Diffraction and transmission of a detonation into a bounding explosive layer

Abstract

The results of an experimental and analytical study of the propagation of a gaseous detonation past a bounding explosive layer are presented. Two adjacent 1.6 cm square detonation tubes, separated at the test section by a 50 nm thick cellulose film, were used to observe the interaction which occurs when a normal detonation in the primary gas comes into contact with a bounding explosive mixture. A pulsed argon ion laser and a high speed camera system were used to obtain Schlieren framing photographs of the interacting waves at 2μ sec intervals with an exposure time of 12 ns. Wave velocities and the pressure variation behind the incident detonation and the wave induced in the bounding gas were also determined. Experiments were made using a stoichiometric H 2 −O 2 mixture as the primary explosive, and using H 2 −O 2 mixtures with equivalence ratios ranging from 0.15 to 4.5 for the secondary bounding explosive. At the first instant of contact a bubble or blast wave was observed to propagate into the secondary explosive, and in some cases a micro-explosion in this bubble led to almost instant transition to oblique detonation. Otherwise an oblique shock is induced in the bounding explosive which is reflected from the shock tube wall. As the equivalence ratio of the bounding mixture increases, the reflection of the induced oblique shock changes from a regular to a Mach reflection, and in many cases a detonation is initiated behind the reflected wave. Shock polar analysis was used to compute the details of the interaction at the interface between the primary and secondary explosives. A simplified method for rapid computation of oblique detonation polars was developed for this purposes, and used to compute the conditions behind the induced oblique detonations and shock waves. There was good agreement between computed and measured shock angles; but computed oblique detonation angles, while showing the proper variation with the equivalence ratio, were always lower than the observed values. Calculation of induction lengths indicated that initiation was only possible behind reflected oblique shocks in agreement with experimental observation.