A Critical Review of Scramjet Combustion Simulation (Invited)

Abstract

A review of scramjet combustion simulation is provided in this paper. The topics covered include the fundamental problem of supersonic mixing layers, high-speed combustion modeling efforts, and actual calculations of realistic scramjet combustors. The review shows that the RANS approach dominates the turbulence modeling of the system, with only a handful of LES work. Also, virtually all the numerical works are based on low-order schemes, and the combustion models that have been used for realistic simulations solve the species evolution equations with assumed PDF closures, although there seems to be a growing use of the flamelet methods. Spray modeling for scramjet combustion has not received enough attention.2 for high-speed combustion applications, emphasizing on the salient features and shortcomings of the averaged equation set. Some of the models in Ref [1] have been implemented in VULCAN, a widely-used, multi-grid, fluxdifference-split, finite-volume code, developed by the Air Force and NASA for high-speed (ramjet, scramjet) reacting flow simulation. The precise models and solution procedure in this code are documented in White and Morrison. 2 Tishkoff et al. 3 presents the state of supersonic combustion research, including modeling and simulation, as an outcome of a joint AFRL/NASA meeting in May of 1996, in which the state-of-the-art in hydrocarbon and/or hydrogen-fueled scramjet research was examined, with suggestions for the future direction and needs of basic research in support of scramjet technology.Other relevant reviews, albeit applicable mostly to low-speed combustion, include Givi, 4 who provides a review of the state-of-the-art in subgrid scale modeling as required for large-eddy simulation (LES) of turbulent combustion. Closure complexities caused by chemical reactions are the focus, while Givi 5 presents a review devoted primarily to subgrid scale (SGS) closure based on the filtered density function (FDF), which is a method that is analogous to the probability density function (PDF) modeling. A more recent review of the FDF method is provided in Ref [6]. Heinz 7 highlights the fundamental differences between the Reynolds-averaged Navier-Stokes (RANS) and LES combustion models for premixed and non-premixed turbulent combustion.Finally, Grinstein 8 addresses modeling issues relevant to CFD of turbulent non-premixed jet flames, including subgrid and supergrid modeling. Transitional jet diffusion flames of the hydrogen/air and propane/air types are reviewed, while the jets studied involve laminar initial conditions, preferential diffusion effects, weak axial forcing, negligible streamwise vorticity, and negligible azimuthal non-uniformities, as well as the impact of aspect ratiodependent vortex topological and dynamical features on the development of the jet diffusion flame. Models for turbulence, chemical reactions, volume expansion, and heat release are discussed, as are models for the dominant features of the couplings between the various phenomena.Unlike the foregoing reviews, which either did n...

A flamelet model for turbulent diffusion combustion in supersonic flow

“…Because of these advantages, the flamelet model is fairly popular in low-mach-number turbulent combustion flow [6] and becomes one of the most appropriate model to consider the interaction between turbulence and combustion. In a review paper about numerical simulation of combustion in scramjet chamber [7], Ladeinde pointed out that the flamelet model has become a research highlight on the modeling of turbulent diffusion combustion in supersonic flow. Though a few researchers have attempted to apply the flamelet model in numerical simulation of supersonic turbulent combustion flow [8,9], the physical foundation of the extension of the flamelet model to supersonic flow has not been discussed in detail, and the experimental data used for validation in these works were not sufficient.…”

Numerical Simulation of Hydrogen Jet Injection and Ignition in Supersonic Flow

“…Therefore, at the first stage, several kinetic schemes were tested [25][26][27]. Calculations were provided in terms of experimental data [28] where hydrogen was injected into a supersonic flow through a tube of d 1 = 9.5 mm diameter.…”

“…This is followed by LES, whereas DNS is still too costly for realistic engineering problems. However, the success of the RANS approach is dependent on type and complexity of flow in combustion chamber, and the procedure needs to be verified for every class of problem, making it non-universal [23]. Moreover, the approach is inherently steady and cannot deal with unsteady large-scale structures that determine the dynamics of many important flow problems.…”

Theoretical analysis of flamelet model for supersonic turbulent combustion

“…The flamelet model is developed based on the low-Mach number assumption and has been extensively used in low-speed flows. In his review of the numerical simulations of the combustion inside a scramjet, Ladeinde [10] pointed out that the flamelet model may become the main direction in the future numerical simulations of supersonic combustion due to its unique advantages. Although the flamelet model has been used in simulation of supersonic combustion, the rationality of using the flamelet model to simulate the supersonic combustion has never been analyzed or proved systematically.…”