This paper addresses the challenge of reconstructing nonuniformly orientated fiber-reinforced polymer composites (FRPs) with three-dimensional (3D) geometric complexity, especially for fibers with curvatures, and proposes a framework using micro X-ray computed tomography (μXCT) images to quantify the fiber characteristics in 3D space for elastic modulus prediction. The FRP microstructure is first obtained from the μXCT images. Then, the fiber centerlines are efficiently extracted with the proposed fiber reconstruction algorithm, i.e., iterative template matching, and the 3D coordinates of the fiber centerlines are adopted for quantitative characterization of the fiber morphology. Finally, Young’s modulus is predicted using the Halpin-Tsai model and laminate analogy approach, and the fiber configuration averaging method with the consideration of the fiber morphology. The new framework is demonstrated on both injection-molded short and long carbon fiber-reinforced polymer composites, whose fiber morphology and predicted mechanical properties are validated through previous pyrolysis and quasi-static tensile tests, respectively.

Introduction

Micro X-ray computed tomography (µXCT), as a typical nondestructive imaging technique, has demonstrated its advantages to explore the detailed three-dimensional (3D) internal structure of carbon fiber-reinforced polymer (CFRP) composites including unidirectional, laminated, injectionmolded, and chopped-fiber composites [1–5]. By leveraging the variation of X-ray attenuations owing to the differences in density and atomic number, the captured microscale XCT images can unveil the composite constituents, e.g., fibers, matrix, and defects [3–8], where the high-density material (e.g., fibers) appears brighter than the low-density material (e.g., matrix). At present, µXCT is effectively used to understand the initiation and evolution of damage and to determine the in-situ fracture mechanics of CFRP composites [4,6,9–14]. However, only limited quantitative image analyses of µXCT images have been reported for non-uniformly orientated CFRP composites, especially for those consisting of curved fibers. This is because the appropriate post-image processing algorithms such as the Bayesian inference theory-based and machine learning-based approaches depend considerably on the image quality and material nature [15–17]. Emerson et al. [18] proposed a dictionary-based probabilistic segmentation technique to indicate the likelihood of a voxel belonging to a fiber or the matrix, which required the user’s inputs, including dictionary patch size and representative labeled patches to identify fiber centroid from 2D images.