Doctor of Philosophy (PhD)


Mechanical Engineering

Document Type



Shock sensitivity of heterogeneous explosive composites is dependent on the formation of hot-spots which are small regions of elevated temperatures within the material. Changes in the initial meso-structure (i.e. packing density, composition, particle size and shapes) of the explosives can significantly alter the hot-spot fields in the material and thereby affect its shock sensitivity. In this study, an explicit, 2D, Lagrangian finite and discrete element technique is used to numerically simulate the deformation induced heating of granular mixtures of explosive (HMX), and metal (Al) particles due to piston supported uniaxial deformation waves (400 ≤ Up ≤ 800 m/s). A number of simulations are performed by systematically varying the effective initial packing densities φs, metal mass fractions λm, and particle size distributions. Emphasis is placed on charactering how the inclusion of metal (Al) affects both the effective wave end states (Hugoniots) and the hot-spot fields within the explosive (HMX) component relative to neat HMX. Variations in hot-spot volumetric quantities such as number density and volume fraction are characterized since these quantities can be used in the ignition and growth models to describe macro-scale material sensitivity. Predictions indicate that porosity has a leading order effect on the shock sensitivity of the material due to enhanced dissipation resulting from plastic pore collapse. For a fixed porosity and piston speed, inclusion of metal is found to enhance the effective plasticity in the material due to higher pressures. This leads to larger hot-spots within the metalized formulations. However, due to the high thermal conductivity of the metal, frictional induced hot-spots are suppressed within the material since most of the frictional dissipation at the Al-HMX interfaces is absorbed by the metal. Additionally, hot-spot formation is found to have a highly non-linear dependence on Al particle size with a substantial decrease in hot-spot number density and volume fraction predicted with increasing metal particle size. Meso-structural stochasticity arising due to random seeding of particles, and/or large particle clustering were found to affect the hot-spot statistics minimally.



Document Availability at the Time of Submission

Release the entire work immediately for access worldwide.

Committee Chair

Gonthier, Keith