Shock front reaction and initiation of detonation in low density ammonium perchlorate

• Published in: Combustion and flame 1969-08, Vol.13 (4), p.375-391
• Main Authors: Seely, L.B., Blackburn, J.H., Reese, B.O., Evans, Marjorie W.
• Format: Article
• Language: English
• ISSN: 0010-2180; 1556-2921
• DOI: 10.1016/0010-2180(69)90107-2

Shock transit times in granular ammonium perchlorate at 1·0 g cm−3 bulk density have yielded a locus of shocked states that exhibits a region of apparently positive slope in the pressure/volume plante between 3 and 10 kilobars and shocked specific volumes greater than the normal value of the crystal between 10 and 70 kbar. Correspondingly, the locus is non-linear in the shock-velocity/particle-velocity plane. Thus, except for extremely weak shocks, ammonium perchlorate fails to follow the collapsing aggregate model, a behaviour which is ascribed to chemical reaction in initiating shocks. For potassium chloride aggregates, in which no reaction can occur, the shocked specific volumes are indistinguishably close to those of the solid crystal up to about 20 kbar. They are significantly larger than those of the crystal at about 40 kbar, but still less than the specific volume of the uncompressed crystal. The shocked specific volumes of potassium chloride at 40 kbar are about 30 per cent less than the shocked specific volumes of ammonium perchlorate aggregates at the same pressure. Because the shocked states are calculated from shock and particle velocities measured when the shock first enters the ammonium perchlorate aggregate, it is proposed that reaction takes place in the shock front. Hot spots capable of causing such localized reaction have been photographed in large particle-sized aggregates at nominal pressures below 15 kbar. Calculations under two sets of assumptions have reproduced the general shape and location of the locus of shocked states. Pressure measurements with manganin-wire gauges show a steady buildup in the initiating shock even though the velocity appears to change little as the transition to detonation is approached. This is ascribed to reaction in the non-steady flow behind the front.