Session O6: EM.1 Cook-off

Observations and Modeling of the Component Mechanisms in Deflagration

We have used dynamic x-ray and proton radiography to observe the behavior of a series of HMX based energetic materials formulations undergoing thermal explosions. The result of these observations is a mechanism for deflagration based on both gas phase convective burning and solid phase conductive burning. The velocities for both the convective and conductive burns are tied together by the single combustion pressure driving both in a single experiment. The convective rate is directly measured as the burn front in the radiographs. The pressure associated with that rate is inferred from independently measured burn rate verses pressure data. This same pressure is then assumed to drive the conductive burning which begins as the convective burn front lights the material surface. Using a single, independently validated particle size distribution for damaged HMX, the combination of pressure driven convective lighting and conductive consumption is then calculated and compared to the measured transmission data sets. This same model with different initial pressurizations is used to successfully model deflagration in PBX9501, PBXN-9, and LX-07. In addition, a correlation between initial pressurization, convective/conductive velocity, and final ``reaction violence'' is observed. This leads us to the use of convective velocity as a metric for final energy release rate and therefore overall reaction violence.   

A thermodynamically based definition of fast verses slow heating in secondary explosives

The thermal response of energetic materials is often categorized according to the rate of heating as either fast or slow, $e.g.$ slow cook-off. Such categorizations have most often followed some operational rationale, without a material based definition. We have spent several years demonstrating that for the energetic material octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) a single mechanism of thermal response reproduces times to ignition independent of rate or means of heating over the entire range of thermal response. HMX is unique in that bulk melting is rarely observed in either thermal ignition or combustion. We have recently discovered a means of expressing this mechanism for HMX in a reduced form applicable to many secondary explosives. We will show that with this mechanism a natural definition of fast versus slow rates of heating emerges, related to the rate of melting, and we use this to illustrate why HMX does not exhibit melting, and why a number of other secondary explosives do, and require the two separate categories.   

A Simple Model for the Pressure field from a Distribution of Hotspots

At the APS SCCM in 2009, Hill, Zimmerman and Nichols showed how, assuming that burn fronts propagate at constant speed from individual point hotspots distributed randomly in a volume, the reaction rate history could then be determined. In this paper a simple analytic approximation is found for the time history of the pressure in the volume. Using acoustic theory, the time history of the pressure field for burning from a single spherical, isolated hotspot of finite radius is developed. Then at any point in the volume, the overall pressure history is determined from the sum of the pressure fields from all the individual hotspots. The results are shown to be in qualitative agreement with full mesoscale calculations of the reaction and burning from a finite size spherical hotspot.   

Microwave frequency material properties and ignition predictions of neat and plastic bound explosives

Microwave energy has been considered for ignition, enhanced burning, and detection/defeat of energetic materials. However, the very limited data set of electromagnetic properties for both neat and plastic bound explosives has severely limited design and implementation of detection, defeat, and initiation devices. In this work, we report complex permittivity measurements for both neat and plastic bonded energetic materials such as HMX, RDX, PBX9501, etc. These measurements provide a new, more extensive set of self-consistent data that can be used to predict the response of such materials to electromagnetic energy. Using this data in conjunction with finite element analysis software, a high localized field experimental microwave applicator was designed and microwave heating predictions were calculated. Predictions show the feasibility of heating low-loss energetic materials in such cavities with high local electric fields without the need for susceptor particles. For the plastic bound materials, the effect of the binder is presented, showing that electromagnetic energy is preferentially absorbed in the more absorptive binder, resulting in significant gradients within individual energetic crystals (e.g., HMX crystals in PBX 9501). These predictions are now being used to aid experimental work with the applicator cavity and have demonstrated the feasibility of volumetrically heating energetic materials in short time scales with low microwave power levels.   

Modeling the Effects of Confinement during Cookoff of Explosives

In practical scenarios, cookoff of explosives is a three-dimensional transient phenomenon where the rate limiting reactions may occur either in the condensed or gas phase. The effects of confinement are more dramatic when the rate-limiting reactions occur in the gas phase. Explosives can be self-confined, where the decomposing gases are contained within non-permeable regions of the explosive, or confined by a metal or composite container. Self-confinement is prevalent in plastic bonded explosives at full density. The time-to-ignition can be delayed by orders of magnitude if the reactive gases leave the confining apparatus. Delays in ignition can also occur when the confining apparatus has excess gas volume or ullage. Explosives with low melting points, such as trinitrotoluene (TNT) or cyclotrimethylenetrinitramine (RDX) are complex since melting and flow need to be considered when simulating cookoff. Cookoff of composite explosives such as Comp-B (mixture of TNT and RDX) are even more complex since dissolution of one component increases the reactivity of the other component. Understanding the effects of confinement is required to accurately model cookoff at various scales ranging from small laboratory experiments to large real systems that contain explosives.