Analysis of particle contact using frustrated total internal reflection

Abstract

Within the field of soil mechanics a continuum assumption is generally adopted in order to avoid the complications of modelling micro-mechanical behaviour. However, certain constitutive behaviour can only be explained by investigating particle level interactions. Numerical investigations, such as those using the Discrete Element Method (DEM) to model soil particles as clusters of spheres, have delivered a greater understanding of the micro-mechanical behaviour. One of the limiting factors in current DEM approaches is modelling of the particle–particle or particle–surface contact behaviour. Hence, an experimental methodology has been developed and used to study particle–surface contact behaviour. The experimental methodology involves loading particles onto a piece of sapphire glass and observing the resulting contact area. In order to distinguish between the contacted area and the rest of the particle, the principle of frustrated total internal reflection and evanescent waves was used which results in only objects in very close proximity to the glass being illuminated and visible. This methodology hence allows the number of contacts and the area of those contacts to be tracked during loading and over time. This paper presents the validation of the experimental methodology by comparing the observed contact behaviour of plastic beads against Hertzian contact theory. In addition, the results from tests on sand samples are presented which show a density of 0.40 and 0.80 contacts per D_{50}^2 for coarse and fine grained sand respectively at an isotropic stress state which subsequently increases to 0.90 to 1.00 contacts per D_{50}^2 at peak deviatoric stress. It was also found that the fine sand particle contacts carried a maximum load of approximately 0.27 N per contact whereas the coarser sand was able to carry substantially higher loads.