Thermoplastic-based pyrotechnic compositions

Abstract

Conventional pyrotechnic time delays are based on powder mixtures pressed into tubes of precise lengths. Pyrotechnics based on filled polymer systems may offer some advantages. This includes continuous manufacture via extrusion processes. Thermoplastic-based pyrotechnic compositions can be formulated by filling conventional polymer matrices with oxidants such as potassium nitrate. More interesting are fluoropolymer matrices as these strong oxidants enable the design of non-gassing systems with suitable fuels. The presence of fillers dramatically increases the melt viscosity. The random packed limit for monodisperse spherical particles corresponds to a volume fraction of 0.637 (Krieger, 1959). Pyrotechnic compositions with oxidant filler volume fractions below this critical level are only viable for conventional polymers with very high C:H atomic ratios, e.g. polystyrene. Pyrotechnic compositions comprising of filled thermoplastics were simulated using EKVI thermodynamic software. This allowed for the calculation of the adiabatic flame temperatures, variation of the product composition and gas evolved with varied filler content. The EKVI thermodynamic simulations showed that polystyrene filled with potassium nitrate or potassium permanganate were unlikely to be viable as global maximum temperatures were not achieved below 78 vol.%. A fully fluorinated polymer filled with aluminium, magnesium, magnalium and calcium carbide were shown to be viable between 20 wt.% and 70 wt.% reducing agent. Co-polymers of tetrafluoroethylene and vinylidene fluoride filled with aluminium or magnesium showed similar adiabatic temperatures as compositions based on perfluorinated polymers with the same reducing agents. The energy required to decompose the thermoplastic binder would, however, lower the amount of energy available for the pyrotechnic reaction. Open flame burn test showed that the polystyrene-based compositions did not generate enough energy to decompose the thermoplastic fraction to sustain chemical reaction. Viton B filled with calcium carbide could not sustain burning. Compositions using aluminium and magnalium with Viton B as oxidant sustained burning over the range 20 wt.% to 60 wt.%. The effect of morphology was tested by using two grades of aluminium; atomized aluminium, and flake-like aluminium particles. Energy measurements of Viton B filled with aluminium and magnalium indicated that the maximum energy output occurred in the range 30 wt.% to 40 wt.% fuel for aluminium-based compositions and between 40 wt.% and 50 wt.% for magnalium-based compositions. A maximum burn rate of 82 mm s1 was achieved using a magnalium/Viton B composition. Friction and impact test showed that the compositions are insensitive. XRD analysis of the combustion residue of Viton B-based compositions using aluminium as fuel showed that the most abundant products formed were Al4C3, AlF3, carbon and an amorphous phase. An elemental balance indicated that the amorphous phase consisted of carbon, aluminium and fluorine. The XRD spectra of the residues of magnalium-based compositions had unidentified reflections. Quantitative XRD was, therefore, not possible on the reaction products of the magnalium-based compositions. TGA analysis on the combustion residue indicated that the combustion residue contained unreacted reagents. Scanning electron microscopy of the reaction residue revealed the presence of agglomerated cubic particles. EDX analysis indicated that the cubic particles consisted of aluminium and fluorine.