Revealing the structure of a granular medium through ballistic sound propagation

Estimation of Particle Location in Granular Materials Based on Graph Neural Networks

Particle locations determine the whole structure of a granular system, which is crucial to understanding various anomalous behaviors in glasses and amorphous solids. How to accurately determine the coordinates of each particle in such materials within a short time has always been a challenge. In this paper, we use an improved graph convolutional neural network to estimate the particle locations in two-dimensional photoelastic granular materials purely from the knowledge of the distances for each particle, which can be estimated in advance via a distance estimation algorithm. The robustness and effectiveness of our model are verified by testing other granular systems with different disorder degrees, as well as systems with different configurations. In this study, we attempt to provide a new route to the structural information of granular systems irrelevant to dimensionality, compositions, or other material properties.

Machine learning approach to force reconstruction in photoelastic materials

Photoelastic techniques have a long tradition in both qualitative and quantitative analysis of the stresses in granular materials. Over the last two decades, computational methods for reconstructing forces between particles from their photoelastic response have been developed by many different experimental teams. Unfortunately, all of these methods are computationally expensive. This limits their use for processing extensive data sets that capture the time evolution of granular ensembles consisting of a large number of particles. In this paper, we present a novel approach to this problem that leverages the power of convolutional neural networks to recognize complex spatial patterns. The main drawback of using neural networks is that training them usually requires a large labeled data set which is hard to obtain experimentally. We show that this problem can be successfully circumvented by pretraining the networks on a large synthetic data set and then fine-tuning them on much smaller experimental data sets. Due to our current lack of experimental data, we demonstrate the potential of our method by changing the size of the considered particles which alters the exhibited photoelastic patterns more than typical experimental errors.

Dilation as a precursor in a continuous granular fault

We analyze the dilation of the system in a cylindrical granular fault consisting of one single layer of disks submitted to both normal pressure and continuous and slow shear, which results in intermittent and sudden energy release events that reproduce the main laws of seismicity. The dilation of the system can be separated into two parts: a smooth increase of dilation, plus sudden changes both contracting and dilating the medium, which are correlated to abrupt jumps -both positive and negative- in the measured resisting torque. We explain the four possible (and existing) general scenarios combining those two variables: dilation jumps and torque jumps, thanks to the assumption of an optimal local angle in the direction of force chains, and each reorganization of the structure as a replacement of the force chain holding most of the applied stress. The average rate of increase of global dilation varies monotonically with the size of the energy release event, making dilation a plausible candidate to predict catastrophic events in such earthquake-like systems.

Acoustic localisation in a two-dimensional granular medium

We focus on localizing the source of acoustic emissions within a compressed two-dimensional granular ensemble of photoelastic disks, having as main information the arrival times of the acoustic signal to 6 sensors located in the boundaries of the system. By estimating, thanks to the photoelasticity of the grains, the wave speed at every point of the structure, we are able to compute the arrival times from every point of the system to the sensors. A comparison between the arrival time differences between every set of computed values to those from the actual measurements allows finding the source of the acoustic emissions.

Continuously Sheared Granular Matter Reproduces in Detail Seismicity Laws

We introduce a shear experiment that quantitatively reproduces the main laws of seismicity. By continuously and slowly shearing a compressed monolayer of disks in a ringlike geometry, our system delivers events of frictional failures with energies following a Gutenberg-Richter law. Moreover, foreshocks and aftershocks are described by Omori laws and interevent times also follow exactly the same distribution as real earthquakes, showing the existence of memory of past events. Other features of real earthquakes qualitatively reproduced in our system are both the existence of a quiescence preceding some main shocks, as well as magnitude correlations linked to large quakes. The key ingredient of the dynamics is the nature of the force network, governing the distribution of frictional thresholds.

Effect of an interstitial fluid on the dynamics of three-dimensional granular media

The propagation of mechanical energy in granular materials has been intensively studied in recent years given the wide range of fields that have processes related to this phenomena, from geology to impact mitigation and protection of buildings and structures. In this paper, we experimentally explore the effect of an interstitial fluid on the dynamics of the propagation of a mechanical pulse in a granular packing under controlled confinement pressure. The experimental results reveal the occurrence of an elastohydrodynamic mechanism at the scale of the contacts between wet particles. We describe our results in terms of an effective medium theory, including the presence of the viscous fluid. Finally, we study the nonlinear weakening of the granular packing as a function of the amplitude of the pulses. Our observations demonstrate that the softening of the material can be impeded by adjusting the viscosity of the interstitial fluid above a threshold at which the elastohydrodynamic interaction overcomes the elastic repulsion due to the confinement.

Contact laws between nanoparticles: the elasticity of a nanopowder

Studies of the mechanical contact between nanometer-scale particles provide fundamental insights into the mechanical properties of materials and the validity of contact laws at the nanoscale which are still under debate for contact surfaces approaching atomic dimensions. Using in situ Brillouin light scattering under high pressure, we show that effective medium theories successfully predict the macroscopic sound velocities in nanopowders if one takes into account the cementation of the contacts Our measurements suggest the relevance of the continuum approach and effective medium theories to describe the contact between nanoparticles of diameters as small as 4 nm, i.e. with radii of contact of a few angstroms. In particular, we demonstrate that the mechanical properties of nanopowders strongly depend on the surface state of the nanoparticles. The presence of molecular adsorbates modifies significantly the contact laws.

River-bed armouring as a granular segregation phenomenon

River bed-load transport is a kind of dense granular flow, and such flows are known to segregate grains. While gravel-river beds typically have an “armoured” layer of coarse grains on the surface, which acts to protect finer particles underneath from erosion, the contribution of granular physics to river-bed armouring has not yet been investigated. Here we examine these connections in a laboratory river with bimodal sediment size, by tracking the motion of particles from the surface to deep inside the bed, and find that armour develops by two distinct mechanisms. Bed-load transport in the near-surface layer drives rapid, shear rate-dependent advective segregation. Creeping grains beneath the bed-load layer give rise to slow but persistent diffusion-dominated segregation. We verify these findings with a continuum phenomenological model and discrete element method simulations. Our experiments suggest that some river-bed armouring may be due to granular segregation from below—rather than fluid-driven sorting from above—while also providing new insights on the mechanics of segregation that are relevant to a wide range of granular flows.

River beds often exhibit armouring, in which formation of a coarse surface layer shields the finer underlying grains from erosion. Here, using experiments in a laboratory river and discrete and continuum models, the authors demonstrate that river-bed armouring is driven by vertical granular segregation.

Photoelastic force measurements in granular materials

Photoelastic techniques are used to make both qualitative and quantitative measurements of the forces within idealized granular materials. The method is based on placing a birefringent granular material between a pair of polarizing filters, so that each region of the material rotates the polarization of light according to the amount of local stress. In this review paper, we summarize the past work using the technique, describe the optics underlying the technique, and illustrate how it can be used to quantitatively determine the vector contact forces between particles in a 2D granular system. We provide a description of software resources available to perform this task, as well as key techniques and resources for building an experimental apparatus.

Structure of force networks in tapped particulate systems of disks and pentagons. I. Clusters and loops

The force network of a granular assembly, defined by the contact network and the corresponding contact forces, carries valuable information about the state of the packing. Simple analysis of these networks based on the distribution of force strengths is rather insensitive to the changes in preparation protocols or to the types of particles. In this and the companion paper [Kondic et al., Phys. Rev. E 93, 062903 (2016)10.1103/PhysRevE.93.062903], we consider two-dimensional simulations of tapped systems built from frictional disks and pentagons, and study the structure of the force networks of granular packings by considering network's topology as force thresholds are varied. We show that the number of clusters and loops observed in the force networks as a function of the force threshold are markedly different for disks and pentagons if the tangential contact forces are considered, whereas they are surprisingly similar for the network defined by the normal forces. In particular, the results indicate that, overall, the force network is more heterogeneous for disks than for pentagons. Such differences in network properties are expected to lead to different macroscale response of the considered systems, despite the fact that averaged measures (such as force probability density function) do not show any obvious differences. Additionally, we show that the states obtained by tapping with different intensities that display similar packing fraction are difficult to distinguish based on simple topological invariants.

Mechanical impulse propagation in a three-dimensional packing of spheres confined at constant pressure

Mechanical impulse propagation in granular media depends strongly on the imposed confinement conditions. In this work, the propagation of sound in a granular packing contained by flexible walls that enable confinement under hydrostatic pressure conditions is investigated. This configuration also allows the form of the input impulse to be controlled by means of an instrumented impact pendulum. The main characteristics of mechanical wave propagation are analyzed, and it is found that the wave speed as function of the wave amplitude of the propagating pulse obeys the predictions of the Hertz contact law. Upon increasing the confinement pressure, a continuous transition from nonlinear to linear propagation is observed. Our results show that in the low-confinement regime, the attenuation increases with an increasing impulse amplitude for nonlinear pulses, whereas it is a weak function of the confinement pressure for linear waves.