Elsevier

Combustion and Flame

Volume 162, Issue 9, September 2015, Pages 3437-3445
Combustion and Flame

Autoignition-affected stabilization of laminar nonpremixed DME/air coflow flames

https://doi.org/10.1016/j.combustflame.2015.06.007Get rights and content

Abstract

The structure and stabilization mechanism of laminar nonpremixed autoignitive DME/air coflow flames were investigated. Computations were performed at 30 atmospheres with uniform inlet velocities of 3.2 m/s for both streams, and the coflow air boundary temperatures were 700, 800, 900, and 1100 K. The heat release rate and species profiles were examined for each case. Further investigation with Chemical Explosive Mode Analysis (CEMA) and Lagrangian Flamelet Analysis (LFA) were performed to identify the controlling chemistry and elucidate the dominant combustion mode and stabilization mechanism. At 700–900 K, autoignition was observed to be the dominant stabilization mechanism, and NTC chemistry determines the stabilization point in mixture fraction space. Conversely, at 1100 K, the kinematic balance between the premixed flame propagation velocity and the incoming flow velocity becomes the dominant stabilization mechanism, and the classical triple flame structure was observed. Extended stabilization regimes, in terms of increasing boundary temperature, are therefore identified, including frozen flow, kinetically stabilized, autoignition–propagation-coupled stabilized, kinematically stabilized, and burner stabilized regimes.

Introduction

Nonpremixed jet flames have been extensively studied to understand the combustion processes in rocket and diesel engines. The stabilization and structure of jet flames determine the lift-off height of the flame and are therefore integral to engine design. Due to the mixing process of the fuel and oxidizer streams in lifted flames at nonautoignitive conditions, the combustion mode is partially premixed, leading to the observation of a two-dimensional tribrachial flame (also known as triple flame) [1]; specifically, a lean and a rich premixed flame wing with a trailing diffusion flame branch. The point where the three branches intersect is called the triple point and is generally considered to be the stabilization point for nonautoignitive situations. The dynamic balance between the local flame propagation speed and the incoming flow speed is characterized as the stabilization mechanism. A recent review by Chung [2] discussed the stabilization, propagation, and instability of tribrachial flames, including the effects of concentration gradient [3], [4], [5], velocity gradient [6], and burned gas expansion [7], [8], [9], [10]. These studies, however, were limited to nonautoignitive conditions, but real engines are operated at elevated pressures and temperatures, where autoignition is activated and could interact with the tribrachial flame.

Chung and co-workers [11], [12], [13] further conducted a series of experiments to investigate the autoignition characteristics of laminar C1 to C4 fuel jets in a heated air coflow and found that, above certain coflow temperatures, lifted flames could be established through autoignition. In these studies, both the tribrachial structure for most autoignited cases and a repetitive behavior of extinction and reignition at the critical condition near blowout were observed. However, the role that autoignition plays in the stabilization mechanism as well as its influences on the tribrachial flame structure are still less understood.

Furthermore, practical hydrocarbon-based fuels generally have two-stage ignition processes, in which the first stage ignition is governed by low temperature chemistry and the second stage ignition by high temperature chemistry. In both low and high temperature regimes, the ignition delay time decreases as the initial temperature increases. However, in the intermediate temperature regime, the transition of the ignition chemistry results in increased overall ignition delay time as the initial temperature increases, exhibiting the negative temperature coefficient (NTC) phenomena, which has been extensively studied in homogeneous systems as a major feature of large hydrocarbon autoignition [14]. For engine applications, however, the coupling between NTC chemistry and transport processes should be considered, for nonuniformities invariantly exist in realistic combustion systems. When the transport time scale becomes comparable to that of the NTC chemical time scale, the two processes are expected to be strongly coupled. As a consequence, the global response of the inhomogeneous system can also be affected by NTC chemistry. Recently, a series of computational and experimental studies adopting the nonpremixed counterflow configuration by Law and co-workers [15], [16], [17] have demonstrated that, with the existence of nonuniformities in the flow, species, and temperature fields, the ignition characteristics of nonpremixed flames can be fundamentally affected by NTC effects, especially at elevated pressures and/or reduced strain rates.

Therefore, NTC-affected stabilization of nonpremixed lifted jet flames can be potentially important, yet few literatures provide detailed analysis. Krisman et al. [18] recently conducted a numerical study of dimethyl ether (DME)/air mixing layer at 40 atmospheres and air coflow temperatures ranging from 700 to 1500 K and observed multibrachial structures in the heat release rate profiles. The mixture fractions corresponding to the stabilization points defined based on the hydroxyl radical (OH) mass fraction and the first stage autoignition kernels based on the methoxymethylperoxy radical (CH3OCH2O2) were compared with the most reactive mixture fractions computed from homogeneous autoignition under the same initial conditions. A transport budget analysis based on selected species was performed to differentiate deflagration from autoignition.

In light of the reported multibrachial structure, showing a modified flame shape from autoignition in the mixing layer, further investigation is warranted to identify the detailed chemical structure and stabilization mechanism of the multibrachial flame. For example, tools for computational diagnostics, especially for identifying locally dominant chemical reactions, can be employed to understand the controlling chemistry. Moreover, a direct comparison to homogeneous autoignition is insufficient to understand the transport processes in the current configuration. In the two-dimensional mixing layer, transport processes in two directions are important: parallel and normal to the mixture fraction gradient, which are due to transverse stratification of temperature and species and streamwise flow and (flame back) diffusion, respectively. These considerations would significantly improve the understanding of the role of autoignition upstream of the flame structure and quantitatively identify the controlling kinetics and stabilization mechanism.

In the present study, nonpremixed DME/air coflow flames were computed at 30 atmospheres with the oxidizer stream heated to activate autoignition. With fixed inlet velocities, only the oxidizer stream boundary temperature was varied to investigate the corresponding lifted flame morphology, chemical structure, and dominant reaction pathways. In the following we shall first present the computational details of the study. The thermal and chemical structures are then described with heat release rate and selected species profiles in Section 3. The evolution of the controlling chemical pathways are subsequently identified with Chemical Explosive Mode Analysis (CEMA) and the stabilization mechanism determined with Lagrangian Flamelet Analysis (LFA) in Section 4. Finally, the transition of the dominant stabilization mechanism is analyzed in Section 5, with extended stabilization regimes constructed for completeness.

Section snippets

Computational details

The flow configuration is an axisymmetric DME stream at 300 K in a heated coflow of air (700, 800, 900, and 1100 K) at 30 atmospheres. The fuel nozzle diameter D is 0.8 mm, and the fuel and air are initially separated with an adiabatic, no-slip wall with thickness D/20. The coflow outer boundary is specified as an adiabatic slip wall, and its diameter is large enough such that increasing the width of the domain does not influence the computation. Uniform inlet velocities of 3.2 m/s were specified

Thermal and chemical structure

To visualize the flame structures, the heat release rate profiles for the four cases (700, 800, 900, and 1100 K) are shown in Fig. 2. Qualitatively, the most upstream point on the largest heat release contour (the leading point), colored1 by red, will be referred to as the stabilization point.

At 700 K, a tribrachial thermal structure is observed, and the stabilization point is located around Z=0.15,

Computational diagnostics and analysis

The above heat release rate and species profiles demonstrate the thermal and chemical structure of the reacting fronts at different boundary temperatures. However, more detailed computational diagnostics and analysis are needed to further demonstrate the controlling chemistry and the stabilization mechanism.

Stabilization mechanism

With the above analysis based on species profiles, Chemical Explosive Mode Analysis, and Lagrangian Flamelet Analysis, the transition of the stabilization mechanism and the coupling between autoignition chemistry and flame propagation can be clearly identified. In the current study, two fundamental stabilization mechanisms are relevant: the kinetic stabilization mechanism, due to the balance between the autoignition delay time and flow residence time, and the kinematic stabilization mechanism,

Conclusions

In the present study, two-dimensional nonpremixed DME flames in heated air coflows were computed. The computations were conducted at 30 atmospheres to observe the influence of NTC chemistry on the stabilization mechanism. A uniform and fixed inlet boundary velocity was specified, and four coflow temperature (700, 800, 900, and 1100 K) cases were studied.

The heat release rate profile and characteristic species profiles for low and high temperature chemistry, autoignition, and premixed flame

Acknowledgments

This research was supported in part by the Air Force Office of Scientific Research (AFOSR) under the technical management of Dr. Mitat Birkan.

References (40)

  • J. Buckmaster

    Prog. Energy Combust. Sci.

    (2002)
  • S.H. Chung

    Proc. Combust. Inst.

    (2007)
  • J.W. Dold

    Combust. Flame

    (1989)
  • M.K. Kim et al.

    Proc. Combust. Inst.

    (2007)
  • B.J. Lee et al.

    Combust. Flame

    (1997)
  • T. Plessing et al.

    Combust. Flame

    (1998)
  • P.N. Kı̃oni et al.

    Combust. Flame

    (1999)
  • B.C. Choi et al.

    Combust. Flame

    (2009)
  • B.C. Choi et al.

    Combust. Flame

    (2010)
  • B.C. Choi et al.

    Combust. Flame

    (2012)
  • J. Zádor et al.

    Prog. Energy Combust. Sci.

    (2011)
  • C.K. Law et al.

    Combust. Flame

    (2012)
  • P. Zhao et al.

    Combust. Flame

    (2013)
  • S. Deng et al.

    Combust. Flame

    (2014)
  • A. Krisman et al.

    Proc. Combust. Inst.

    (2015)
  • H. Pitsch et al.

    Combust. Flame

    (1998)
  • X.L. Zheng et al.

    Proc. Combust. Inst.

    (2005)
  • X. Qin et al.

    Proc. Combust. Inst.

    (2005)
  • A. Bhagatwala et al.

    Proc. Combust. Inst.

    (2015)
  • O. Desjardins et al.

    J. Comput. Phys.

    (2008)
  • Cited by (0)

    View full text