Ephraim B. Washburn

Modeling multiphase effects in the combustion of HMX and RDX

RDX and HMX have similar structures and bur ning rates. However, the burning -rate temperature sensitivity ( �p) is significantly different between RDX and HMX at low pressures. Recent efforts to mathematically model the steady -state combustion of RDX and HMX with detailed chemical kinetics in the gas phase and distributed decomposition in the condensed phase hav e succeeded in modeling burning rates at a specific initial temperature. However, all have failed to calculate thep trends of HMX at low pressure and differentiate thep of RDX and HMX. RDX and HMX both burn with a thin multi -phase surface of bubbles i n liquid. A liquid -bubble submodel was developed to improvep calculations. Calculations including the liquid -bubble submodel produced the desired trends in both the HMX and RDXp values. To predict the observed HMXp values with the model, first, evapo ration in the sub -surface was limited near the gas -liquid surface. Second, the difference in surface temperature at different initial temperatures was adjusted to follow trends in experimental data. Third, the Marangoni effect was added to the calculation of the bubble velocities. At low pressures, the Marangoni effect was found to be greater in the higher initial temperature calculations because the temperature gradient was steeper. For RDX, there was little change in the calculatedp values with the addition of the liquid - bubble submodel. This is the first combustion model with detailed gas phase kinetics to predict the properp trends for both HMX and RDX.

Fundamental Simulation of Aluminum Droplet Combustion

The Liang and Beckstead aluminum-particle combustion model has been successfully joined with a detailed chemical-kinetic mechanism. The model has been used to investigate the effect of oxidizer concentration, particle size, and pressure on the combustion of solidpropellant combustion products and micrometer-sized aluminum particles. The simulated burn times of the aluminum particles were the same with the traditional propellant combustion products and the nitrate ester plasticized polyether (NEPE) propellant combustion products except at low pressures and aluminum particle sizes. In the simulations, the additional chlorine content interfered with the reactions to form aluminum oxide and caused the transition to kinetic-controlled burning of the particle to occur sooner. In the simulations with explosive products at constant pressure equilibrium, they could not produce enough heat feedback from the gas-phase flame to sustain combustion expect at high pressures and large diameter particles. However, the simulations with explosive products’ concentrations at the Chapman-Jouguet (C-J) point burned with a gas-phase flame without difficulty for most of the conditions studied. Some large-scale blast codes use the C-J concentrations instead of those found at the constant pressure equilibrium. The implications of these results are that if large-scale blast codes show significant burning of the aluminum particles before they are exposed to the atmospheric O2, this could be an unphysical result. The simulations with the C-J product concentrations also showed a sooner transition to kinetic-controlled combustion because of the presence of chlorine species. Based on a multi-scale approach, a one-dimensional, spherically symmetric set of simulations was completed. Although it was originally hypothesized that multicomponent diffusion effects would be important, results indicate that a mixture-averaged description is sufficiently accurate. Further analysis with the 1-D code is completed.