Improved unit cells to predict anisotropic thermal conductivity of three-dimensional four-directional braided composites by Monte-Carlo method

Abstract

The analysis of 3D4d braided composites is challenging due to its complicated mesostructure. A parameterized modified three-dimensional four-directional (3D4d) braided composites unit cell model is developed in this work, along with a high-precision computation approach for it. A parametric representative volume element (RVE) method for three-dimensional braided composites is established, which is based on the coupling braiding parameters and discrete anisotropic characteristics. Additionally, a Monte-Carlo cycle framework for performing random effect verification of defects is established. To accomplish high-precision computation of effective thermal conductivity (ETC), the methodologies described in this paper include RVE unit modeling, yarns' anisotropy treatment and probability statistics on porosity can be applied to any multidirectional braided and heterogeneous materials. The ETCs of 3D4d braided composites with various fiber volume fractions, braiding angles and porosities are evaluated by the methodologies. The results demonstrate that the relative errors between the simulation results and the available experimental data are all limited within 2.8%, the thermophysical properties of 3D4d braided composites are obviously anisotropic and that the variation of ETCs with parameters exhibit quasi linear and divergent distribution characteristics. This work proposes and verifies two influence mechanisms of porosity on composites, which also provide a possibility to realize composites properties' interval estimate.