Three-Dimensional Imaging and Numerical Reconstruction of Graphite/Epoxy Composite Microstructure Based on Ultra-High Resolution X-Ray Computed Tomography

摘要

A combined experimental and computational study aimed at high-resolution 3Drnimaging, visualization, and numerical reconstruction of fiber-reinforced polymerrnmicrostructures at the fiber length scale is presented. To this end, a sample ofrngraphite/epoxy composite was imaged at sub-micron resolution using a 3D X-rayrncomputed tomography microscope. Next, a novel segmentation algorithm wasrndeveloped, based on concepts adopted from computer vision and multi-targetrntracking, to detect and estimate, with high accuracy, the position of individual fibersrnin a volume of the imaged composite. In the current implementation, thernsegmentation algorithm was based on Global Nearest Neighbor data-associationrnarchitecture, a Kalman filter estimator, and several novel algorithms for virtualfiberrnstitching, smoothing, and overlap removal. The segmentation algorithm wasrnused on a sub-volume of the imaged composite, detecting 508 individual fibers. Thernsegmentation data were qualitatively compared to the tomographic data,rndemonstrating high accuracy of the numerical reconstruction. Moreover, the datarnwere used to quantify a) the relative distribution of individual-fiber cross sectionsrnwithin the imaged sub-volume, and b) the local fiber misorientation relative to thernglobal fiber axis. Finally, the segmentation data were converted using commerciallyrnavailable finite element (FE) software to generate a detailed FE mesh of therncomposite volume. The methodology described herein demonstrates the feasibilityrnof realizing an FE-based, virtual-testing framework for graphite/fiber composites atrnthe constituent level.