Lianfeng Liu

A simulation study of the effects of dynamic variables on the packing of spheres

The packing of uniform spheres has been studied by means of Discrete Element Method with special reference to variables affecting the packing dynamics, with the results analysed in terms of packing density, radial distribution function (RDF) and coordination number. It is shown that packing density increases with dropping height and restitution coefficient, and decreases with deposition intensity and friction coefficient, which is consistent with previous experimental findings. Both RDF and coordination number distribution vary with these variables, in line with packing density. For a packing of high density, it has a clear split second peak in its RDF, like that observed for the dense random packing. However, as packing density decreases, the first component of the split second peak will gradually vanish, giving an RDF more comparable to those observed in a sequential addition simulation. Mean coordination number can be correlated with packing density; it increases with dropping height and restitution coefficient, and decreases with deposition intensity and friction coefficient.

Dynamic simulation of the centripetal packing of mono-sized spheres

This paper presents a study of the centripetal packing of mono-sized spherical particles simulated by means of the granular dynamic or discrete element method. A packing is formed by imposing an assumed centripetal force on particles randomly generated in a spherical space. Different from the conventional simulation techniques, dynamic information of individual particles including transient forces and trajectory is traced in the present simulation. Structural properties, such as packing density, radial distribution function, coordination number distribution and homogeneity, are analyzed, with particular reference to the effects of the magnitude of the centripetal force and the number of particles. Comparison with the literature results suggests that such a dynamic model can satisfactorily simulate the dynamics of forming a packing and produce more realistic structural information. In particular, it is confirmed that a centripetal packing is not homogeneous in structure, becoming looser as its size or the number of particles increases. The packing has a limit packing density of 0.637–0.645, an overall mean coordination number of around 6.0 and a radial distribution function of clear split second peak. The centripetal force affects the rate of densification and the mean coordination number but not packing density and radial distribution function.

Simulation of microstructural evolution during isostatic compaction of monosized spheres

A numerical simulation of isostatic compaction of monosized spheres by discrete element method is presented in this paper. The dynamic microstructural changes of the particle compact during compaction were first investigated in terms of mean packing density and coordination number, and then characterized using percolation theory. The constitutive relationship between density and pressure was also examined, moreover used to validate the empirical relations obtained by experiments.

Computer simulation of the packing of particles

This paper presents a brief review of the recent progress in computer simulation of particle packing with special reference to the dynamic simulation method. The need for and the development of computer simulation are first discussed. Then, the dynamic simulation method is briefly described, followed by illustrative examples to demonstrate its applicability under different packing conditions, and the results are discussed in terms of both micro- and macroscopic properties such as porosity, radial porosity oscillation, radial distribution function and co-ordination number. It is concluded that the discrete dynamic simulation can provide realistic structural information that can be used in the quantification of structural properties of a packing of particles.

Total Lagrangian perspectives on analytical sensitivities for flexible beams

This paper is set in the context of nonlinear structural optimisation. Analytical sensitivity expressions are developed from two contrasting Total Lagrangian beam formulations. Strong geometrically nonlinear structural responses are introduced through the assumption of inextensibility of the curvilinear in one case and modification of the curvature in the second. Though performing similarly in predicting structural behaviour, the two approaches are shown to require careful consideration when extended to sensitivity determination. Conditioning of the sensitivity formulation is required to avoid contamination of the solution by even a small violation of the inextensibility constraint. An arguably more robust algorithm can be developed from the modified-curvature approach at the cost of computation efficiency. Recommendations regarding application of each approach are made.

Energy Dissipation During Impact of an Agglomerate Composed of Autoadhesive Elastic-Plastic Particles

Discrete Element Method is used to simulate the impact of agglomerates consisting of autoadhesive, elastic-plastic primary particles. In order to explain the phenomenon that the elastic agglomerate fractures but the elastic-plastic agglomerate disintegrates adjacent to the impact site for the same impact velocity, we increase the impact velocity and lower the yield strength of the constituent particles of the agglomerate. We find that increasing the impact velocity can lead to the increased number of yielded contacts, and cause the elastic-plastic agglomerate to disintegrate faster. Mostly importantly, the energy dissipation process for the elastic-plastic agglomerate impact has been investigated together with the evolutions of the yielding contacts, and evolutions of velocity during impact.