Thermal Explosion Experiment Research Articles

Thermal–mechanical analysis for confined HMX-based polymer-bonded explosives

It is known that the reaction rate of the thermal decomposition of polymer-bonded explosives exposed to cook-off has a certain relation with temperature, confining pressure and some other factors, which were verified by many experiments. Temperature-dominated thermal-decomposition models were developed for various high explosives and applied to study their decomposition process, such as HMX- and TATB-based polymer-bonded explosives. These models have reasonable accuracy. For example, the multistep thermal-decomposition model of PBX 9501 (which consists of 95% HMX, 2.5% Estane and 2.5% BDNPA/F) proposed by Tarver considers decomposition of both HMX and polymer binders. The temperature-dominated thermal-decomposition model only applies to preignition thermal decomposition. After ignition occurs, the dominant mechanism of the reaction transforms to deflagration and subsequent explosion, where the effect of pressure can no longer be neglected. Furthermore, the time scale of deflagration and explosion (millisecond or microsecond) differs significantly from the time scale of slow thermal decomposition (hours or minutes). The numerical model of postignition phenomenon (deflagration and final explosion) is still under investigation and is far from maturity. The predicted violence scale resulting from thermal explosion does not agree with experiment very well. An alternative method is to conduct a thermal–mechanical analysis for preignition stage, which takes advantage of a developed temperature-dominated thermal-decomposition model, and to analyze the stress caused by quasi-static thermal expansion. Herein, a thermal–mechanical analysis is implemented for a one-dimensional time-to-explosion experiment (ODTX) and a scaled thermal explosion experiment (STEX) with HMX-based polymer-bonded explosives inside using the finite element method. Then, the finite element model is applied to investigate the thermal decomposition of PBX 9501 inside an explosive device exposed to cook-off. The regions that have maximum temperature, maximum hydrostatic pressure and maximum von Mises stress are identified based on simulation results, which can benefit future improvement of the explosive device.

Numerical study of countermeasure against thermal stimuli for HMX-based polymer-bonded explosives

ABSTRACTAn accurate thermal decomposition model of Octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX)-based polymer-bonded explosives is indispensable to the design of countermeasure against environmental thermal stimuli for certain explosive devices. However, the complicated chemical reactions occurring in decomposition of HMX-based polymer-bonded explosives pose great challenge to scientists. In this study, the thermal ignition kinetic model proposed by Tarver is implemented to study thermal decomposition of HMX-based polymer-bonded explosives using commercial software Abaqus, which does not only consider the thermal decomposition of HMX but also the polymer binders. The simulation results are compared to ODTX and Scaled Thermal Explosion Experiment (STEX) and reasonably good agreements are achieved. Then the thermal decomposition model is applied to analyze an explosive device exposed to environmental thermal stimuli. Furthermore, countermeasure against environmental thermal stimuli is suggested and analyzed quantitatively for the explosive device. The time to explosion and environmental temperature before ignition is calculated and analyzed for the explosive device under various heating rates.

The Thermal Properties of Nitrocellulose: From Thermal Decomposition to Thermal Explosion

ABSTRACTNitrocellulose is an extremely dangerous explosive substance, which will decompose at a relatively low temperature. This article studied the thermal properties transformed from thermal decomposition to thermal explosion for a less violent nitrocellulose sample (nitrogen content: 10.92%). Four experiments were implemented: XPS analysis and SEM analysis, DSC-TG analysis, and small-scale thermal explosion experiment. From XPS analysis and SEM analysis, high temperatures led to the decrease of N1s and O1s (O-NO2 bond) and the break or even “melt” of the nitrocellulose fibers. The results demonstrated that nitrocellulose would become more unstable if it was exposed to high temperature. From the DSC-TG analysis and small-scale thermal explosion experiment, the upper critical ambient temperatures (UCATs) of the nitrocellulose were calculated. The UCATs of the nitrocellulose sample would decrease from 132.5ºC to 96.4ºC with the dimensions increased from 0.1 m to 1.6 m. The results demonstrated that the risk of spontaneous combustion and thermal explosion increased significantly with the increase of the volume of nitrocellulose.

Thermal Explosion Experiment of Energetic Materials

For the thermal safety evaluation of energetic materials, the experimental devices, test method, conditions of determination, numerical calculation and simulation are listed in this paper on thermal explosion experiment both at home and abroad. By listing typical examples and comparing the existing researches, current study and developing trend of thermal explosion experiment are summarized and prospected.

The evolution of solid density within a thermal explosion. I. Proton radiography of pre-ignition expansion, material motion, and chemical decomposition

We report proton transmission images obtained during direct heating of a sample of PBX 9501 (a plastic bonded formulation of the explosive nitramine octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX)) prior to the ignition of a thermal explosion. We describe the application of proton radiography using the 800 MeV proton accelerator at Los Alamos National Laboratory to obtain transmission images in these thermal explosion experiments. We have obtained images at two spatial magnifications and viewing both the radial and the transverse axes of a solid cylindrical sample encased in aluminum. During heating we observe the slow evolution of proton transmission through the samples, with particular detail during material flow associated with the HMX β-δ phase transition. We also directly observe the loss of solid density to decomposition associated with elevated temperatures in the volume defining the ignition location in these experiments. We measure a diameter associated with this volume of 1–2 mm, in agreement with previous estimations of the diameter using spatially resolved fast thermocouples.

Modeling and Analysis of Compressive Properties of Porous NiTi Shape Memory Alloy Using Artificial Neural Network

Artificial neural network (ANN) is an intriguing data processing technique. Over the last decade, it was applied widely in the chemistry field, but there were few applications in the porous NiTi shape memory alloy (SMA). In this paper, 32 sets of samples from thermal explosion experiments were used to build a three-layer BP (back propagation) neural network model. According to the registered BP model, the effect of process parameters including heating rate ( ), green density ( ) and particle size of Ti ( d ) on compressive properties of reacted products including ultimate compressive strength ( v D σ ) and ultimate compressive strain (ε ) was analyzed. The predicted results agree with the actual data within reasonable experimental error, which shows that the BP model is a practically very useful tool in the properties analysis and process parameters design of the porous NiTi SMA prepared by thermal explosion method.

Simulating thermal explosion of octahydrotetranitrotetrazine-based explosives: Model comparison with experiment

A model comparison with measurements for the thermal, chemical, and mechanical behaviors in a thermal explosion experiment is presented. Confined high explosives (HEs) are heated at a rate of 1°C∕h until an explosion is observed. The heating, ignition, and deflagration phases are modeled using an arbitrarily Lagrangian-Eulerian (ALE3D) code that can handle a wide range of time scales that vary from a structural to a hydrodynamic time scale. During the preignition phase, quasistatic mechanics and diffusive thermal transfer from a heat source to the HE are coupled with the finite chemical reactions that include both endothermic and exothermic processes. Once the HE ignites, a hydrodynamic calculation is performed as a burn front propagates through the HE. Two octahydrotetranitrotetrazine (HMX)-based explosives, LX-04 and LX-10, are considered, whose chemical-thermal-mechanical models are constructed based on measurements of thermal and mechanical properties along with small-scale thermal explosion measurements. The present HMX modeling work shows the violence calculations with thermal predictions associated with a confined thermal explosion test. The simulated dynamic response of HE confinement during the explosive phase is compared to measurements in larger scale thermal explosion tests. The explosion temperatures for both HEs are predicted to within 1°C. Calculated and measured wall strains provide an indication of vessel pressurization during the heating phase and violence during the explosive phase.

Test-based thermal explosion model for HMX

We present a thermal explosion (cookoff) model for an HMX-based plastic bonded explosive (LX-10). The thermal–chemical–mechanical response of LX-10 is modeled based on the measurements from the scaled thermal explosion experiment (STEX) at the Lawrence Livermore National Laboratory. Confined LX-10 is heated at a rate of 1 °C/h until an explosion is observed. The modeled cookoff problem is simulated by the Arbitrarily Lagrangian–Eulerian hydrocode (ALE3D) that can handle a wide spectrum of time scales that vary from a structural to a high speed shock physics time scale. In addition to a comprehensive model for energetic material, the confinement material namely an AerMet 100 steel is modeled as a Steinberg–Guinan material with a Johnson–Cook failure model with a statistical failure distribution. By using the size distribution data from the fragmentation experiment, the metal fracture and fragmentation due to an explosion are modeled. The explosion temperature is predicted to within 1°. Calculated wall strain provides violence associated with the thermal explosion process and agrees favorably with the measured STEX data.

Simulating thermal explosion of cyclotrimethylenetrinitramine-based explosives: Model comparison with experiment

We compare two-dimensional model results with measurements for the thermal, chemical, and mechanical behavior in a thermal explosion experiment. Confined high explosives (HEs) are heated at a rate of 1°C∕h until an explosion is observed. The heating, ignition, and deflagration phases are modeled using an Arbitrarily Lagrangian-Eulerian code (ALE3D) that can handle a wide range of time scales that vary from a structural to a dynamic hydrotime scale. During the preignition phase, quasistatic mechanics and diffusive thermal transfer from a heat source to the HE are coupled with the finite chemical reactions that include both endothermic and exothermic processes. Once the HE ignites, a hydrodynamic calculation is performed as a burn front propagates through the HE. Two cyclotrimethylenetrinitramine-based explosives, C-4 and PBXN-109, are considered, whose chemical-thermal-mechanical models are constructed based on measurements of thermal and mechanical properties along with small scale thermal explosion measurements. The simulated dynamic response of HE confinement during the explosive phase is compared to measurements in larger scale thermal explosion tests. The explosion temperatures for both HEs are predicted to within 5°C. Calculated and measured wall strains provide an indication of vessel pressurization during the heating phase and violence during the explosive phase. During the heating phase, simulated wall strains provide only an approximate representation of measured values indicating a better numerical treatment is needed to provide accurate results. The results also show that more numerical accuracy is needed for vessels with lesser confinement strength. For PBXN-109, the measured wall strains during the explosion are well represented by the ALE3D calculations.

Chemical Kinetic Modeling of HMX and TATB Laser Ignition Tests

Recent high-power laser deposition experiments on octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) and 1,3,5-triamino-2,4,6-trinitrobenzene (TATB) produced ignition times from milliseconds to seconds. Global chemical kinetic thermal decomposition models for HMX and TATB developed to predict thermal explosion experiments lasting seconds to days are applied to these laser ignition experimental data. Excellent agreement was obtained for TATB, while the calculated ignition times were longer than experiment for HMX at lower laser fluxes. Inclusion of HMX melting and faster reaction for liquid HMX in the HMX decomposition model improved the agreement with experiment at lower laser energies.

Thermal decomposition models for HMX-based plastic bonded explosives

Global multistep chemical kinetic models for the thermal decomposition of octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazine (HMX)-based plastic bonded explosives (PBXs) using endothermic or exothermic binders are developed for calculation of the times to thermal explosion as functions of heating rate and geometry in the Chemical TOPAZ heat-transfer computer code. The decomposition mechanisms of the binder materials are treated separately from that of HMX, and the chemical reactions of each constituent are assumed to occur independently. Experimental data and theoretical predictions of the thermal properties, decomposition pathways, and chemical kinetic reaction rate constants are used to develop reaction sequences for each of the components present in the PBX at various weight percentages. The measured times to thermal explosion at various initial temperatures in a new One-Dimensional Time-to-Explosion (ODTX) apparatus are compared to the Chemical TOPAZ predictions. Two series of pristine HMX spheres formulated with coarse and fine particles are tested for the first time in an ODTX apparatus. The pure HMX data clearly show that the presence of an endothermic binder in a PBX increases the times to thermal explosion, while the presence of an exothermic binder decreases the times to explosion. The magnitudes of these changes in explosion time depend upon the chemical stabilities and heats of reaction of these binders. A four-step decomposition model is developed for HMX, which includes the β to δ solid-phase transition as the first endothermic reaction. This model accurately reproduces the pure HMX curves. Decomposition models for the various binder components are then used with the HMX model to accurately reproduce the ODTX time-to-explosion curves. Comparisons are also made to times to thermal explosion obtained in various experiments involving aged PBXs, ramped temperature rate increases, unconfined explosives, and a larger size, cylindrical geometry called the scaled thermal explosion experiment.

Towards a predictive thermal explosion model for energetic materials

We present an overview of models and computational strategies for simulating the thermal response of high explosives using a multi-physics hydrodynamics code, ALE3D. Recent improvements to the code have aided our computational capability in modeling the behavior of energetic materials systems exposed to strong thermal environments such as fires. We apply these models and computational techniques to a thermal explosion experiment involving the slow heating of a confined explosive. The model includes the transition from slow heating to rapid deflagration in which the time scale decreases from days to hundreds of microseconds. Thermal, mechanical, and chemical effects are modeled during all phases of this process. The heating stage involves thermal expansion and decomposition according to an Arrhenius kinetics model while a pressure-dependent burn model is employed during the explosive phase. We describe and demonstrate the numerical strategies employed to make the transition from slow to fast dynamics. In addition, we investigate the sensitivity of wall expansion rates to numerical strategies and parameters. Results from a one-dimensional model show that violence is influenced by the presence of a gap between the explosive and container. In addition, a comparison is made between 2D model and measured results for the explosion temperature and tube wall expansion profiles.

Quantitative experiments on phreatomagmatic explosions

An experimental set-up for controlled generation of thermal explosions (Fuel Coolant Interactions) has been built by an interdisciplinary research group of volcanologists, physicists and engineering scientists: The TEE-Haus (Thermal Explosion Experiment). During the last 3 years more than 500 experimental runs were performed. Aim of the project was not to build a “mini volcano” but to produce igniteable mixtures of water and hot melt on a small scale in order to investigate basic physical aspects. Such small scale mixtures, however, most probably also occur during phreatomagmatic events in nature. Mixtures of water and magma on a large scale may be modelled by combination of small elements. Earlier experimental studies of thermal explosions were carried out with the use of metal melts nearly exclusively. Therefore, a carbonate melt (approx. 0.3 kg of 1:1 mass ratio Na 2CO 3 / K 2CO 3) was used for initial experimental series because of easy handling and more “magma similarity”. By injection of water into this melt explosive mixtures were produced and ignited (self-triggering). Explosive interactions occurred in a wide melt-temperature range (720 to 1040°C). At melt temperatures exceeding 1000°C, CO 2 increasingly degassed. Presence of non-condensible gas bubbles in the melt strongly reduced the explosivity. The intensity of explosions only slightly increased with higher melt temperatures but strongly increased with higher water injection velocities, because of more intensive mixing. A minimum injection velocity was found to be approx. 0.7 m s −1. Experiments with silicate melts, derived from diverse remelted volcanic rocks were carried out with a slightly modified set-up. Explosions were ignited by an additional trigger: a shock wave of approx. 8 J. In volcanic systems, shock waves of this magnitude are abundant (volcanic tremor, etc.). Strong explosions with silicate melts were produced at melt temperatures between 1350 and 1750°C. The produced explosions exerted repulsion forces of up to 24,000 N with a typical pulse duration of 1 ms (carbonate melt) resp. 1.5 ms (silicate melts). The ejection velocity was found to be in a range between 200 and 400 m s −1. Explosion energy values up to 500 J were calculated. The maximum explosion pressure was calculated to be in the range between 10 and 100 MPa. Only part of the melt (so-called interactive melt) reacts explosively with injected water. Most of the melt is passively ejected by and after the thermal explosion. Different fragmentation processes result in specific grain size and grain shape populations that can be distinguished from each other. Particles resulting from the fine fragmentation process that precedes and causes the thermal explosion are characterized by angular to subrounded shapes and grain sizes ranging from 20 μm to 180 μm. The amount of interactive melt is proportional to the intensity of the explosion. The maximum amount of water vapourized at the explosive interaction (so-called interactive water) was estimated. Thus, the water/melt mass ratio for explosive interactions could be calculated. The values range between 1 25 and 1 35 (carbonate melt) resp. between 1 6 and 1 25 (silicate melts).