Search

The pre-chamber combustion (PCC) concept is a proven lean or diluted combustion technique for internal combustion engines with benefits in engine efficiency and reduced NOx emissions. The engine lean operation limit can be extended by supplying auxiliary fuel into the pre-chamber and thereby, achieving mixture stratification inside the pre-chamber over the main chamber. Introducing liquid fuels into the pre-chambers is challenging owing to the small form factor of the pre-chamber. With a conventional injector, the fuel penetrates in liquid form and impinges on the pre-chamber walls, which leads to increased unburned hydrocarbon emissions from the pre-chamber. In this study, a prototype liquid fuel injector is introduced which preheats the fuel within a heated chamber fitted with an electrical heating element before injecting an effervescently atomized spray into the pre-chamber. The experiments were conducted in a heavy-duty pre-chamber research engine using ethanol as the primary fuel. In the first set of experiments, only the pre-chamber was fueled to investigate the influence of injector operating parameters on the mass of fuel injected and the stability of the fuel flow rate. In the second set of experiments, fuel was supplied to both the pre- and the main chamber to investigate the engine performance at different air-fuel ratios at a fixed intake airflow. As a proof of concept, preheating the fuel prior to injection into the pre-chamber was found to improve the combustion stability with simultaneous reductions in engine-out unburned fuel and carbon monoxide emissions while requiring only low power requirements for effective fuel preheating.

Pre-chamber combustion (PCC) modeling has been progressing in recent years, while there are lingering questions on fundamental modeling aspects, whether a flame-based or an ignition-based model predicts the combustion with higher fidelity. This mode of ignition concept is known to enable a stable engine operation at ultralean conditions with a short combustion duration, thus enhancing engine efficiency. The current work utilizes computational fluid dynamics to assess well-known combustion models: multi-zone well-stirred reactor (MZ-WSR) and G-Equation. The former models combustion as an ignition-based phenomenon while the latter as a flame propagation type of combustion. A pre-chamber containing twelve nozzles divided into two layers on a narrow throat was chosen. The jets from the two layers of nozzles and the local thermodynamic conditions differ substantially, which makes it a suitable configuration for assessing the predictive capabilities of distinct combustion models. The fuel utilized was methane and the global air-fuel ratio (λ) was varied, ranging from global-λ of 1.6, 1.8, and 2.0, and the total fuel injected through the pre-chamber was varied for one of the cases (3%, 7%, and 13%). The results suggest that both combustion models can potentially match experimental engine performance data upon appropriate calibration; however, fundamental differences in jet topology arise since the G-Equation formulation accounts for turbulence-chemistry interaction, while MZ-WSR does not.

Pre-chamber combustion (PCC) has re-emerged in recent last years as a potential solution to help to decarbonize the transport sector with its improved engine efficiency as well as providing lower emissions. Research into the combustion process inside the pre-chamber is still a challenge due to the high pressure and temperatures, the geometrical restrictions, and the short combustion durations. Some fundamental studies in constant volume combustion chambers (CVCC) at low and medium working pressures have shown the complexity of the process and the influence of high pressures on the turbulence levels. In this study, the pre-chamber combustion process was investigated by combustion visualization in an optically-accessible pre-chamber under engine relevant conditions and linked with the jet emergence inside the main chamber. The pre-chamber geometry has a narrow-throat. The total nozzle area is distributed in two six-hole rows of nozzle holes. A novel optical pre-chamber assembly was designed and tested on an optical engine, with combustion only in the pre-chamber volume, using methane (99.5% purity CH4) as fuel. A separate measurement was carried out using a metal pre-chamber assembly to study the emergence of the pre-chamber jets into the main chamber to assess the difference between the pre-chambers. The combustion process was captured using a broadband camera and pressure transducers in both chambers (pre- and main chamber). The combustion inside the pre-chamber starts at the spark location; the flame propagates along the cone and subsequently propagates fast inside the throat. The broadband combustion luminosity exhibited a non-proportional relation with the heat release rate (HRR) due to the CO2* chemiluminescence. An HSV-model (Hue Saturation Value) analysis revealed the predominance of premixed combustion during the first stage of the process, followed by infrared flame radiation. Finally, although the optical pre-chamber assembly showed a faster pressure increase, the combustion process is similar to the metal pre-chamber assembly on a mass burnt scale.

An optimized active pre-chamber geometry was obtained by combining computational fluid dynamics (CFD) and machine learning (ML). A heavy-duty engine operating with methane under lean conditions was considered. The combustion process was modeled with a multi-zone well-stirred reactor (MZ-WSR) with a skeletal methane oxidation mechanism. The simulations were run for a complete cycle. For the optimization study, the pre-chamber was parametrized; six independent and three dependent variables were considered, while the volume was kept constant. Three hundred pre-chamber designs were generated, and a one-shot design of experiments (DoE) optimization was first considered. A merit function was adopted to rank the designs, and an optimum design was found from the DoE results, which yielded considerable improvements in merit ranking, considering fuel consumption, engine-out emissions, noise, and safety; secondly, machine learning algorithms were trained by utilizing the DoE results aiming at finding a globally optimum geometry for the considered operating condition. Five sequential iterations were performed, and the ML algorithms were capable of proposing a new design with superior performance compared to the best DoE.