Reactant Mixing Effects in Rotating Detonation Engines

Rotating detonation engines (RDEs) represent a promising technology for achieving a step change improvement in the efficiency of gas turbine and rocket engines for propulsion and power generation applications. However, the complexity of the rotating detonation cycle presents a challenge for researchers and the effects of all the interdependent mechanisms are not yet well understood. In particular, the reactant mixing process appears to be a predominant facet of cycle dynamics that can have significant effects on engine behavior and performance. The small spatial and temporal scales that govern the mixing process creates difficulty for computational modeling efforts and experimental works have been unable to characterize mixing effects due to limitations imposed by measurement capabilities and the harsh environment of the combustion chamber. In an effort to improve understanding of reactant mixing effects in rotating detonation engines, a series of experiments were conducted in operational detonation engines using conventional and state-of-the-art measurement tools.

    Simultaneous 100 kHz acetone planar laser-induced fluorescence (PLIF) and OH* chemiluminescence were applied in a linear detonation channel that emulates RDE dynamics to visualize reactant mixing and heat release processes. Detonation waves were observed to propagate through highly stratified mixtures and poor mixing in the injector near field caused a weakening of the detonation. The lower heat release rates at the base of the detonation wave allowed unburned fuel to survive for 17% of the cycle period before being consumed in a deflagrative manner, likely reducing the performance of the device. Fuel recovery times were measured and were found to lie between 14-18% of the cycle period as a function of detonation cycle time and fuel manifold pressure. A mismatch between recovery times was shown to prevent reactant mixing and to create a non-reacting layer at the top of the detonation wave, lowering combustion efficiency and highlighting the need to design for equivalent injector recovery times.

     A variably premixed rotating detonation engine was then developed that allows for modulation of mixing quality from non-premixed to fully premixed conditions. Initial testing revealed low detonation wave speeds and unstable behavior until a product gas recirculation zone was induced near the injector face by the addition of a backward-facing step. The step injector demonstrated stable premixed operation at 86% of the theoretical Chapman-Jouguet velocity. Comparisons with a simple two-dimensional computational model revealed a high level of agreement and significant reactant heating and vitiation in the refill zone, likely due to the product recirculation zone.

     More comprehensive testing of an improved design continued to show poor operability at near-atmospheric chamber pressure unless a backward-facing step injection scheme was used. Detonation wave speeds were observed to increase by up to 16% relative to the Chapman-Jouguet speed as premixing was increased from 0 to 60\%. When the chamber pressure was raised above two atmospheres, operability trends indicated that improved mixing conditions in concert with product gas recirculation and increased kinetic rates served to spawn multiple detonation waves, leading to cycle failure in the extreme. Simultaneous 100 kHz OH PLIF and OH* chemiluminescence were then implemented in the variably premixed RDE to image the evolution of the heat release field in three dimensions. Under various premixing conditions, the results show that a poor reactant mixing field can lead to highly lifted detonation waves with multiple heat release zones. OH PLIF images of the chamber revealed significant product gas recirculation in the wake of the backward-facing step. Interaction between incoming reactants and recirculating products induced refill zone deflagration that altered the detonation wave structure. Surprisingly, refill zone deflagration was found to decrease with increasing premixing as the momentum and recovery time of the non-premixed fuel jets were found to be the controlling factors of deflagration levels. Reactant recovery times were approximated using the PLIF images and found to be between 15-23% of the detonation cycle, in line with prior literature.

     These studies have improved understanding on the effects of reactant mixing on detonation structure, reactant deflagration, cycle frequencies, and global operability in rotating detonation engines. The observations of alternate processes, such as product gas recirculation and injector dynamics, have also advanced comprehension of the interdependence between multiple facets of the RDE cycle.