The autoignition of Liquefied Petroleum Gas (LPG) in spark-ignition engines

https://doi.org/10.1016/j.proci.2014.06.070Get rights and content

Abstract

This paper investigates the autoignition of C3/C4 hydrocarbon mixtures in a CFR octane rating engine. The four species examined – propane, propylene (propene), n-butane and iso-butane – are the primary constituents of Liquefied Petroleum Gas (LPG), and are also important intermediates in the oxidation of larger hydrocarbons. In-cylinder pressure data was acquired for both autoigniting and non-autoigniting engine operation at the same test conditions. The latter was used to calibrate a two-zone model of the CFR engine in a prior work, thus enabling the inclusion of the unburned charge chemical kinetics for further examination in this paper. The in-cylinder heat transfer and residual gas composition are both shown to affect autoignition significantly. In particular, physically reasonable concentrations of nitric oxide (NO) are found to be a strong promoter of autoignition in almost all cases, in keeping with several, more fundamental studies. The inclusion of NO in the residual gas is also required to obtain good agreement between the measured and modelled autoignition timing. This in turn suggests that kinetic interaction between hydrocarbon fuels and NO plays a vital role in octane rating, and its inclusion is important when modelling the autoignition of hydrocarbons in spark-ignition engines more generally.

Introduction

Liquefied Petroleum Gas (LPG) is a mixture of variable content, but is primarily composed of propane, propylene (propene), n-butane and iso-butane [1]. LPG has several attractive features as an alternative fuel. First, LPG vehicles can have lower emissions of both regulated pollutants and greenhouse gases than gasoline and diesel vehicles [2], [3]. When stored in liquid form, LPG also has an energy density that is comparable to other liquid transport fuels. Finally, LPG is usually of relatively low cost.

There are nonetheless only a few studies of LPG’s autoignition and knock in engines. These studies most commonly utilise correlation or reduced chemistry, e.g. [4], [5], [6], and so provide limited insight into the general autoignition problem. Further, since the components of LPG are intermediates in the oxidation of larger hydrocarbons, an improved understanding of their autoignition contributes to our understanding of autoignition more generally.

The aim of this paper is therefore to model the autoignition of arbitrary LPG blends in a spark-ignition engine. The autoignition modelling is undertaken using a kinetic model derived from those of Healy et al. [7] and Dagaut and Dayma [8]. This kinetic model is coupled with a two-zone engine model that was validated for non-autoigniting operation in a prior work [9]. The in-cylinder heat transfer and residual gas composition are both shown to affect autoignition significantly. In particular, nitric oxide (NO) is found to be a strong promoter of autoignition, in keeping with several, more fundamental studies. Indeed, the inclusion of NO in the residual gas at physically reasonable concentrations enables good agreement between the measured and modelled autoignition timing in almost all cases.

Section snippets

Experimental methods

The experimental data presented in this paper was acquired with a standard Cooperative Fuel Research (CFR) engine operated in accordance with the ASTM Research method (RON) test condition [10]. The standard engine was equipped with a gaseous fuel system that prepared mixtures of propane, propylene, n-butane and iso-butane. Each fuel was examined at the air–fuel ratio (λ) that registered the maximum Knock Intensity (KI) value on the ASTM Knockmeter unit, as summarised in Table 1. The compression

CFR engine model

The two-zone model of the CFR engine was developed using GT-Power [15]. The engine model incorporated the combustion chamber, together with sub-models of the intake and exhaust systems and the gas exchange processes. The full engine model was validated for non-autoigniting operation in a prior work [9].

In-cylinder heat transfer sub-model

The in-cylinder convective heat transfer coefficient was modelled using the Woschni correlation without swirl [16],hc=ζB-0.2p0.8T-0.55w0.8.The term ζ denotes a convective heat transfer

Effect of in-cylinder heat transfer on autoignition

The effect of heat transfer on the autoignition timing of propane was examined by adjusting the value of the convective heat transfer coefficient multiplier (ζ) in Eq. (1). The value of ζ was varied from 0.5 to 1.5, with all other model parameters remaining unchanged. Figure 3 indicates that heat transfer strongly influences the autoignition timing in a physically intuitive manner. In the case of reducing ζ, the lower in-cylinder heat transfer between the mixture and the combustion chamber

Conclusions

This paper investigated the autoignition of C3/C4 hydrocarbon mixtures in a CFR octane rating engine. The four species examined – propane, propylene (propene), n-butane and iso-butane – are the primary constituents of Liquefied Petroleum Gas (LPG), and are also important intermediates in the oxidation of larger hydrocarbons. In-cylinder pressure data was acquired for both autoigniting and non-autoigniting engine operation at the same test conditions. The latter was obtained by adding a small

Acknowledgments

This research was supported by the Advanced Centre for Automotive Research and Testing (ACART) and the Australian Research Council.

References (33)

  • R.M. Shanmugam et al.

    Performance and emission characterization of 1.2L MPI engine with multiple fuels (E10, LPG and CNG)

    SAE Int. J. Fuels Lubr.

    (2010)
  • M. Pecqueur, K. Ceustermans, P. Huyskens, D. Savvidis, Emissions Generated from a Suzuki Liane Running on Unleaded...
  • R. Chiriac, B. Radu, N. Apostolescu, Defining Knock Characteristics and Autoignition Conditions of LPG with a Possible...
  • B. Radu, G. Martin, R. Chiriac, N. Apostolescu, On the Knock Characteristics of LPG in a Spark Ignition Engine, SAE...
  • P. Dagaut et al.

    The high-pressure reduction of nitric oxide by a natural gas blend

    Combust. Flame

    (2005)
  • K.J. Morganti et al.

    Design and analysis of a modified CFR engine for the octane rating of liquefied petroleum gases (LPG)

    SAE Int. J. Fuels Lubr.

    (2014)
  • Cited by (38)

    View all citing articles on Scopus
    View full text