Microchannel combustion dynamics of low gas, environmentally friendly time delay compositions

Eric J Miklaszewski, Purdue University

Abstract

The purpose of this study is to investigate environmentally friendly time delay compositions. Replacement reactive systems for the widely used tungsten delay composition (W/BaCrO4/KClO4/diatomaceous earth) are needed due to concerns over the toxicity of hexavalent chromium and perchlorates. The ideal system is environmentally benign and gasless in nature (<10 mg of gas formed per gram of>reactive) with burning rates in the range of the W/BaCrO4/KClO4 composition (0.6 to 150 mm·s–1). In this work, condensed phase reactives (e.g, Ti/C-3Ni/Al and Ti/C-Ni/Al-Al2O3), Si/metal-oxide reactives and Mn/MnO2 are systematically explored as potential time delay compositions. Additionally, combustion performance predictions from a 2-D axisymmetric COMSOL 4.3a Multiphysics model using measured thermal properties and experimentally determined Arrhenius kinetics are presented. Systems based on condensed phase reactions, that are typically used in combustion synthesis (e.g., Ti/C or Ni/Al) are of interest as replacements due to their wide range of combustion velocities and potentially low environmental impact. The combustion characteristics of the Ti/C-3Ni/Al reactive system were examined in microchannels with inner diameters ranging from 3.0 – 6.0 mm (i.e., similar to that of a common delay housing). It was found that this reactive system could be tailored to overcome the heat losses associated with small diameter microchannels by changing the relative amounts of Ti/C and 3Ni/Al. At 40 wt.% Ti/C content, the failure diameter was found to be between 3.0 and 4.0 mm, while at 30 wt.% Ti/C the failure diameter was between 4.8 and 6.0 mm. Measured combustion temperatures in metal microchannels were approximately 1700 K while those of unconfined pellets were around 100 K greater. Increasing Ti/C content resulted in faster combustion velocities while decreasing microchannel diameter resulted in slower combustion velocities. At these small sizes the effects of adding a thermal barrier (specifically Grafoil™) to minimize radial heat losses to the microchannel were shown to be minimal with respect to combustion velocity. The Ti/C-3Ni/Al system was shown to be a suitable delay fuze composition with tunable combustion velocities ranging from 2.1 - 38.1 mm•s-1 in aluminum microchannels with diameters ranging from 4.0 – 6.0 mm. The Mn/MnO2 reactive system was also investigated as a suitable replacement for the traditional W/BaCrO4/KClO4/diatomaceous earth delay composition. The delay performance, ignition sensitivity, and aging characteristics were examined in aluminum microchannels similar in diameter to common delay housings (4.7 mm). Stoichiometries with measured combustion temperatures between 1358 and 2113 K were self-sustaining with combustion velocities ranging from 2.4 to 7.3 mm•s-1. The Mn/MnO 2 system produced less gas than W/BaCrO4/KClO4/diatomaceous earth compositions allowing consideration for use in sealed delay housings. Accelerated aging at 70°C and 30% relative humidity for eight weeks resulted in no measurable loss of performance. Safety characterization showed that this composition is not sensitive to ignition by friction or electrostatic stimuli. The combustion products (as determined by X-ray diffraction) appear to be benign based on current regulations. Therefore, the Mn/MnO2 system appears to be a suitable low gas-producing, non-sensitive, less toxic delay composition with good longevity. Predictive computational models to complement laboratory experiments could help to rapidly develop new pyrotechnic systems. Typically a time intensive experimental approach is necessary to demonstrate applicability. In this effort, a simplified model is developed for the Mn/MnO2 reactive system and compared to experimental data taken previously. First, kinetic parameters and thermal properties were experimentally determined. Effective thermal properties of a Mn/MnO2 powder compact were measured using the transient plane source method and kinetic parameters were obtained using the Boddington method. An effective activation energy of 56.2 ± 11.7 kJ/mol was determined. A 2-D axisymmetric COMSOL 4.3b Multiphysics model was then developed using global one step first order Arrhenius kinetics. By using this approach, a model of a reacting unconfined pellet predicts combustion velocities similar to those observed experimentally. However, simulations were not able to predict the effect of metal microchannel confinement, changing channel diameter or channel material to those observed experimentally. In order to more accurately simulate these effects, a considerably more complex model with multiphase heat transfer effects would likely be needed.

Degree

Ph.D.

Advisors

Son, Purdue University.

Subject Area

Chemical engineering

COinS