Discrete Element Model And Numerical Simulation Of Granular Mixing Processs

Granular mixing is the most common unit operation in many industrial applications,and a deeper understanding of the kinematic law of particles is the premise to optimize and control the mixing behavior of granular materials.With the increasingly advanced computer technology,the discrete element method(DEM)has become the main tool for particle studies.However,it is still limited by the computing competence of computer for solving industrial-scale problems by DEM.In order to solve the numerical simulation bottleneck of granular mixing in engineering,the main contents of this thesis include:(1)Establishing a neighbor-searching method based on the dynamic mesh called Dynamic Mesh Cell-linked List(DMCL)and a Coarse-grain(CG)DEM based on the conservation principle of energy dissipation rate,which are suitable for granular mixing in industrial-scale applications.To save computer memory in large-scale simulations,the conventional method is modified by using dynamic mesh here,ie.,DMCL,where the domain with particles will be partitioned by smaller mesh while the domain with no particle will be partitioned by larger mesh.To verify the DMCL method,simulations for particles transported by a screw conveyor and a belt conveyor are done.The simulation results indicate that using the DMCL method can extraordinarily reduce the memory usage and is as efficient as the conventional method.Based on the conservation principle of energy dissipation rate,the CG-DEM is built here especially for dense flows dominated by contact force to improve the computational efficiency in large-scale dense flows.Both the experiment and simulation results confirm that in this CG-DEM,the CG system can reasonably predict the particle behavior in the original system.Using CG-DEM is efficient but cut back on the accuracy of granular flow,which still enables the application of DEM in industrial-scale particle researches.(2)Understanding the diffusion mechanism of both size bidisperse spheres and monodisperse ellipsoids and establishing diffusion scaling laws for them,respectively,which can be combined with the continuum model to predict the granular motion in large-scale free-surface flows.First,the simulation shear cell with Lees-Edwards boundary condition is built.Then the relation between the diffusion coefficient and the flow parameters in dense shear flows for size bidisperse spheres and monodisperse ellipsoids is systematically investigated.The simulation results indicate the diffusion coefficient is proportional to the shear rate and the square of the characteristic size,but inversely proportional to the solid fraction,for the ellipsoids,the diffusion coefficient first decrease and then increase with the increase of aspect ratio.Finally,the scaling laws of diffusion coefficient for size bidisperse spheres and monodisperse ellipsoids are built,respectively,which compare well with each other.The modified continuum model by the quantified relation can accurately predict the granular segregation in bounded heap flow,which confirms the scaling law.(3)Applying the CG-DEM to the simulation for granular mixing in the industrial-scale double-screw conical mixer,and making clear the influence of the mixer structure and the operating condition on granular mixing.The influence of the ratio of the pitch to the screw diameter,the screw diameter ratio,the rotation and sweeping speeds on mixing are systematically studied.The mixing index,energy spectra,equipment power and wear are mainly investigated for these simulations,and the CG-DEM is applied to these numerical simulations.The simulation results indicate that when the ratio of the pitch to the screw diameter,the screw diameter ratio and the rotation ratio are at specific values,the mixer performs comparably better in all aspects.In addition,the CG-DEM can reasonably predict the mixing tendency and flow pattern of the original particles,and can accurately predict the equipment power consumption and wear rate for the original system.