Simulation of Grain Behaviour for Explosive Formulations

For production of high quality PBX charges, it is essential to have low viscosity formulations available during the casting process. As generally known, for a certain filling grade, casting viscosity depends strongly on particle size distribution and on particle shape. We have shown in previous work, that a high tap density of the solid energetic material is one strong requirement to get low viscosities after being mixed with corresponding binder ingredients. So far, the use of particle size distributions in formulation work was based on empirical or semi-empirical approaches, but there was, up to now, no theoretical background to understand in more detail the behaviour of solid energetic material in view of optimised density or flowability. In this paper we present a tool, which enables to perform simulations of packing densities and grain flowability based on a certain particle size distribution. The model used therefore consists of a distinct element method (DEM), observing each particle until its final position in the ensemble is reached. Considering a realistic particle size distribution, spanning several orders of magnitude and reasonable geometric conditions, which allow to neglect vessel border effects, at least several hundred thousands of particles have to be taken into account for a simulation. Therefore a very efficient algorithm to detect particle contacts, based on 3-dimensional Delaunay triangulation had to be established. But to simulate particles behaviour during a pouring or a vibrating process, this contact model had to be combined with an appropriate physical model to describe energy dissipation. Considering the influence of tapping frequency on average movement and segregation of particles we can show our first simulation results. Furthermore we are able to show the influence of different particle distribution on flowability.