Multiscale modeling for mechanical property estimation of polymer matrix composites with nano- and micro-scale reinforcements

The current work proposes a combined atomistic-continuum numerical modeling to analyze the constitutive response of a novel polymer composite consisting of multiscale reinforcement. This composite consists of unidirectional carbon fibers in a polymer matrix along with two nanoscale reinforcements: (i) a “fuzzy” interphase consisting of radially aligned carbon nanotubes (CNT) around the carbon fiber and (ii) additional CNTs randomly dispersed in the polymer matrix. While radially aligned CNTs on fiber surface improve the fiber-matrix interfacial bonding, randomly dispersed CNTs enhance the matrix toughness. This special arrangement at multiple length scales demands a multiscale homogenization procedure, which is developed in this work. At the atomistic length scale, CNT embedded in polymer matrix is modeled using molecular dynamics (MD) simulations. The effective elastic properties of CNT dispersed matrix and fuzzy interphase are obtained using appropriate unit cells with periodic boundary conditions (PBCs). At the continuum length scale, the composite material microstructure is modeled as a repeating unit cell (RUC) with the elastic properties of CNTs derived from the atomistic scale. The novelty of this work lies in developing a three-phase RUC with PBCs to analyze this unique hybrid composite. Extensive parametric studies have been conducted to assess the influence of CNT content in the matrix and interphase, and volume fraction of the primary fiber reinforcement on the effective elastic properties. Results indicate a 60%–80% enhancement of transverse elastic properties with CNT reinforcement in addition to fiber reinforcement.