The mechanism of detonation attenuation by a porous medium and its subsequent re-initiation

Abstract

The attenuation and re-initiation mechanism of detonations transmitted through a porous section consisting of a two-dimensional array of staggered cylinders was investigated experimentally and numerically for acetylene–oxygen mixtures. It was found that the leading order attenuation mechanism is the wave diffraction around the cylinders. The local re-amplification permitting the self-propagation of the wave was due to wave reflections from adjacent obstacles. The critical conditions for transmittance of a detonation wave were found to correspond approximately to a pore size equal to approximately 30–60 detonation induction lengths, or one to two cell sizes. For quenched detonations, the re-initiation mechanism was found to rely on wave reflections from neighbouring pores. Depending on the mixture sensitivity, one or several shock reflections may be necessary to re-amplify the attenuated detonation wave back to a self-sustained wave. For the latter case, a novel mechanism was identified, where each shock reflection gives rise to a significant enhancement of the gas reactivity and burnout of large portions of unreacted gas. This leads to a slow acceleration of the leading front, punctuated by small-scale local sudden re-accelerations. The resulting wave interactions give rise to a topologically complex reaction zone structure consisting of alternating layers of reacted and unreacted gas. The role of turbulent diffusive burning during this transient is discussed.

Influence of hydrodynamic instabilities on the propagation mechanism of fast flames

“…Upon the next reflection of the associated transverse waves, new Mach shocks form, which in turn give rise to a second generation of hot spots. This sequential initiation of hot spots, without the establishment of a detonation, is consistent with the mechanism deduced by Radulescu & Maxwell from their simulations and previous experiments in acetylene-oxygenargon mixtures performed by Lee and Papyrin, reported in [6]. The second generation of hot spots is unable to amplify the shock waves enough to sustain them.…”

“…It is difficult to account for such high deflagration speeds based on turbulent deflagration velocities; experiments in fan-stirred bombs show that turbulent velocities eventually saturate as the turbulent intensity is increased [4]. It is also difficult to account for these waves as a diffusionless phenomenon alone, since the auto-ignition delay in the gas compressed by the shock is not compatible with the reaction wave propagation, unless strong temperature fluctuations are present to trigger local ignition kernels (hot spots) with ignition delays compatible with the observed reaction wave propagation [5,6]. Based on indirect and often speculative experimental observations, it is believed that these fast deflagration velocities are maintained via the action of pressure waves within this deflagration complex (see [3] and references to earlier work by Lee and his collaborators [7]).…”

“…Ohyagi et al and Teodorczyk et et al have focused on the conditions leading to the prompt re-initiation of the detonation from Mach reflections [11,12]. Radulescu and Maxwell [6] investigated the re-initiation of detonation waves downstream of a porous medium, which comprised a two-dimensional array of staggered cylinders. They reported that such high-speed deflagrations require auto-ignition spots via shock compressions to drive strong pressure wave activity.…”

“…Open shutter photograph illustrating the reactions along the trajectories of shear layers and the eventual onset of detonation in acetylene-oxygen, adapted from[6].…”

High resolution numerical simulation of triple point collision and origin of unburned gas pockets in turbulent detonations

“…Even with this simple detonation model, the simulations are in remarkable qualitative agreement with the experiments (e.g. [16,29,30,32,34,35,58]). A small number of simulations have been carried out with detailed chemical reaction mechanisms using the Euler formulation for hydrogen-oxygen-argon mixtures in two-dimension [12,14,59].…”