Computational prediction of probabilistic ignition threshold of pressed granular octahydro-1,3,5,7-tetranitro-1,2,3,5-tetrazocine (HMX) under shock loading

The probabilistic ignition thresholds of pressed granular octahydro-1,3,5,7-tetranitro-1,2,3,5-tetrazocine explosives with average grain sizes between 70 μm and 220 μm are computationally predicted. The prediction uses material microstructure and basic constituent properties and does not involve curve fitting with respect to or prior knowledge of the attributes being predicted. The specific thresholds predicted are James-type relations between the energy flux and energy fluence for given probabilities of ignition. Statistically similar microstructure sample sets are computationally generated and used based on the features of micrographs of materials used in actual experiments. The predicted thresholds are in general agreement with measurements from shock experiments in terms of trends. In particular, it is found that grain size significantly affects the ignition sensitivity of the materials, with smaller sizes leading to lower energy thresholds required for ignition. For example, 50% ignition threshold of the material with an average grain size of 220 μm is approximately 1.4-1.6 times that of the material with an average grain size of 70 μm in terms of energy fluence. The simulations account for the controlled loading of thin-flyer shock experiments with flyer velocities between 1.5 and 4.0 km/s, constituent elasto-viscoplasticity, fracture, post-fracture contact and friction along interfaces, bulk inelastic heating, interfacial frictional heating, and heat conduction. The constitutive behavior of the materials is described using a finite deformation elasto-viscoplastic formulation and the Birch-Murnaghan equation of state. The ignition thresholds are determined via an explicit analysis of the size and temperature states of hotspots in the materials and a hotspot-based ignition criterion. The overall ignition threshold analysis and the microstructure-level hotspot analysis also lead to the definition of a macroscopic ignition parameter (J) and a microscopic ignition risk parameter (R) which are statistically related. The relationships between these parameters are established and delineated.