Stepwise DE/FE combined approach for estimating effective thermal conductivity of frozen spherical particulate media

A stepwise discrete element-finite element (DE/FE) combined numerical framework was proposed for efficiently estimating the effective thermal conductivity of spherical particulate media, which considers the thermal interaction of particle-to-particle and particle-to-pore filling material. The particulate composition was numerically modeled by the discrete element method (DEM), which was followed by a series of heat-transfer analyses formulated in the finite element method (FEM). In order to overcome difficulty of generating meshes in the FE modeling, a novel particle-size-reduction method was developed in the course of the mesh generation. The effective thermal conductivity of glass-bead specimens was measured at the porosity of 0.32, 0.35 and 0.38 under saturated and saturated-frozen conditions. The numerical framework resulted in the accurate estimation of effective thermal conductivity of glass-bead specimens in comparison with the laboratory measurements with the errors less than 5%. In addition, to verify the applicability of the proposed numerical framework to actual geologic materials, the effective thermal conductivity of Jumunjin standard sand specimens was considered at the porosity of 0.41, 0.43 and 0.45 under saturated and saturated-frozen conditions. The estimation errors in the numerical framework for the Jumunjin standard sand were less than 3%, which reveals the applicability of the proposed method.