Session S2: TMS: Mesoscale Explosive Initiation I

Reaction rate model of energy release in shocked RDX

Shock-to-detonation transition in secondary explosives is driven by energy localization at microstructural heterogeneities, wherein local temperature spikes ("hot spots") trigger chemical reactions and formation of self-sustained reaction fronts. Emerging mesoscale methods, including sub-scale hydrodynamics and coarse-grained particle methods, can explicitly model these processes but require reaction rate models as input. I will present the development of such a model for RDX that combines first-principles predictions for condensed-phase decomposition under GPa pressures with a reduced model of thermal oxidation of RDX decomposition products. I will also discuss inherent shortcomings of the model due to uncertainties in the first-principles predictions and the scarcity of kinetic data for high-temperature and high-pressure conditions.   

Molecular Dynamics-informed RDX chemistry model and continuum static hot spot simulations

Reactive molecular dynamics (MD) simulations are able to describe the complex chemistry of high energy density materials and its coupling to mechanics. However, MD cannot reach the microstructural scales required to model hotspot formation and shock to detonation. Thus, we developed a multiscale model that uses MD simulations to inform a continuum model capable of reaching the microstructural scales. We developed a two-step reduced order chemistry model from reactive simulations of RDX and computed other critical physical properties like specific heat, thermal conductivity and equations of state. This information is used in a continuum model implemented in a finite elements code. We validate the continuum model via explicit MD hotspot calculations and then use it to characterize hotspot criticality for system sizes relevant to experiments.   

Multi-scale Modelling of Shock Sensitivity in Energetic Materials

Shock compression of energetic materials exposes the need for a wide variety of modelling and simulation capabilities due to critical length- and time-scales that easily span many orders of magnitude. For example, defects and other microstructure features give rise to hotspots which initiate chemical reactions on the nanometer and sub-nanosecond timescales, all of which complicate our predictions of sensitivity. To improve the safe use and reliability of energetic components in a time and cost effective manner, our modelling and simulation efforts need to be tightly coupled yet flexible enough to avoid shortcomings due to approximations baked into these tools. This talk will discuss the current progress and future outlook of these multi-scale efforts, with a focus on understanding the role of microstructure on the shock-to-detonation transition.   

Atomistic simulations for granular explosives under shock compression

Solid explosives are composite materials made of molecular crystal grains binded by a polymer matrix. Although most of their detonation properties are related to the explosive molecule itself, the granular structure can give a non-negligible contribution due the stacking porosity, the granular surface energy, or the properties of the binder: for instance detonation nanodiamonds synthesized from carbon-rich explosives display granularity-dependent size-distributions. We present a series of numerical studies based on molecular dynamics simulations which aim at improving our understanding of the impact of granularity on the shock/detonation behavior of these solid explosives. We show that their shock-induced properties depend mostly on the initial porosity amount and structuration. The local perturbation of the shock temperature and pressure are likely to modify the chemical decomposition process, suggesting a potential impact on the detonation products.   

Molecular dynamics study of shock loading followed by release in (100)-oriented }$\alpha $-RDX

Molecular dynamics simulations were used to study shock loading and release in (100)-oriented $\alpha $-RDX. A reverse-ballistic configuration was used where the sample, initially at 300 K, impacts onto a rigid, stationary piston at an initial velocity of either 1 km/s or 2 km/s, generating a supported shock wave. Failure is induced by allowing the simulations to continue through the release process from the maximum compression out to some final sample length that is about 1.5 times the starting value. The material response to the shock compression differs significantly for the 1 km/s and 2 km/s impacts. In the former case, extended, intersecting shear bands oriented at approximately 45\textdegree $^{\mathrm{\thinspace }}$to the shock direction form fill the [010] zone of the sample. For the latter case, a high density of localized translational defects develop but these do not grow or glide, presumably because they are formed immediately at the shock front and are ``locked in'' due to their very high density and interactions. These qualitatively differing states exhibit decidedly different behaviors during release, namely failure mediated by the shear bands in the weaker shock case and cavitation in the case of the stronger shock. Efforts to characterize these phenomena and draw contrasts with what is typically observed for spall in metals will be discussed.