Semi-resolved CFD-DEM simulation of fine particle migration with heat transfer in heterogeneous porous media

Abstract

A comprehensive understanding of the migrating phenomena of microscale particles associated with fluid flows and heat exchange in porous media is essential for developing unconventional geo-resources. In this work, we further develop the kernel-based semi-resolved CFD-DEM and combine it with grain-scale reconstruction to numerically investigate particulate flows and associated heat transfer processes in porous media. To characterize wide grain size distributions and heterogeneity of rock skeleton, a 3D spherical-packed model is quantitatively assembled based on CT images of real rock. After a series of validations, it is found that the results from the improved semi-resolved CFD-DEM, including motion, heat transfer, and pressure drop behaviors of particulate systems, match better with experimental observations and analytical solutions than that obtained from the unresolved CFD-DEM. On this basis, the particle-scale migration and distribution characteristics of fine particles are further studied. It is found that the penetration distance, temperature, and energy distributions of particle systems rely on the variation of powder properties (i.e., concentration and size) and the effects of matrix anisotropy. In a higher powder concentration system, more deposited particles at the upper of the packed bed prevent the downward migration of powder particles and lower the quantity of heat exchange between the fine and matrix particles. In addition, this can also lead to a more pronounced jamming phenomenon and reduce the seepage capacity of the packed bed. As the powder diameter increases, the heat transfer capacity of fine particles increases and gradually dominates the temperature of particulate systems. The deposition patterns and temperature variation behaviors of fine particles along the migration direction caused by the anisotropy of the porous medium are also identified. Therefore, the developed computational framework can be a powerful tool to reproduce the dynamics of fine particles in heterogeneous porous materials and reveal the underlying mechanisms.