Two-Scale Prediction of Effective Thermal Conductivity of 3D Braided C/C Composites Considering Void Defects by Asymptotic Homogenization Method

Abstract

In order to predict the effective thermal conductivities of three-dimensional (3D) braided carbon/carbon (C/C) composites with randomly distributed void defects. Two-scale prediction model is developed based on the asymptotic homogenization method. Unit cell models both on fiber-scale and fiber bundle-scale are established according to the scanning electron microscopy observation of the material, and the randomly distributed void defects are considered. The effective thermal conductivities of fiber bundles with void defects are predicted firstly, then the effective thermal conductivities of the 3D braided C/C composites are predicted considering void defects in matrix pocket and interface by introducing the predicted thermal conductivities of fiber bundles. The predicted effective thermal conductivities agree well with the experimental results, demonstrating the validity of the two-scale prediction model. A parametric study is then conducted to analyze the effects of void volume fraction and interfacial thermal conductivity on the predictions of the developed model. The results show that the random distribution of void defects has a little effect on the effective thermal conductivities, while the void volume fraction has a significant effect on the effective thermal conductivities. The thermal conductivities decrease generally linearly with the increase of void volume fractions, and the effect of void volume fraction of matrix pocket is greater than that of fiber reinforcement. The effective thermal conductivities increase with the increase of interfacial thermal conductivity, and the effect of void volume fraction of interface becomes larger with the increase of interfacial thermal conductivity. A higher interfacial thermal conductivity have a greater effect on the effective thermal conductivities of the material than a smaller interfacial thermal conductivity.