Contribution for the modelling of three-dimensional granular matter : study of railway ballast.

Discrete element methods are an alternative to finite element approach to modelise thebehaviour of granular materials. This work is devoted to the study of a granular matter submittedto cyclic loadings : the ballast of the railway track submitted to repeated train passing. We choosethe NonSmooth Contact Dynamics approach for the resolution method, and we present all technicalaspects for the two-dimensional and three-dimensional cases : contact detection algorithmbetween convex polygons and convex polyhedra, the kinematic parametrization, the resolution ofthe contact and friction problem. We investigate the best parameters for numerical computation ofcyclic loadings. In the following we present some results in order to show the ability of discreteelement methods to describe the behaviour of granular matter under cyclic loading. A comparisonbetween experiment and numericals computations underline the special behaviour of this kind ofsystem. We focuse on the thin granular layer behaviour with a specific mechanical reponse dueto the creation of special structures. The three-dimensional model has been used as an exampleto study well known behaviour of masonery structure. We present a numerical study of lateralresistance of railway track based on an experiment with digitised ballast grains.

Data and Resources

Additional Info

Field Value
Source https://theses.hal.science/tel-00077519
Author Saussine, Gilles
Maintainer CCSD
Last Updated May 15, 2026, 08:30 (UTC)
Created May 15, 2026, 08:30 (UTC)
Identifier tel-00077519
Language fr
Rights https://about.hal.science/hal-authorisation-v1/
contributor Laboratoire de Mécanique et Génie Civil (LMGC) ; Université de Montpellier (UM)-Centre National de la Recherche Scientifique (CNRS)
creator Saussine, Gilles
date 2004-10-14T00:00:00
harvest_object_id a397cea3-6777-45e1-a52f-2815a7cc744e
harvest_source_id 3374d638-d20b-4672-ba96-a23232d55657
harvest_source_title test moissonnage SELUNE
metadata_modified 2025-07-17T00:00:00
set_spec type:THESE