DETONATION PROPAGATION IN A ROTATING DETONATION ENGINE ANALOGUE WITH NONPREMIXED FUEL-OXYGEN INJECTION

Abstract

In industrial explosions, the combustible mixture often accumulates on a ceiling, or forms on the ground. An investigation into detonation propagation through a combustible gas layer in a narrow channel bounded by an inert gas was undertaken using schlieren video and the soot foil technique to obtain the detonation cell structure. A gas layer was generated as a premixed hydrogen-oxygen-nitrogen gravity current over an inert gas, or by injecting hydrogen and oxygen separately into the inert gas from a plenum located at the top of the channel. Both argon and nitrogen were used as the inert gas. For the gravity driven layer tests, the layer development time was varied to control the diffusion of the inert gas into the layer. For times of 1.5 s and less, the detonation propagated through most of the layer at the local argon-diluted theoretical detonation velocity; where a CFD simulation carried out by a lab colleague predicted the argon distribution. For longer times, where significant dilution occurred at the leading edge of the gravity wave, the detonation failed before the end of the layer. Detonations through a layer of reduced reactivity were unstable, with progressively earlier premature detonation failure characterized by decoupling and reinitiation. Nitrogen was shown to affect the detonation propagation more than argon, as it produced a critical propagation distance corresponding to 1.5 s, above which time the detonation did not propagate any further. A detonation failure criterion based on the calculated induction zone length was found to predict the extent of detonation propagation recorded on the soot foil records and was especially accurate along the bottom diffuse interface of the layer. The jet layer tests successfully produced a detonation propagating through the entire layer. Based on the “chugging” of the detonation front, and the nonuniform, patchy appearance of the cell structure, it was demonstrated that there was a significant mixing issue, with detonable gas pockets formed leading to unsteady propagation. Asymmetry in the soot foil records from the two sides of the channel demonstrated that the high momentum oxygen jets dominated the hydrogen jets, such that most of the combustible mixture ended up on the hydrogen plenum-side window.