Modeling of frontal polymerization of carbon fiber and dicyclopentadiene woven composites with stochastic material uncertainty

Abstract

A novel finite element modeling framework is proposed for simulating dicyclopentadiene (DCPD) frontal polymerization in carbon fiber (CF) woven composites. A mesoscale CF/DCPD representative volume element (RVE) model was developed with two triggering directions, resulting in out-of-plane and in-plane DCPD polymerization. The Arrhenius equation coupled with a modified Prout-Tompkins autocatalytic model was used to evaluate the dynamic DCPD polymerization process. The rate and degree of DCPD cure were calculated by the material property uncertainties of DCPD with and without random void inclusions (3 % used as a reference in this work). The DCPD frontal polymerization in CF/DCPD composite was strongly influenced by the triggering direction. In the RVE model considered in this work, DCPD frontal polymerization was slightly faster in the in-plane (warp/weft yarn) directions than the out-of-plane (thickness) direction; polymerization occurs first in DCPD resin and is followed by CF/DCPD interface. Material property uncertainty and void inclusion had significant effects on both out-of-plane and in-plane DCPD polymerization. A large variation in DCPD material properties and the presence of void inclusions significantly delayed the rate and degree of DCPD frontal polymerization. This work provides the preliminary estimation of the frontal polymerization of DCPD-based composites and guides the structural applications of these materials.