A novel multiscale semi-analytical approach for thermal properties of fuzzy fiber reinforced composites

Abstract In order to comprehensively understand the thermal conductive behaviors CFRCs with carbon nanotubes, a novel multiscale semi-analytical model is presented in this work to study the effective thermal conductivities and localized responses. Firstly, an interlayer model is established at the first level including aligned CNTs and embedded polymer matrix. At the second level of the hierarchical structure, a three-phase model is established to simulate representative unit cells of two distinct geometries for fuzzy-fiber reinforced composites (FFRCs). The locally exact homogenization theory (LEHT) is further extended to predict the effective thermal conductivities and localized fields at each level. After guaranteeing the convergence of the proposed approach, the predicted results of fuzzy-fiber reinforced nanocomposites are verified against the available simulations in the literature and finite element method (FEM). The results demonstrate that the longitudinal thermal conductivity is least affected by the CNT coating, while the transverse thermal conductivity is significantly enhanced for the CFRCs. The thermal conductive performance of FFRCs is then demonstrated by varying the CNTs’ geometric and thermal parameters. More importantly, from a top-down procedure, the localized heat-flux/temperature distributions are recovered using the semi-analytical multiscale model, to predict the possible crack initiations starting within the microstructures.