Particle crushing mechanics

Modeling Ballast Particle Crushing as a Phase Change

Crushing implies the division of particles into smaller parts, and may involve a change in shape. Particle crushing occurs when high contact forces are produced within a granular medium, mainly: gouge formation at fault faces, ballast degradation, and dynamic soil compaction. Continuum-Based models give an average representation of elastic weakening, often in the form of a visco-plastic deformation law. For instance, dashpots can be used in parallel with springs in Finite Element models to predict the frequency modes of fouled ballast. Discrete Element Methods (DEM) are better suited to capture the loss of compression and shear strength with particle crushing. In DEM, a Representative Elementary Volume typically contains several thousands of particles (or "balls"), within "walls" representing boundary conditions. As a result, DEM-based models designed with realistic grain sizes cannot be used to simulate large-scale particle assemblies involved in faults, railways or embankments. Moreover, DEM aims to mimic the physical processes occurring at the grain scale, which requires powerful observation and image analysis tools.

The objective of this study is to propose an alternative approach to the modeling of grain crushing, more descriptive than a volumetric creep law, but less computation-intensive than particulate mechanics. A DEM model of crushable ballast particle was designed and calibrated against experimental results of uniaxial compression tests on single sand grains. The calibrated model was then used to study the influence of the coordination number on the peak force at initial failure ("shielding effects"). On-going research focuses on the influence of particle size on particle strength, the micromechanical processes of crushing in granular assemblies subject to under monotonic and cyclic loading, and on the forms of energy dissipated by an assembly of bonded aggregates during crushing.

Pictures: Wang, 2015

Pictures: Bakhtiary, 2013; Wang, 2015