[Paper Review] Simulation nnnumérique discrète et comportement mécanique des matériaux granulaires

Jean-Noël Roux, François Chevoir|ArXiv.org|Jan 22, 2009

Granular flow and fluidized beds39 references20 citations

TL;DR

This paper presents discrete numerical simulation as a computational tool to investigate the micromechanical origins of macroscopic mechanical behavior in granular materials. By modeling grain-scale interactions using contact mechanics and simulating quasi-static deformation and dense flow, the study reveals how microstructure, friction, and dilatancy govern macroscopic responses such as strain localization and flow resistance.

ABSTRACT

As a complementary tool to laboratory experiments, discrete numerical simulation, applied to granular materials, provides valuable information on the grain and contact scale microstructure, thereby enabling one to better understand the microscopic origin of macroscopic mechanical behaviours. We first introduce different simulation methods, which we regard as techniques for numerical experimentation, in connection with micromechanical models for intergranular contacts. We lay special emphasis on the important issue of sample representativity and stress the usefulness of dimensional analysis in the definition of relevant control parameters. We then apply this approach to two important rheological regimes of granular systems: solid-like, slowly strained granular materials, which might be ruled by elastoplastic constitutive relations; and liquid-like, dense granular flows, either confined or with a free surface, described by suitable friction and dilatancy laws

Motivation & Objective

• To understand the micromechanical basis of macroscopic mechanical behavior in granular materials such as soils and powders.
• To evaluate the role of grain-scale interactions—elasticity, friction, and interfacial forces—in determining bulk material response.
• To assess the representativeness of simulated samples and the influence of geometric and contact parameter configurations.
• To investigate quasi-static deformation and dense flow behaviors in granular assemblages using numerical experiments.
• To identify key dimensionless parameters governing mechanical response through dimensional analysis.

Proposed method

• Discrete element method (DEM) simulations with spherical or circular grains in 2D or 3D to model individual grain dynamics.
• Implementation of contact laws based on linear elasticity and Coulomb friction to simulate intergranular forces.
• Use of quasi-static time integration to simulate slow deformation processes without dynamic inertia.
• Application of boundary conditions to simulate triaxial compression, shear, and inclined plane flows.
• Employment of dimensional analysis to identify dominant control parameters such as friction coefficient and coordination number.
• Analysis of microstructural variables including particle coordination, force network anisotropy, and local volume fraction.

Experimental results

Research questions

• RQ1 How do grain-scale interactions give rise to macroscopic plasticity and strain localization in granular solids?
• RQ2 What role does particle geometry and coordination number play in determining the mechanical response during quasi-static deformation?
• RQ3 How do friction and dilatancy laws govern the rheology of dense granular flows under confinement or on inclined planes?
• RQ4 To what extent can discrete simulations reproduce and explain experimental observations of granular flow and failure?
• RQ5 Which dimensionless parameters most significantly influence the macroscopic behavior of granular assemblages?

Key findings

• Discrete simulations successfully reproduce macroscopic behavior such as strain softening and hardening in granular solids, with microstructural evolution linked to force network reorganization.
• The coordination number and particle shape significantly influence the onset of plastic deformation and the development of force chains.
• Friction and dilatancy laws are critical in determining the flow resistance and velocity profiles in dense granular flows on inclined planes.
• Simulations show that the transition from solid-like to flow-like behavior is governed by a balance between frictional resistance and dilative volume changes.
• Dimensional analysis identifies the friction coefficient and coordination number as key control parameters for macroscopic rheology.
• Numerical experiments confirm that sample preparation and boundary conditions critically affect the representativeness of simulated mechanical responses.