View article

A random packing algorithm is implemented to construct numerically models of heterogeneous propellants. These models consist of random distributions of spheres in a periodic cube. The spheres can have various sizes, randomly assigned if desired. Packing fractions are calculated for a bimodal model (spheres of two different sizes) and compare well with old experimental data for steel shot. Slices through the cube de ne surfaces that represent propellant surfaces, and the stoichiometry of these surfaces is a stochastic variable. Several realizations of the propellant cube are constructed and variations in surface stoichiometry are examined using histograms and Fourier series. A sample ame calculation is presented and solved using data de ned by a single surface realization.