Numerical resolution of pulsating detonation waves

Hwang, Paul S; Fedkiw, Ronald; Merriman, Barry; Aslam, Tariq D.; Karagozian, Ann; Osher, Stanley

doi:10.1088/1364-7830/4/3/301

Cited by 70 publications

(46 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The problem of determining a computational domain width suitable to investigate periodic detonation structures is a nontrivial one. Using a 1-D simulation of a pulsating overdriven detonation wave in a mixture with E a ∕RT 0 50, Q∕RT 0 50, and γ 1.2, Hwang et al [44] reported the minimum domain length of L 75 HRL. This was required to keep the flow properties unaffected by the truncation of the computational domain.…”

Section: Boundary and Initial Conditions And Grid Resolutionmentioning

confidence: 99%

Hydrodynamic Instabilities in Gaseous Detonations: Comparison of Euler, Navier–Stokes, and Large-Eddy Simulation

Mahmoudi

Karimi

Deiterding

et al. 2014

Journal of Propulsion and Power

View full text Add to dashboard Cite

A large-eddy simulation is conducted to investigate the transient structure of an unstable detonation wave in two dimensions and the evolution of intrinsic hydrodynamic instabilities. The dependency of the detonation structure on the grid resolution is investigated, and the structures obtained by large-eddy simulation are compared with the predictions from solving the Euler and Navier-Stokes equations directly. The results indicate that to predict irregular detonation structures in agreement with experimental observations the vorticity generation and dissipation in small scale structures should be taken into account. Thus, large-eddy simulation with high grid resolution is required. In a low grid resolution scenario, in which numerical diffusion dominates, the structures obtained by solving the Euler or Navier-Stokes equations and large-eddy simulation are qualitatively similar. When high grid resolution is employed, the detonation structures obtained by solving the Euler or Navier-Stokes equations directly are roughly similar yet equally in disagreement with the experimental results. For high grid resolution, only the large-eddy simulation predicts detonation substructures correctly, a fact that is attributed to the increased dissipation provided by the subgrid scale model. Specific to the investigated configuration, major differences are observed in the occurrence of unreacted gas pockets in the high-resolution Euler and Navier-Stokes computations, which appear to be fully combusted when large-eddy simulation is employed.

show abstract

Section: Boundary and Initial Conditions And Grid Resolutionmentioning

confidence: 99%

Hydrodynamic Instabilities in Gaseous Detonations: Comparison of Euler, Navier–Stokes, and Large-Eddy Simulation

Mahmoudi

Karimi

Deiterding

et al. 2014

Journal of Propulsion and Power

View full text Add to dashboard Cite

show abstract

“…In fact, it can be shown (though it is not done in this work) that if the length scale in the Figure 1 profiles is changed from the tube length to the so-called reaction halfwidth (the distance from wherever the reaction commences to where half of the reactant mass fraction has been consumed), the K0=300 and 3000 profiles are identical (Ref. 21). Not surprisingly then, the high fidelity numerical schemes of References 12 and 13, along with countless PDE experiments show that in tubes (i.e., planar detonations), as long as detonations are achieved, they travel at the CJ speed (Ref.…”

Section: Reaction Rate Effects In a One-dimensional Tubementioning

confidence: 91%

“…A simple reaction model of this type precludes capturing the cellular detonation structure, or the "galloping" (i.e., pulsating) planar detonation phenomenon necessary for its development (Ref. 21). Nevertheless, it captures effects within the flowfield that are germane to this study.…”

Section: Simulation Descriptionmentioning

confidence: 99%

Examination of Wave Speed in Rotating Detonation Engines Using Simplified Computational Fluid Dynamics

Paxson

2018

2018 AIAA Aerospace Sciences Meeting

View full text Add to dashboard Cite

A simplified, two-dimensional, computational fluid dynamic (CFD) simulation, with a reactive Euler solver is used to examine possible causes for the low detonation wave propagation speeds that are consistently observed in air breathing rotating detonation engine (RDE) experiments. Intense, small-scale turbulence is proposed as the primary mechanism. While the solver cannot model this turbulence, it can be used to examine the most likely, and profound effect of turbulence. That is a substantial enlargement of the reaction zone, or equivalently, an effective reduction in the chemical reaction rate. It is demonstrated that in the unique flowfield of the RDE, a reduction in reaction rate leads to a reduction in the detonation speed. A subsequent test of reduced reaction rate in a purely one-dimensional pulsed detonation engine (PDE) flowfield yields no reduction in wave speed. The reasons for this are explained. The impact of reduced wave speed on RDE performance is then examined, and found to be minimal. Two other potential mechanisms are briefly examined. These are heat transfer, and reactive mixture nonuniformity. In the context of the simulation used for this study, both mechanisms are shown to have negligible effect on either wave speed or performance.

show abstract

“…Example 5.3. In this example, we considered a detonation case with more complex waves [4,6,7,44] . And the Heaviside model with following parameters are used, 2 , 50 , 230 .…”

Section: Simplified Reactive Euler Systemmentioning

confidence: 99%

The dual information preserving method for stiff reacting flows

et al. 2017

View full text Add to dashboard Cite

a b s t r a c tIn this paper, we construct a new numerical method to solve the reactive Euler equations to cure the numerical stiffness problem. The species mass equations are first decoupled from the reactive Euler equations, and then they are further fractionated into the convection step and the reaction step. In the convection step, by introducing two kinds of Lagrangian points (cell-point and particle-point), a dual information preserving (DIP) method is proposed to resolve the convection characteristics. In this new method, the information (including the transport value and the relative coordinates to the center of the current cell) of the cell-point and that of the particle-point are updated according to the velocity field. The information of the cell-point in a cell can effectively restrict the incorrect reaction activation caused by the numerical dissipation, while the information of the particle-point can help to preserve the sharp shock front once the strong shock waves are formed. Hence, by using the DIP method, the spurious numerical propagation phenomenon in stiff reacting flows is effectively eliminated. In addition, a numerical perturbation method is also developed to solve the fractional reaction step (ODE equation) to improve the stability and efficiency. A series of numerical examples are presented to validate the accuracy and robustness of the new method.

show abstract

Numerical resolution of pulsating detonation waves

Cited by 70 publications

References 18 publications

Hydrodynamic Instabilities in Gaseous Detonations: Comparison of Euler, Navier–Stokes, and Large-Eddy Simulation

Hydrodynamic Instabilities in Gaseous Detonations: Comparison of Euler, Navier–Stokes, and Large-Eddy Simulation

Examination of Wave Speed in Rotating Detonation Engines Using Simplified Computational Fluid Dynamics

The dual information preserving method for stiff reacting flows

Contact Info

Product

Resources

About