segmentation methods when applied to low-contrast, heterogeneous composites, enabling accurate defect identification and quantification. A novel hybrid digital volume correlation technique is introduced, which combines subset-based local analysis with global-based algorithms, considering different deformation increments. Additionally, an enhanced finite element simulation framework based on CT image geometry is proposed, leveraging peridynamics simulation to overcome the limitations of crack propagation modeling. The integration of these methods makes it possible to visualize internal defect evolution in three dimensions during in-situ tensile, compressive, and fatigue tests. The review concludes the future directions, including the development of highly integrated and automated CT data processing platform, aimed at addressing the challenges in damage prediction for composites in aerospace and rail transportation.