Studies on thermal conductivity prediction of fiber reinforced material with microscopic structure identification

摘要

The aim of this article is to establish an innovative thermal conductivity prediction method based on microstructure features recognition. Digital image processing technology is applied to analyze electron micrograph of typical long fiber reinforced composites (FRC). A novel practice and the criteria of determining the critical size of Representative Volume Element (RVE) are proposed. The mean value and the standard deviation of equivalent thermal conductivity could be achieved which can represent more real physical meanings due to considering the real random distribution of fibers. The thermal conductivity of a typical carbon fiber reinforced epoxy matrix was predicted by this novel method. Meanwhile laser flash experimental method was applied to get the equivalent the thermal conductivity. The result of calculation fit well with the measured value, the relative error is a 10.1%.