EN FR
GeoConferences.ca

DEM simulation with elliptical particles for micro-level geotechnics simulations

Mohamed Chekired, Varvara Roubtsova, Elham Kheradmand Nezhad, Serge Prudhomme, Marc Laforest

In the proceedings of: GeoOttawa 2017: 70th Canadian Geotechnical Conference; 12th joint with IAH-CNC

Session: Soil Mechanics - Numerical Methods

ABSTRACT: Soil stability under a water flow is a complex problem in geotechnical sciences. Physical phenomena such as liquefaction, and internal and external erosion are the result of soil particles interacting with fluid. Virtual laboratory SiGran has been developed at IREQ (Institut de recherche d™Hydro-Québec) to provide key microscopic information about these interactions and gain a better insight of granular media. Thanks to recent advances in numerical methods, particle interactions can now be simulated with high precision. This makes it possible to determine the contributions of the different processesŠelastic particle interaction, interparticle friction, flow drag force, pore-pressure increases and so onŠby quantifying the absorbed energy of each process at the micro level. The proposed virtual approach uses a microscale coupling of the Discrete Element Method (DEM) and the Marker and Cell (MAC) method for fluid-particle interactions. To date, particles in SiGran have been represented by spheres. One obvious advantage of using spheres is that it is straightforward to find the contact properties between two of them. Unfortunately, spheres are different from real soil particles in terms of shape, position of their centers of mass with respect to contact points, and hence, the manner by which one rotates around the other. The present work is a first implementation of ellipsoidal particles into SiGran code, which is tested on a numerical example showing that the energy balance during the interaction of several particles is preserved. An example of oscillating ellipsoidal particles within a fluid is also presented to illustrate the evolution of the streamlines pattern around ellipsoidal particles.

RÉSUMÉ: La stabilité du sol sous l™effet d™un écoulement constitue un problème complexe en géotechnique. Les phénomènes physiques tels que la liquéfaction et l'érosion interne et externe sont le résultat de l'interaction des particules du sol avec le fluide. Le laboratoire virtuel SiGran a été développé à l'IREQ (Institut de recherche d'Hydro-Québec) pour fournir des informations clés à l™échelle microscopique sur ces interactions et obtenir une meilleure connaissance des milieux granulaires. Grâce aux progrès récents dans les méthodes numériques, les interactions des particules peuvent maintenant être simulées avec une grande précision. Cela permet de déterminer les contributions des différents processus tels que l™interaction élastique des particules, la friction interparticulaire, la force de traction d'écoulement, l™augmentation de pression des pores, etc. - en quantifiant l'énergie absorbée de chacun de ces processus au niveau micro. L'approche virtuelle proposée utilise un couplage à micro-échelle de la Méthode d'élément discrète (DEM) et de la méthode Marker and Cell (MAC) pour les interactions fluides-particules. À ce jour, les particules ont été représentées par des sphères dans SiGran. Un avantage évident d'utiliser des sphères est qu'il est simple de trouver les propriétés de contact entre deux d'entre elles. Malheureusement, les sphères sont différentes des particules réelles du sol en termes de forme, de position de leurs centres de masse par rapport aux points de contact, et donc de la façon dont l™une tourne autour de l'autre. Le présent travail est une première mise en œuvre de particules ellipsoïdales dans le code SiGran, qui est testé sur un exemple numérique montrant que le bilan énergétique lors de l'interaction de plusieurs particules est préservé. Un exemple de particules ellipsoïdales oscillantes dans un fluide est également présenté pour illustrer l'évolution des lignes de courant autour des particules ellipsoïdales. 1. INTRODUCTION Studying the physical characteristics of soil becomes necessary in order to gain a better understanding of civil engineering problems. Significant advances in numerical modeling methods in recent decades provide a unique opportunity to simulate soil behavior at very fine scales. Soil is an assembly of particles where the space between particles is partially or completely filled with water. The Discrete Element Method (DEM) describes the motion of particle assemblies and has been widely used as a mathematical tool for studying their behavior. DEM was introduced by Cundall (1971) to analyze rock mechanics and was applied to soils by Cundall and Strack (1979). Since then, the method has been extended to include new features. Originally, the authors proposed an algorithm for two-dimensional simulations, but since then efficient programs for three-dimensional simulations have been developed by Langston et al. (1997). The linear spring model for normal and tangential forces was replaced by a model based on Hertz (1882) contact theory for normal direction and Mindlin (1949) contact theory for tangential direction. Rolling resistance due to elastic hysteresis losses or viscous dissipation was studied by Zhou et al. (2002). Particles in DEM are usually represented as discs and spheres, in two-dimensional and three-dimensional

Please include this code when submitting a data update: GEO2017_600

Access this article:
Canadian Geotechnical Society members can access to this article, along with all other Canadian Geotechnical Conference proceedings, in the Member Area. Conference proceedings are also available in many libraries.

Cite this article:
Mohamed Chekired; Varvara Roubtsova; Elham Kheradmand Nezhad; Serge Prudhomme; Marc Laforest (2017) DEM simulation with elliptical particles for micro-level geotechnics simulations in GEO2017. Ottawa, Ontario: Canadian Geotechnical Society.

@article{geo2017Paper600, author = Mohamed Chekired; Varvara Roubtsova; Elham Kheradmand Nezhad; Serge Prudhomme; Marc Laforest,
title = DEM simulation with elliptical particles for micro-level geotechnics simulations,
year = 2017
}