Effective thermal conductivity prediction model for vacuum insulation cores

Abstract

Accurate prediction of thermal conductivity of porous granular materials enables the identification and rapid optimisation of new composite core materials for Vacuum Insulation Panels (VIPs). To date, no computer model has reported the use of multi-sized particles with a combined Finite Element Analysis (FEA) and Discrete Element Method (DEM) approach to predict thermal conductivity of VIP core. To fill this knowledge gap, we propose a FEA and DEM based thermal conductivity prediction model for powdery composites, particularly suited for VIPs. The geometry of the model was formed using random packing of multisized spherical particles using DEM and a MATLAB PDE-based thermal model solver was used to obtain a solution for the generated geometry. The model can predict effective thermal conductivity whilst accounting for the VIP-specific fundamental heat exchange phenomena. The results from the model are validated against those obtained from accompanying thermal conductivity measurements performed using the Transient Hot Wire method, with results falling within the error range for temperatures <491.15 K. Effective thermal conductivity of a perlite and Silicon Carbide (SiC) core at 0.1 mbar over a temperature range of 303 K to 803 K, with the proportion of perlite varying from 100 % to 50 % (by weight), as predicted by the model, is presented. The thermal conductivity of the 50 % perlite-50 % SiC composite had the lowest rate of increase of thermal conductivity of 46.3 %, with the value increasing from 12.1 mW m-1 K-1 at 303 K to 17.7 mW m-1 K-1 at 803 K.