Filling a silo with a mixture of grains: Friction-induced segregation

– We study the filling process of a two-dimensional silo with inelastic particles by simulation of a granular-media lattice gas (GMLG) model. We calculate the surface shape and flow profiles for a monodisperse system and we introduce a novel generalization of the GMLG model for a binary mixture of particles of different friction properties where, for the first time, we measure the segregation process on the surface. The results are in good agreement with a recent theory, and we explain the observed small deviations by the nonuniform velocity profile. Mixtures of grains tend to separate as a response to virtually any type of external perturbation [1], an effect that can be a major problem in some practical situations, and useful in others. Indeed, understanding the underlying processes of segregation phenomena in granular mixtures is an intriguing problem of interest to scientists from a wide range of disciplines. Segregation occurs when a mixture is poured onto a horizontal base and a pile builds up. When the grains of the mixture differ in size, the large grains tend to gather at the bottom of the pile [2], while when the grains have different shapes, the more faceted grains are found preferentially near the top. An even more surprising effect can be observed when a mixture of small rounded and large faceted grains is poured between two parallel plates: stratification is observed —i.e., the grains organize spontaneously into stripes [3]. Particle size distribution crucially influences segregation. However, the particles generally differ not only in size but also in other properties, such as frictional ones. Frictional effects play a relevant role in band formation in 3-D rotating drums, in segregation in thin rotating drums, and also segregation and stratification in 2-D silos. Thus the study of the particular case