A quasi-dimensional burn rate model for pre-chamber-initiated SACI combustion

Abstract

Tackling climate change demands an intersectoral restructuring towards a sustainable economy and way of living. Since the transport sector accounts for considerable amounts of manmade greenhouse gas emissions, incisive technical solutions are required for its fast and sustainable defossilization. Evaluating different drive types seems rational, considering the expanse of application areas with highly diverging requirements. This includes developing highly hybridized and advanced combustion engines operated with sustainable fuels and striving for optimum energy conversion efficiencies. For the latter, fast-running quasi-dimensional simulations allow conducting robust design studies and complement further testing methodologies. Therefore, phenomenological models able to assess the behavior of different combustion mechanisms based on the boundary conditions are required. This work addresses the development of such a model for a promising combustion method: Pre-Chamber (PC) initiated flame propagation coupled with Spark-Assisted Compression Ignition (SACI) combustion. Despite significant development efforts to optimize the spark ignition engine combustion (quasi-hemispherical flame propagation combustion) while avoiding knock phenomena, achieved indicated efficiencies remain around 35–40 %. Further optimizations are enabled by significant dilution or increased combustion speed. However, flammability limits and decreasing flame speeds with increasing dilution prevent substantial improvements. PC-initiated jet ignition combustion systems improve flame stability and shift flammability limits towards higher dilution levels due to increased turbulence and a larger flame area in the early Main-Chamber (MC) combustion stages. Simultaneously, the much-increased combustion speed reduces knock tendency, allowing the implementation of the innovative PC-initiated SACI combustion. The jets penetrating the MC establish a flame propagation combustion that triggers a controlled volume reaction in the remaining charge. The resulting ultra-fast combustion process converges to the ideal thermodynamic constant-volume cycle leading to indicated efficiencies of > 45 %. Physically sound modeling is enabled by first developing and validating quasi-dimensional burn rate models for the mentioned individual combustion mechanisms: SACI and PC-initiated jet ignition. The former model is based on a previous work assessing homogeneous charge compression ignition combustion. The SACI model includes the initial flame propagation combustion by employing a two-zone entrainment approach, while the volume reaction is covered by a multi-pseudo-zone approach based on a temperature-distributed auto-ignition integral. The auto-ignition integral calculation approach is based on wide-ranging reaction kinetic calculations of different fuels, including gasoline blends, ethanol, and methanol. Moreover, a calculation specification is implemented to consider the volume reaction conversion time relevant for highly diluted mixtures. Based on literature research, a burn rate model for PC-initiated flame propagation is implemented, considering two thermodynamic systems (PC and MC) connected through orifices. Both systems use a two-zone entrainment model for flame propagation combustion. The effects of the PC combustion on the MC are displayed by considering the jet-shaped flame area, the jet-induced turbulence, and a separate jet entrainment term. Moreover, the jet turbulence’s impact on the MC WHT is implemented. Finally, the burn rate models for PC-initiated flame propagation and SACI combustion are merged, including physical interaction terms between the PC effects and the volume reaction. The combined model gets along with a manageable number of understandable calibration parameters. The models are integrated into the so-called cylinder module developed at the Institute of Automotive Engineering Stuttgart. Furthermore, the individual and combined burn rate models are validated with experimental data from different geometrical engine setups. The experimental campaigns cover different loads and speeds, fuels (ethanol, standard gasoline with up to 10 vol. % ethanol), excess air dilutions (λ = 1-2.8), and compression ratios (12.6-16.4). Overall, the burn rate models show noteworthy predictive capabilities employing a fixed set of calibration parameters for a given geometric engine setup. Although significant simplifications are (necessarily) used, satisfactory prediction of the burn rates and pressure curves can be achieved for the different combustion mechanisms.