Detailed analysis of combustion stability in a spark-assisted compression ignition engine under nearly stoichiometric and heavy EGR conditions

2020 
Abstract Extending the load range of low-temperature combustion is of priority to meet future CO2 and emission targets for reciprocating engine applications. Spark assist is a feasible solution to this challenge for mono-fuel homogeneous charge compression ignition (HCCI). This paper explains how spark-assisted compression ignition (SACI) enables ultra-low NOX targets to be met, with acceptable pressure rise rates and combustion stability, at high load boundary conditions, favourable for HCCI/SACI transition. The work provides new methods of combustion analysis which give better understanding of the mechanisms and their implementation for real-time control of SACI engines. The goals are achieved by a combination of single-cylinder engine research and high-fidelity/high-speed, model-based calculations, performed on an individual cycle basis. The results show that determining the start of the kinetic phase in SACI is possible via standard combustion indicators. The new method is two orders of magnitude faster than the commonly used spline-Wiebe approach. With real-time capability and proven correlation to temperature evolution, triggered by propagating flame, the method enables in-cycle predictive control. Additionally, it gives a deeper insight of the mechanisms underpinning the demonstrated superior performance. The study shows the capability to run SACI at indicated mean effective pressure (IMEP) of 0.5 MPa with engine-out NOX below Euro VI’s heavy-duty engine limit and with specific fuel consumption of 207 g/kWh. Importantly, pressure rise rate and variation in IMEP do not exceed 0.25 MPa/CAD and 3% respectively. Margins for critical parameters are far greater than for baseline autonomous HCCI, providing significant load extension potential.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    35
    References
    7
    Citations
    NaN
    KQI
    []