| The work described involves the study of shock-induced reactions in binary/heterogeneous reactive powder mixtures that produce little or no gas and the possibility that a detonation could be produced in such mixtures. The first task was to identify the mechanism by which these mixtures reacted when shocked. Critical initiating shock pressure and autoignition temperature were determined for a number of mixtures by recovery of shocked samples and differential thermal analysis, respectively. In addition, recovery experiments using Mn+S in two different charge geometries and at two different nominal densities (60% TMD and 90% TMD) were carried out to study the effect of the rigid confining walls. It was found that in general mixtures that were the most thermally sensitive were not necessarily the most shock sensitive, and vice-versa. For the majority of mixtures tested, the minimum shock energy required to cause the entire sample mixture to react was much less than the minimum autoignition enthalpy. This result implies that initiation caused by the shock wave must be due to local shock energy concentration at particle contact interfaces and at locations where pores exist, where hot spots are formed, rather than bulk heating of the sample.;Finally, the possibility of detonating reactive powder mixtures that produce little or no gas was considered in light of the known shock initiation mechanism. It was shown that two conditions must be satisfied: 1) the volumetric expansion of the products must be sufficient to support a shock wave in the reactants and 2) the reaction time must be short compared to lateral relief time. It was shown that due to the fact that products usually contract upon reaction, compared to the reactants, few mixtures can satisfy the first requirement especially when the initial reactants are very porous. Those mixtures that satisfy the first condition typically produce large reaction temperatures and/or products that have low densities and large thermal expansion coefficients. The second condition could be satisfied when the charge diameter is large enough so that the lateral relief time scale is greater than the reaction time scale. This condition ensures that the expansion occurs mainly in the direction of shock propagation. A quasi one-dimensional model of the detonation reaction zone was developed to account for the competition between the energy release rate and the momentum and energy losses due to lateral expansion. Using this model with a relation for the reaction rate derived from the experimental results mentioned above and a model that accounts for the lateral losses, it was shown that unless burning velocities can be increased significantly (by at least two orders of magnitude for Ni-Al mixtures and three orders of magnitude in Mn-S mixtures) compared to that in mixtures of micron-sized powders, detonations may not be observable in charges with diameters less than one meter.;The shock initiation process was then directly observed in different powder mixtures (Mn+S, Zn+S, 2Al+Bi2O3, 2Al+3PbO, 8Al+3Pb 3O4, 8Al+MoO3, 4Al+Fe2O3, Ti+2B, Ti+B, Ni+Al, 5Ti+3Si, and Ti+Si; note that some of those mixtures were also mechanically activated in a ball-milling machine) contained in recovery capsules over long time scales using light detectors and thermocouples. Shortly following shock passage (a few μs), it appeared that only a small fraction of the mixture had reacted. The bulk of the mixture reacted only after many milliseconds. The observed time of bulk reaction was found to be mostly independent of shock pressure but seemed to correlate with the burning speed of the mixtures. This further confirmed the hypothesis that initiation onset first took place in hot spots and continued from there via a burning reaction, which is relatively slow and pressure-independent in the powders considered in this study. |
