Simulation of nonlinear effective thermal conductivity using CAE Fidesys

This article is devoted to the numerical estimation of the effective thermal conductivity coefficients of heterogeneous media. An estimation algorithm is described, modified for the non-linear case (when the thermal properties of the material components depend on the temperature). Using the Fidesys CAE system, linear calculations are performed to estimate the effective thermal conductivity coefficients of construction heat-insulating composites made of polymer reinforced with hollow glass microspheres. The cases of simple cubic packing of filler particles and their random distribution in the matrix are investigated. The modified algorithm is implemented as a research software module for solving non-linear problems of estimating effective thermal conductivity. Using it, model problems of estimating non-linear effective thermal conductivity for orthogonally reinforced and particle-reinforced composites are solved. The configuration of an orthogonally reinforced composite is revealed, the effective thermal conductivity of which does not depend on temperature. A significantly non-monotonic dependence of the effective thermal conductivity on temperature is obtained for a particle-reinforced composite.