3-D Numerical Simulation On Dispersion Behavior Using Discrete Element Method

The dispersion of solid particles, such as inorganic nano-particles and polymer latexes, ina fluid is an important process for commercial products containing particulate materials. It’sof vital importance to investigate the dispersion mechanism and dynamic details ofnano-particle clusters in flow fields, which can be helpful to design efficient mixingequipments and determine fine mixing techniques to enhance the product performance.However, limited by the experimental condition, the problem was investigated mainly bycomputer numerical simulation. Numerical simulation has been one of the main tools inscientific research, as the development of computer techniques. The numerical methodsadopted by the commercial software, such as Finite Element Method and Finite VolumeMethod, can get inspiring results when solving continuum problems. While facing thedispersion process of particle clusters which are discrete medium, those numerical methodscan’t reflect the physical character of discrete medium. The Discrete Element Method(DEM)based on the discontinuous hypotheses, was originally used to analyze the rock joint. It’sparticularly suitable for solving the discontinuous problems. Afterward, DEM bloom into aneffective analysis method to study the dispersion behavior of particulate material in fluidfields.In this paper, basic principle, algorithm implementation and the programming thought ofDEM are introduced at first, and the Vander Waals force and fluid drag force models areintroduced in the basic principle of DEM. On these basises, we developed the3-D discreteelement software DEMix3D by Visual Studio2010, using VC++language. DEMix3D iscomposed of three modules: pre-processing module, solver module and post-processingmodule. The pre-processing module realizes the generation of fractal structure of clusters byDLA model, and uses the graphical user interface(GUI) to input the physical parameters,initial conditions and the control condition. The solver module realizes the calculation of theclusters’ dispersion process by iterative algorithm in each time steps. The post-processingmodule realizes the visualization of the results by using OpenGL library function, and thestatistic analysis of results. It can display the animation of the clusters’ dispersion process andthe cloud chart. This paper conducted the simulation of random packing process of hardsphere balls in a rigid box, and compared the results with the previous similar work to verifythe correctness of DEMix3D. Finally, DEMix3D is used to simulate the dispersion process ofnano-particle clusters of polystyrene in shear flow fields with different shear stress. Dynamicdetails of deformation, movement and failure of particle clusters are achieved, which provide an effective approach to study the dispersion behavior of nano-particle clusters. It’sdemonstrated that the dispersion process of clusters in flow fields first exhibit deformationand then disperse in the manner of erosion and rupture, and secondary reunion of fragmentscan be seen in the dispersion process. Conclusion can also be drawn that the clusters withhigher flow intensity have higher dispersion efficiency. The initial clusters with lower fractaldimension have higher dispersion efficiency. And the dispersion efficiency of elongationalflow is higher than the shear flow.