A short reaction mechanism for the combustion of dimethyl-ether
Introduction
Dimethyl-ether (DME: CH3OCH3) is an attractive alternative to conventional diesel fuel for compression-ignition (CI) engines because it auto-ignites favorably and burns with little soot formation [1], [2]. The main property of DME relevant to its engine compatibility is its high cetane number (CN>55), resulting in low auto-ignition temperatures along with its rapid vaporization. In addition, DME is an oxygenated hydrocarbon, with a low carbon-to-hydrogen ratio and the absence of a C–C bond, leading to very low emissions of particle matter (PM) during CI combustion. The main disadvantages of using DME in CI-powered vehicles are related to its viscosity, which is lower than that of diesel fuel, enhancing both leakage from fuel-supply systems and surface wear of moving parts within fuel-injection systems. An additional disadvantage of DME is its lower combustion enthalpy, compared with that of diesel, thereby requiring a larger injected volume to deliver the same amount of energy as that provided by diesel. Direct and indirect methods can be used to produce DME [3], [4]; direct synthetic methods make it directly from natural gas, while indirect synthetic methods generate it through a dehydration reaction after synthetic production of methanol.
Since DME is considered a clean alternative to diesel fuel, with notably low PM, for example, a combustion model to predict its ignition delays, flame propagation, and emission properties is critical to designing practical combustion devices for using this fuel optimally. Also, as a simple oxygenated fuel, DME provides a useful reference compound for comparison with other more complex oxygenated fuels, such as biofuels. In addition, having a low-temperature combustion (LTC) path, DME exhibits a negative-temperature-coefficient (NTC) phenomenon and two-stage ignition. This is of great interest in CI engines, where the LTC mode, combined with the other favorable DME properties, ultimately may enable cleaner and more efficient engines to be designed.
Several DME oxidation mechanisms, reported in the literature [5], [6], [7], [8], [9], [10], [11], have been tested against data from numerous types of experiments [5,[12], [13], [14], [15], [16], [17], [18], [19]], including pressure-flow reactors, jet-stirred reactors, shock tubes, rapid-compression machines, and direct sampling from flames. These chemical mechanisms have also been used for modeling additional combustion processes, mainly in flames or for ignition at high-pressure conditions [20], [21], [22], [23], [24], [25], [26], [27], [28], [29]. Some recent flow-reactor experiments at atmospheric conditions [30], [31], [32], [33], which provide new insight into the DME ignition delay, are, however, not modeled accurately by the above mechanisms. New reactions and rate parameters have to be used to update these chemical models [32], [34] to better describe these recent experimental data.
Because the mechanisms in these publications are too complex for several practical purposes, a need exists for deriving reduced mechanisms that can be validated against the wider range of experimental data that extends to atmospheric conditions. The present paper addresses this need by taking as a base mechanism the so-called San Diego mechanism (http://combustion.ucsd.edu), which has been designed to be a short mechanism, for use in applications in which larger mechanisms become impractical [35], [36].
Section snippets
The reaction mechanism
The approach to be adopted here takes into account recent chemical-kinetic studies of DME combustion processes, especially flow-reactor experiments under atmospheric conditions. It also makes use of applicable steady-state approximations and lumping procedures for reducing the size of the mechanism. In addressing DME oxidation at atmospheric conditions, literature on elementary reaction rates relevant to this condition [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48] was
Validation tests
The canonical testing ground for chemical-kinetic combustion mechanisms is the prediction of laminar burning velocities. For any mechanism to be successful, it must at least provide reasonably agreement with measured burning velocities, within experimental uncertainties. In a sense, this is not entirely fair because burning velocities depend on transport properties as well as the chemical kinetics, so disagreements can arise from improper transport descriptions. However, for fuel molecules as
Conclusions
This work has shown that by adding a relatively small set of reactions and species to the San Diego mechanism, without changing any reactions or rate parameters in the base mechanism, it is possible to model well the combustion of DME and its blends with C1–C3 alkanes. The approach takes into account extensive recent chemical-kinetic experiments, particularly studies of low-temperature combustion processes at atmospheric pressure, which are not described well by rate parameters available in the
Acknowledgments
JCP acknowledges support from UCSD through the Presidential Chair in Energy and Combustion Research and from Tecnológico Nacional de México that enabled him to spend a year at UCSD. FAW would like to acknowledge support from the US National Science Foundation through Grant number CBET-1404026.
References (57)
- et al.
J. Power Sources
(2006) - et al.
Fuel
(2008) - et al.
Catal. Today
(2005) - et al.
Chem. Eng. Process.
(2014) - et al.
Proc. Combust. Inst.
(1998) - et al.
Prog. Energy Combust. Sci.
(2012) - et al.
Proc. Combust. Inst.
(1996) - et al.
Combust. Flame
(2001) - et al.
Combust. Flame
(2004) - et al.
Proc. Combust. Inst.
(2005)