Page 1 of 23
2010-01-0572
Simulating combustion of practical fuels and blends for modern
engine applications using detailed chemical kinetics
Andrew Smallbone, Amit Bhave
Reaction Engineering Solutions Ltd.
Neal Morgan, Markus Kraft
University of Cambridge
Roger Cracknell and Gautam Kalghatgi
Shell Global Solutions (UK)
Copyright © 2010 SAE International
ABSTRACT
This research describes the potential to adopt detailed chemical kinetics for practical and potential future fuels
using tri-component surrogate mixtures capable of simulating fuel octane "sensitivity". Since the combustion
characteristics of modern fuels are routinely measured using the RON and MON of the fuel, a methodology to
generate detailed chemical kinetic mechanisms for these fuels based on these data is presented. Firstly, a novel
correlation between various tri-component blends (comprised of i-octane, n-heptane and toluene) and fuel RON
and MON was obtained by carrying out standard octane tests. Secondly, a chemical kinetic mechanism for tri-
component fuels was validated using a Stochastic Reactor Model (SRM) suite, an in-cylinder engine
combustion simulator, and a series of engine experiments conducted in HCCI operating mode. Thirdly, the
methodology was applied to predict combustion characteristics of a practical gasoline and fuel blends with
ethanol and di-iso-butylene blends using detailed chemical kinetics. Finally, for the first time the application of
this technique was demonstrated by employing detailed chemistry in the optimization of two engines and two
fuels operating in HCCI mode. Here a parametric study highlighted the adoption of fuels with "sensitivity"
could significantly extend the HCCI peak operating IMEP limit by as much as 60%.
INTRODUCTION
Recent advances in chemical kinetics have brought about ever more robust fuel models capable of computing
the combustion characteristics of the higher molecular weight hydrocarbon fuels [1, 2, 3]. However, due to the
vast number of hydrocarbons blended into practical gasolines [4], surrogates representative of the fuel (usually
based on a simplified alkane with equivalent carbon number) are typically adopted in order to simplify the
chemistry [3]. Conventionally, these have been mono or bi-component surrogates, that is either i-octane or a
Primary Reference Fuel (PRF), where the PRF is adopted in i-octane/n-heptane proportions equivalent to the
RON of the practical fuel, or by subtle tuning of the blend to match with corresponding experimental data [5].
However by definition, a PRF blend has zero fuel “sensitivity”, S=RON-MON, hence meaning it is an
insufficient surrogate for fuels with “sensitivity”. This limits the practical robustness of their adoption in a
predictive sense for the full range of fuels and operating points observed in modern engines [6, 7, 8].
This is the Computational Modelling Group's latest version of the paper.
For the published version please refer to http://www.sae.org/technical/papers/2010-01-0572