Micromechanical analysis for transverse thermal conductivity of composites

Abstract

Micromechanical analyses were conducted for the prediction of transverse thermal conductivity of laminated composites. We reproduced and reinvestigated both analytic and numerical models with regular and randomly distributed fibers in matrix material. A parametric study was conducted for wide ranges of fiber volume fractions and fiber-to-matrix thermal conductivity ratios. The numerical solutions using finite element (FE) analysis were compared with various analytic solutions from simple and enhanced rule or mixtures and an effective inclusion method (EIM). It was found that the EIM yields a reasonably agreeable solution with the FE solution using a hexagonal-array of regular fiber distribution for wide ranges of fiber volume fraction and fiber-to-matrix thermal conductivity ratios, which makes the EIM a useful method in predicting various multiphysical transverse properties of composites. Comparison of the results from the regular- and random-fiber models indicates that the transverse thermal conductivity of composites can significantly be affected by the random fiber distributions, especially at high fiber volume fractions. A similar conclusion was made for the foams with random pore distribution. It was shown that the predictions with the random fiber distribution agree well with the experimental data.