Granular Matter最新文献

We demonstrate for the first time that a lateral impulse experienced by a granular channel can induce an inertial bulk dilation over long distances across a granular medium with a mechanically free surface. The surface dilation requires zero overburden pressure (exposure to vacuum) and is precipitated by the passing of waves traveling barely above the sound speed (> Mach 1.05). We simulate this phenomenon using open source Soft Sphere Discrete Element Method software. We prepare channels of monodisperse, cohesive spherical particles exposed to vacuum and modeled as Hertzian springs. We validate our model by recreating acoustic wave, strong shock, and shear dilation behavior. We then create shocks within the channel to determine the sensitivity of surface dilation to wave speed, wave type, initial packing fraction, and boundary effects. The shocks we create undergo a rapid decay in strength and appear to propagate as solitary waves that can be sustained across the channel. We find that an inertial surface dilation is induced by compressive solitary waves, is insensitive to channel length, decreases with bed height, and increases substantially with initial packing fraction. A hard subsurface floor is required to maintain this wave over the entire channel. Free surface dilation induced by laterally propagating impulse loading could be implicated in the formation of Lunar Cold Spots, distal regions of low thermal inertia surrounding young craters on the Moon.

Multi-layer rigid netting barriers (RNBs) can obstruct the granular flow layer by layer, and all the layers of the structures share the impact load, which has better reliability than single-layer structure. However, the granular flow obstructed by multi-layer RNBs is complicated. The grains between the layers of RNBs may make the forces acting on all structures related to each other. Under the limited testing conditions, it is difficult to obtain several important information such as the obstruction efficiency of structural barriers and the impact forces acting on the RNBs at all layers. In this study, the discrete element method is used to numerically simulate a typical granular flow experiment. Based on the numerical verification, the method was used to study the mechanical characteristics of the RNBs at all layers and the typical process of granular flow under different settings. The results show that the numerical calculation can not only simulate the rebound, run-up, splash, passing-through and other movements of grains during the obstruction process, but also obtain the mechanical characteristics of the RNBs at all layers that are related to each other caused by the grains between the RNBs, and the final deposition of grains in front of the RNBs. The mesh and position settings of the protective structures have a significant influence on the forces acting on the RNBs at all layers, so adjusting the RNB settings through numerical optimization can make the forces on RNBs more reasonable and optimize the design of the protective structures. At the same time, the grain segregation characteristics in front of the RNBs obtained by the numerical simulation can provide a basis for further research on the physical and mechanical characteristics and the stability of the deposition.

In granular flows, grains exhibit heterogeneous dynamics featuring large distributions of forces and velocities. Conventional statistical methods have previously revealed how these dynamical properties scale with the grain size in monodisperse flows. We explore here whether they differ between small and large grains in bi-disperse flows. In simulated silo flows comprised of dense and collisional zones, we use a machine learning classifier to attempt to distinguish small from large grains based on features such as velocity, acceleration and force. Results show that a classification based on grain velocity is not possible, which suggests that large and small grains feature statistically similar velocities. In the dense zones, classification based on force only fails too, indicating that small and large grains are subjected to similar forces. However, classification based on force and acceleration succeeds. This indicates that the classifier is sensitive to the correlation between forces and acceleration, i.e. Newton’s second law, and can thus detect differences in grain size via their mass. These results highlight the potential for machine learning to assist with better understanding the behaviour of granular flows and similar disordered fluids.

The foundation of a lunar habitation may be either smooth or rough with different shapes in the light that various concepts of the lunar base have been proposed, which require a good understanding of the bearing behavior of the lunar soil under plates with different shapes and roughness. Therefore, the three-dimensional distinct element method (3D DEM) is employed here to perform plate load tests on lunar soil simulant using the force-driven method. The soil failure mechanism under different plates is first described at various scales with detailed DEM studies of the load-settlement curve, stress path, ground heave, void ratio changes, and normalized-velocity field. Following these, the shape factor and coefficient of plate roughness are discussed by comparing the DEM and theoretical results. The results show that a typical general failure model can be identified for the strip plate, a local failure model for the circular and square plates, and a Hill model for the smooth plate from the ground heave, void ratio changes, and normalized velocity field. The shape factors for bearing capacity determined by the settlement criterion are close to those by the Terzaghi method with an error of nearly 10%, and the shape factors for the deformation modulus are similar to those in Chinese standard with an error of nearly 20%. In addition, the coefficient of plate roughness for the semi-rough plate is close to those predicted by the Meyerhof and Kumar methods with an error of nearly 5%.

We investigate the effects of morphological features on the angle of repose ((Phi)) of granular materials using a three-dimensional discrete element method (3D-DEM). A commercially available DEM package called particle flow code (PFC3D) was used. The elliptical clumped particles were imported in the form of an STL file, which was modelled by controlling two morphological descriptors (i.e., β and ξ) using a multi-sphere (MS) approach. The β represents the maximum pebble-pebble intersection angle and ξ defines the ratio of smallest to largest pebble diameter within a clump. The results concluded show that the (Phi) increase with an increase in static friction coefficient became insignificant after µ_s = 0.8, whereas the (Phi) decrease with a decrease in the number of pebbles. In addition, the microscopic parameter in terms of coordination number (Z) shows an initial concave response and stabilized with a further subsequent increase in the number of pebbles. These results are of particular interest which could provide insights into the microscopic level of interactions at particulate levels of granular materials.

Producing a consistent layer quality for different raw-materials is a challenge for powder-based additive manufacturing. Interparticle cohesion plays a key role on the powder spreading process. In this work, we characterise the structure of deposited layers in the powder-base additive manufacturing process by numerical simulations using the discrete element method. The effect of particle cohesion on the quality of powder layers is evaluated. It is found that higher interparticle cohesion lead to poor spreadability, with more heterogeneous powder layer structure and enhances particle size segregation in the powder layer. We also compare the powder layer quality deposited on a smooth substrate with that on a powder layer. Deposition on a powder layer leads to inferior layer quality of powder layer with higher heterogeneity and higher particle size segregation effects.

Graphical abstract

A theoretical study on the hydraulic conductivity of fully saturated anisotropic granular materials for a 2D fluid flow has been made by making use of the microstructure tensor as the anisotropy descriptor. The assemblage of particles was assumed to be the representative elementary volume of materials with void spaces as a multiply-connected continuum through which a Stokesian flow can pass. The Navier–Stokes equations have been then solved to find the mean velocity vector under different pressure boundary conditions. A tensorial form of the hydraulic conductivity with constants being functions of the invariants of the microstructure tensor, as the geometric measure of the anisotropy, has been presented based on a number of realizations for different GSD curves. Verifications with available experimental data exhibit a reasonable accuracy of the suggested equation.

Graphical abstract

Modern particle size statistics uses many different statistical distributions, but these distributions are empirical approximations for theoretically unknown relationships. This also holds true for the famous RRSB (Rosin-Rammler-Sperling-Bennett) distribution. Based on the compound Poisson process, this paper introduces a simple stochastic model that leads to a general product form of particle mass distributions. The beauty of this product form is that its two factors characterize separately the two main components of samples of particles, namely, individual particle masses and total particle number. The RRSB distribution belongs to the class of distributions following the new model. Its simple product form can be a starting point for developing new particle mass distributions. The model is applied to the statistical analysis of samples of blast-produced fragments measured by hand, which enables a precise investigation of the mass-size relationship. This model-based analysis leads to plausible estimates of the mass and size factors and helps to understand the influence of blasting conditions on fragment-mass distributions.