Granular Matter最新文献

The post-Darwinian era has been marked by a long-term effort to lay the foundations for a generalized theory of evolution in the broad sense. We suggest throughout this article that most of biological systems, including living species, could stand as multiscale complex systems due to microscopic or mesoscopic properties of the entity interacting with its environment. Intriguing commonalties which exist between the living and non-living species as complex systems give a strong hint that a unified approach could be developed. The paper explores this hypothesis by analyzing how complex systems, such as granular matter, evolve and adapt when brought out of equilibrium. The inherent disorder in most of granular materials gives way to a wide spectrum of structural patterns that can transform according to the external conditions applied. When brought out of equilibrium, phase transitions can occur spontaneously, leading to profound configurational reorganizations where new and unexpected structures can emerge. Using most of the fundamentals derived for granular systems, a material approach of evolution is proposed, whereby living and non-living architectures can be brought together within a rational framework whereby key concepts such as self-organization, emergence, scale effects, fluctuations and memory storage are at the very forefront.

This paper focuses on the study of a vertical spiral stirred mill, thoroughly analyzing the dynamic behavior of the grinding media within the mill barrel, aiming to achieve a comprehensive understanding of the internal operating mechanisms of this type of equipment. Firstly, based on the working principles of the vertical spiral stirred mill, a discrete element method (DEM) simulation model was constructed, and its validity was verified through experiments. Then, to explore the kinematic characteristics of the grinding media in multi-dimensional space, a refined velocity model of the grinding media was developed using vector decomposition techniques. On this basis, key control parameters such as the pitch of the spiral agitator, blade diameter, rotation speed, and grinding media filling were systematically analyzed for their effects on the motion patterns of the grinding media, relying on the validated DEM model. The results indicate that in the axial dimension, the axial velocity of the grinding media, along with the circumferential velocity in the central region of the mill, exhibits high stability, revealing the uniformity of the motion state in this region. Simultaneously, in the radial region between the outer edge of the spiral blades and the mill wall, the grinding media present significant gradients in both circumferential and axial velocities, indicating this area as a crucial grinding zone. Further analysis shows that the pitch of the spiral agitator, blade diameter, and rotation speed significantly affect the circumferential velocity in the radial direction, while both blade diameter and rotation speed also play a dominant role in the axial velocity. In contrast, the filling of the grinding media has a minimal effect on the overall motion patterns, suggesting that the dynamic characteristics of the grinding media are primarily influenced by the mechanical structure design and operational parameters.

This study presents the first experimental investigation of stress wave propagation in 1D granular chains of closed-cell PVC foam disks. Average impact velocities for H130 and H250 foams ranged from 17.6 to 38.1 m/s. The analysis focuses solely on the incident stress wave, excluding the reflected wave. The mid-planes of the disks were chosen for analysis due to their uniaxial force components along the chain's length. The results show that the stress wave speed is faster in the H250 foam chain due to its higher stiffness. Wave speed increases with impact velocity but decreases as it travels along the chain, with a more pronounced reduction in the H130 foam compared to the H250 foam. The peak normal forces in the H250 foam chain disks are approximately three times greater than those observed in the H130 foam chain disks at comparable impact velocities. The peak normal forces in both foam chains decrease rapidly with increasing impact velocity, especially over the first few disks. As the wave propagates further from the impact source, the attenuation rate slows, with a more gradual force reduction in the H250 foam due to its higher density and stiffness. Energy loss is governed by viscoelastic and plastic dissipation at disk contacts, which becomes more significant at higher impact velocities. This study provides new insight into dissipative wave phenomena in granular systems of deformable elements and offers experimental data for future modeling of strongly nonlinear, dissipative granular media.

Recent advances in image-based particle shape characterization allow reliably and rapidly determining particle roundness and sphericity of a statistically significant large number of particles, which enables systematic investigation of the influence of roundness and sphericity on macroscopic engineering behaviors such as strength and dilatancy of sands. This study collects 22 sands with a wide range of particle sphericity, roundness, gradations, and mean particle sizes. A total of 207 direct shear tests are prepared at various relative densities and normal stresses to establish the database. This database is further augmented by experimental data of another 97 sands from published geotechnical engineering sources. Influences of image-based sphericity, roundness, and gradation on the frictional and dilational components of soil strength are analyzed, leading to observations that angular, elongated, and well-graded sands exhibit larger values of critical strength, dilatancy, and peak strength. A material parameter is proposed by integrating roundness and gradation that captures the joint effects of intrinsic properties. The material parameter is used to develop predictive models for critical friction angles, dilation angles, and peak friction angles. The effectiveness and accuracy of the predicted models are validated by various published geotechnical experimental data. The material parameter and predictive models provide insights into relationships between micro particle level properties and macro mechanical behavior of sands and enable researchers and practitioners to rapidly estimate the strength and dilatancy of sands without performing laboratory tests.

Graphical Abstract

An innovative methodology for predicting gradation stability using integrated machine learning technologies is introduced. Current geometric criteria for suffusion assessment rely on a limited set of characteristic particle sizes, which results in a loss of detailed gradation information embedded in grading curves. This study proposes a new framework for evaluating the suffusion sensitivity through predicting the gradation stability of granular soil with a specified grading curve. Two distinct integrated machine learning models are developed to quantitatively assess soil internal stability. The predicted results and performance analysis demonstrate that the PCA-SVM model achieves superior classification accuracy for internal stability, while the PCA-ANN exhibits enhanced predictive capability in estimating the probability of internal stability within the given dataset. The proposed methodology provides a novel application for investigating the relationship between gradation characteristics and stability. This study will facilitate further research on establishing the accurate gradation stability criteria and predicting the soil suffusion sensitivity.

Graphical Abstract

Cemented granular materials, as unique granular substances possessing both permeability and load-bearing characteristics, have found extensive applications in chemical catalysis and geological engineering, and other fields. Given the significant impact of skeleton particle shape on the mechanical properties of cemented granular materials, this paper proposes a bonded polyhedral discrete element method adaptable to arbitrary skeleton particle shapes. Within this method, the adhesive surface is constructed from the contact geometry, and the interaction between particles of different shapes is described by employing an energy-conserving contact model based on strain energy density. The spring-damping model and bilinear constitutive model are utilized to characterize the elastic behavior and damage fracture behavior of cement, respectively. Moreover, the influence of skeleton particle shape on cemented granular materials is elucidated through both mesoscopic and macroscopic analyses using the proposed model. Mesoscopic results indicate that the area of the adhesive surface is a critical factor influencing the destructive force of bonding units. Variations in particle shape cause particles with identical volume and density to form adhesive surfaces with differing shapes and areas under the same conditions, leading to varied destructive forces in the bonding units. The macroscopic results reveal that both the sphericity and aspect ratio of the skeleton particles impact the strength of the cemented granular material. This effect predominantly arises from the differences in the coordination number of the accumulation bodies formed by skeleton particles of varying shapes.

Graphical Abstract

Particle size and shape are critical for characterizing gravel soils, typically quantified through the analysis of two-dimensional (2D) images and three-dimensional (3D) models. However, obtaining 3D particle features can be challenging and time-consuming, often resulting in low efficiency. To address this, many researchers have recently attempted to estimate 3D morphological features from 2D data by establishing relationships between 2D image features and their 3D counterparts. Nevertheless, these methods generally focus on specific particle categories within a limited region, limiting their broader applicability. In response, this study proposes a method for acquiring extensive morphological data for gravelly soils in both 2D and 3D formats through multiple collections. Additionally, it introduces and validates a practical approach for deriving 3D information from 2D image analysis, offering a series of new equations that are compared with previously published models. The result demonstrates that 3D morphological features, including particle size and shape, can be effectively estimated from 2D data using linear and polynomial correlation equations.

This paper examines the simultaneous impact of strain rate, aggregate fragmentation, and free water on the dynamic behavior of concrete in mesoscale uniaxial compression conditions. A concrete specimen measuring 50 × 50 mm² and having a low porosity of 5% was the subject of extensive two-dimensional (2D) dynamic investigations (that is, a research limitation). Its mesostructure was based on laboratory micro-CT images. Concrete’s fracture patterns, strength, brittleness, and fluid pressure distributions were all investigated. A mesoscopic pore-scale hydro-mechanical model based on a unique fully coupled DEM/CFD technique with breakable aggregate particles was utilized to study the behavior of partially or fully saturated concrete. A four-phase material comprising aggregate, mortar, ITZs, and macropores was used to replicate concrete. Groups of small spherical particles were used to simulate the fragmentation of aggregate particles with various shapes and sizes, allowing for intra-granular fracturing among them. A network of fluid channels was assumed in a continuous region between discrete elements. A two-phase laminar compressible fluid flow (air and water) in pores and cracks was suggested for wet concrete. The accurate volumes of pores and cracks were computed for tracking the liquid/gas content. Dynamic numerical compressive tests were performed with strain rates ranging between 1 1/s and 1000 1/s. Strain rate, aggregate fragmentation, and free water flow increased the dynamic compressive strength. Because of free water confinement in pores and cracks, the pore fluid pressures retarded a fracture process, enhancing the concrete dynamic strength.

Asteroid impacts are a significant area of study in astronomy; however, the specific impact properties of binary systems at close distances have not been extensively explored. This study employs a dual approach, combining low-velocity impact experiments with numerical simulations, to investigate the dynamic characteristics of binary projectiles at varying separation distances. Special focus is placed on how the initial separation distance affects the repulsion and attraction phenomena of the projectiles within granular media. Empirical evidence shows that smaller initial separation distances lead to significant repulsion between projectiles upon impact. Once a specific separation distance is reached, the binary projectiles exhibit attractive behavior post-impact. Quantitative simulations clarify the observed repulsive and attractive phenomena by considering the force chain, thereby providing a deeper understanding of the dynamic impact process.

Graphical Abstract

The strength-displacement behaviour of soil-structure interface should be carefully considered during slope stabilisation using soil nail. Experimental evidences have demonstrated that tangential displacement between the soil and structure could deteriorate the microstructure within the interface, resulting in a strength degradation of the soil-structure system. To capture such responses, an elastoplastic model is developed by adopting particle probabilistic entropy to characterise the evolution of particle rearrangement within the interface, where a microstructure-dependent plastic flow rule and a kinematic hardening law are proposed. The capability of the model is verified by simulating a series of interface direct shear tests, where the normal-dilatancy response and strain softening of the interface under low normal stress as well as the distinct normal-contraction under cyclic loads are well simulated. Then, the model is further implemented through FRIC subroutines for finite element (FE) simulation of the pull-out tests on a soil–nail under different overburden pressures. It is found that the FE model can reasonably simulate the pull-out behaviour of a soil nail. The stress and strain fields around the soil nail as well as the pull-out force and displacement response can be reproduced.

Graphical Abstract