Abstract
The first discrete modeling of soils can be traced to Hertz [90] who formulated a contact law between spheres, and Reynolds [167] who proposed a dilatancy theory. Dantu [55] and Schneebli [180] idealized real soils as assemblies of rigid rods, and noticed some striking similarities between the mechanical responses of these mechanical analogs and real soils. Duffy and Mindlin [70], Deresiewicz [59, 60], and Thurston and Deresiewicz [204] examined the response of soil models made of spheres. Biarez [21] used glass beads and duralumium rods to examine the elastic and limit response of soils, and applied his observations to analyze practical problems in geotechnical engineering. These pioneer works were later followed by photoelastic investigations [e.g., 67, 68] to visualize stresses within granular media.