Abstract
Compact reaction schemes capable of predicting auto-ignition are a prerequisite for the development of strategies to control and optimise homogeneous charge compression ignition (HCCI) engines. In particular for full boiling range fuels exhibiting two stage ignition a tremendous demand exists in the engine development community. The present paper therefore meticulously assesses a previous 7-step reaction scheme developed to predict auto-ignition for four hydrocarbon blends and proposes an important extension of the model constant optimisation procedure, allowing for the model to capture not only ignition delays, but also the evolutions of representative intermediates and heat release rates for a variety of full boiling range fuels. Additionally, an extensive validation of the later evolutions by means of various detailed n-heptane reaction mechanisms from literature has been presented; both for perfectly homogeneous, as well as non-premixed/stratified HCCI conditions. Finally, the models potential to simulate the auto-ignition of various full boiling range fuels is demonstrated by means of experimental shock tube data for six strongly differing fuels, containing e.g. up to 46.7% cyclo-alkanes, 20% napthalenes or complex branched aromatics such as methyl- or ethyl-napthalene. The good predictive capability observed for each of the validation cases as well as the successful parameterisation for each of the six fuels, indicate that the model could, in principle, be applied to any hydrocarbon fuel, providing suitable adjustments to the model parameters are carried out. Combined with the optimisation strategy presented, the model therefore constitutes a major step towards the inclusion of real fuel kinetics into full scale HCCI engine simulations.
Nomenclature
HCCI | = | Homogeneous Charge Compression Ignition |
NOx | = | Nitrogen Oxide |
EGR | = | Exhaust Gas Recirculation |
DME | = | Dimethyl Ether |
MB | = | Methylbutanoate |
ILDM | = | Intrinsic Low Dimension Manifold |
CSP | = | Computational Singular Perturbation |
CFD | = | Computational Fluid Dynamics |
LES | = | Large Eddy Simulation |
CA | = | Crank Angle |
NTC | = | Negative Temperature Coefficient |
PRF | = | Primary Reference Fuel |
GA | = | Genetic Algorithm |
LT | = | Low Temperature |
HT | = | High Temperature |
CF | = | Cool Flame |
RMSE | = | Root-Mean-Square Deviation |
ms | = | Millisecond |