Douglas Schwer

Numerical Investigation of Rotating Detonation Engines

Rotating detonation engines (RDE’s) represent an alternative to the extensively studied pulse detonation engines (PDE’s) for obtaining propulsion from the high efficiency detonation cycle. Since it has received considerably less attention, the general flow-field and effect of parameters such as stagnation conditions, combustion chamber length, and fuel mixture on specific impulse are less well understood than for PDE’s. In this paper we develop a model for doing time-accurate calculations of RDE’s in two and three dimensions, using algorithms that have successfully been used for PDE simulations previously. Results are shown for a stoichiometric hydrogen-air RDE operating at 10 atm, 300 K stagnation premixture conditions and 1 atm back pressure. Conditions within the chamber are described as well as inlet and outlet conditions and integrated quantities such as total mass flow, force, and specific impulse. Further computations examined the role of back pressure and inlet stagnation pressure on performance. It was found that the specific impulse was dependent on pressure ratio, whereas the mass flow and propulsive force were primarily dependent on the stagnation properties of the inlet micro-nozzles. The specific impulse varied from 3845 sec to 5560 sec over a pressure ratio of 5 to 30. The specific impulse for the 10 atm stagnation pressure, 1 atm back pressure was 5130 sec.

Feedback into Mixture Plenums in Rotating Detonation Engines

Rotating detonation engines (RDE’s) represent a natural extension of the pulse detonation engine (PDE) concept to a continuous detonation for obtaining propulsion and power from the high efficiency detonation cycle. Although extensively studied for the last few years, RDE’s are still less understood and developed than the PDE concept. In particular, interaction between the detonation wave in the combustion chamber and the upstream mixture plenum is critical to the stability and performance of an RDE, and has not been studied in detail. In this paper, we extend our well-tested RDE models to examine the effect of pressure feedback into the mixture plenum with two and three-dimensional simulations. We examine three main injector configurations to determine their effect on stability, performance, and feedback into the mixture plenum. Results indicate that although the flow-field is considerably different, the performance of the RDE with all three modeled injection systems is close to that of assuming an ideal injection process, especially for a injector throat to injector face area ratio of 0.2. For an area ratio of 0.4, the drop in performance is more noticable, varying between 2.4% at 10 atm plenum pressure to 11.4% at 4 atm. Specific impulses for all the modeled injector cases varied from 3975 s to 5263 s. The stability at high and low pressures becomes much more of an issue when modeling specific injectors, due to an increase in flow unsteadiness from the injection. Pressure feedback into the mixture plenum appears to be relatively low for area ratios of 0.2 (less than 10% of stagnation pressure), but becomes very large for even moderate area ratios of 0.4. The cylindrical micro-injectors had the worst pressure feedback, with underpressures between 1 and 1.77 atm and overpressures between 3.09 and 3.83 atm with the large area ratio.

Numerical Study of the Effects of Engine Size on Rotating Detonation Engines

Rotating detonation engines (RDE’s) represent an alternative to the extensively studied pulse detonation engines (PDE’s) for obtaining propulsion from the high efficiency detonation cycle. Since it has received considerably less attention, the general flow-field and effect of parameters such as stagnation conditions, combustion chamber sizing, and fuel mixture on specific impulse are less well understood than for PDE’s. In this paper we use a model developed previously for doing time-accurate calculations of RDE’s in two and three dimensions to examine the effect of different engine sizing parameters on mass flow rate, performance, and thrust. Specific sizing parameters that are discussed are area ratio of micro-injectors to head-end wall, combustion chamber diameter and length. Additionally, several three-dimensional simulations for small combustion chambers with different thicknesses are shown. Results indicate that for many of these parameters, the characteristics of the engine scale in predictable ways for high plenum pressures. At lower plenum pressures, the results are more difficult to interpret. Specific impulses varied from 3300 s (low-pressure, large chamber length) to 5500 s.

Energy Transfer in a Rotating Detonation Engine

Rotating Detonation Engines (RDEs) have the potential to achieve the high propulsive efficiencies of detonation cycles in a simple and effective annular geometry. A thorough understanding of the thermodynamic cycle is required to achieve this potential. A one- dimensional detonation thermodynamic cycle is modified by the use of velocity triangles and compared to a time-averaged computational model. Rothalpy is introduced to explain the steady state processes in a rotating frame of reference. The derived thermodynamic cycle is in good agreement with the CFD model, and provides a means to explain the energy transfer mechanisms in the RDE.

Modeling Exhaust Effects in Rotating Detonation Engines

Rotating detonation engines (RDEs) represent an alternative to the extensively studied pulse detonation engines (PDEs) for obtaining propulsion from the high efficiency detonation cycle. The current work focuses on the effect of adding an exhaust plenum to the RDE simulation. The addition of an exhaust plenum causes only minimal changes in the RDE flow field, most significantly at the exhaust plane. The specific impulse varied by less than 1% between modeling with and without an exhaust plenum, showing that using a mixed boundary condition for the exhaust instead of an exhaust plenum is valid. The exhaust plume for three different RDEs is also described. The highest plume temperatures are found after the circulation zone behind the centerbody, although the temperature profiles vary greatly depending on the RDE size. Finally, the effect of a simple conical nozzle extending from the centerbody on the exhaust is investigated. This nozzle reduces the size of the recirculation zone and reduces the temperature in the plume, however, it has very little effect on the flow field inside the RDE or at the exhaust plane. For the hydrogen/air RDE with 10 atm injection pressure and 1 atm back pressure, the specific impulse varied from 4930 s to 4980 s for all cases examined in this paper.

A Hybrid Grid Compressible Flow Solver for Large-Scale Supersonic Jet Noise Simulations on Multi-GPU Clusters

ow simulations. Supersonic jet noise simulations require the accurate representation of complex nozzle geometry and thus the use of unstructured grids. While such a grid representation is crucial for the accurate representation of the nozzle geometry, it has the disadvantage of introducing a highly irregular memory access pattern, which violates GPU coalescing requirements resulting in an inecient use of the otherwise high memory bandwidth of GPUs and therefore a bottleneck in computational performance. In order to mitigate this performance bottleneck, a hybrid grid representation is implemented which allows for augmenting the unstructured grid representation in the vicinity of complex nozzle geometry with an ecient structured grid representation in other regions of the ow domain, which is able to fulll GPU coalescing requirements, and thus achieve a signicant improvement in computational performance, leading to a reduction in simulation turnaround time.

Direct Comparison of Particle-Tracking and Sectional Approaches for Shock Driven Flows

Dispersed-phase flows are important for a wide variety of problems, and several numerical approaches for the solution of dispersed-phase flows have been proposed and implemented in the past. The present research implements two popular approaches to dispersed-phase flows: the Lagrangian particle-tracking approach and the Eulerian sectional approach. A direct comparison between the two methods is made for a range of shock driven seeded flow-fields. First, different drag models are investigated using the particle-tracking method for a range of conditions, and then direct comparisons between the two methods are made for shock speed attenuation and shock-wave profiles. In addition, resolution requirements are investigated to determine the number of sections and the number of particles required to obtain good agreement between the methods, and then two-dimensional simulations are done to investigate the effect of each method on more complicated flow-fields. Results showed both methods can be used to obtain very similar results, although each method has benefits and drawbacks. The glass particles were then replaced with water droplets, and the effect of vaporization and droplet breakup were then investigated. Although vaporization was well represented with the sectional approach, different droplet breakup models had to be implemented for the different approaches, with some significant differences in the resultant droplet distributions. The reason for this is that breakup models require a droplet deformation time before breaking up, and thus a droplet history. This droplet history is difficult to implement in sectional approaches (and Eulerian methods in general), and so the breakup model must be changed. Similar profiles could be reproduced with the sectional method, but significant differences persisted. The results did show, however, that the Eulerian sectional approach is a viable method for computing complex, multi-dimensional flow-fields and can provide significant numerical advantages when compared with Lagrangian particle-tracking methods, especially in flooded environments such as examined here.