The mixing of granular materials is an important unit operation in many industries. Due to the complex behaviors of granular flows, general laws and fundamental mechanisms of granular flows in industrial mixers are not completely understood yet. As a detailed numerical approach, the discrete element method (DEM) describes the forces and motions of granular materials at the particle scale, and thus has notable advantages over experimental approaches in the research of mixing mechanisms. With the rapid developments of its models and the computational technologies, this method becomes more and more popular in the simulations of various mixing processes. The effects of particle properties, mixer types, and operating parameters on mixing rate and mixing mechanisms could be investigated comprehensively through DEM, which would be quite valuable for the design and optimization of mixers as well as their optimal operations. Moreover, the high computational cost of industrial-scale simulations could be greatly alleviated by the fast developments of computer hardware, such as the advent of graphics processing unit (GPU). This review summarizes the recent progresses of DEM simulations on mixing, with emphasis on the treatments for non-cohesive particles in different kinds of mixers (rotary and fixed), cohesive particles (fine and wet), non-spherical particles (direct description of shape and multi-sphere method), and large-scale implementations. Finally, future development of the DEM method in mixing simulations is prospected.

The mixing of granular materials is an important unit operation in many industries. Due to the complex behaviors of granular flows, general laws and fundamental mechanisms of granular flows in industrial mixers are not completely understood yet. As a detailed numerical approach, the discrete element method (DEM) describes the forces and motions of granular materials at the particle scale, and thus has notable advantages over experimental approaches in the research of mixing mechanisms. With the rapid developments of its models and the computational technologies, this method becomes more and more popular in the simulations of various mixing processes. The effects of particle properties, mixer types, and operating parameters on mixing rate and mixing mechanisms could be investigated comprehensively through DEM, which would be quite valuable for the design and optimization of mixers as well as their optimal operations. Moreover, the high computational cost of industrial-scale simulations could be greatly alleviated by the fast developments of computer hardware, such as the advent of graphics processing unit (GPU). This review summarizes the recent progresses of DEM simulations on mixing, with emphasis on the treatments for non-cohesive particles in different kinds of mixers (rotary and fixed), cohesive particles (fine and wet), non-spherical particles (direct description of shape and multi-sphere method), and large-scale implementations. Finally, future development of the DEM method in mixing simulations is prospected.