In a unidirectional polymer-matrix composite under axial tensile loading, the failure process is governed by the dynamic localization and clustering of multiple fiber breaks. The composite fails catastrophically once a critical cluster size (10-30 fiber breaks) is reached. The brittle failure of a fiber is a locally dynamic process which leads to stress concentrations in the interface, matrix and the neighboring fibers that can propagate at high speed over long distances.
To study this local dynamic event, a fiber-level finite element (FE) model of a 3-dimensional array of S2-glass fibers embedded in an epoxy matrix with rate-dependent interfacial cohesive traction law will be used. A statistical distribution of defects will be implemented within the fibers by assigning periodic ‘defect planes’ represented by cohesive traction-separation surfaces. The representative spacing of these defect planes in the fiber will be determined based on a combination of the results from single fiber fragmentation experiments and a novel continuous fiber bending experiment where all cross-sections along the length of the fiber are subjected to progressively reducing radii of curvature.
The rate-dependent properties of the matrix will be implemented using a User-defined Material (UMAT) constitutive model in ABAQUS. The rate-dependent interfacial traction law has been determined from FE modeling of micro-droplet experiments spanning six-decades of strain rates on this material system (Tamrakar, PhD Thesis, 2018). The effects of thermal residual stresses and frictional sliding of the debonded fibers along the interface will also be incorporated in the 3D model. In addition to providing a more comprehensive understanding of micro-mechanical damage mechanisms within a composite, our integrative experimental and modeling approach also gives us the opportunity to tailor the matrix and interphase properties to maximize energy absorption under high loading rates.