The process of filling a silo with particles is simulated by placing them one-by-one at random locations. In simulations, a plane problem is considered and the particles are represented by disks. After each disk is placed on the system of disks already in a silo, the distribution of normal and shear stresses is recalculated. If any shear stress exceeds the predetermined limit, then a slip takes place. It is found that a slip can occur between a few particles simultaneously. Such a micro-slip reflects a physical instability in the system. It also results in a numerical instability. A Recursive Inverse Matrix Algorithm (RIMA) is implemented to account for the topological changes in the system during the loading. It allows efficient addition/removal of contact interfaces between the particles by updating the inverse matrix of the system rather then generating and solving a system of equations after each topological change. The extent of instabilities is determined numerically using RIMA update algorithm iteratively. As a result of irreversible events taking place during the filling operation, the final stress distribution in a filled silo is loading history-dependent. A numerical example of filling a silo with 500 particles is given as an illustration.

This content is only available via PDF.
You do not currently have access to this content.