Flame Front Propagation in an Optical GDI Engine under Stoichiometric and Lean Burn Conditions

Santiago Martinez, Adrian Irimescu, Simona Merola, Pedro Lacava, Pedro Curto-Riso
2017 Energies  
Lean fueling of spark ignited (SI) engines is a valid method for increasing efficiency and reducing nitric oxide (NO x ) emissions. Gasoline direct injection (GDI) allows better fuel economy with respect to the port-fuel injection configuration, through greater flexibility to load changes, reduced tendency to abnormal combustion, and reduction of pumping and heat losses. During homogenous charge operation with lean mixtures, flame development is prolonged and incomplete combustion can even
more » ... stion can even occur, causing a decrease in stability and engine efficiency. On the other hand, charge stratification results in fuel impingement on the combustion chamber walls and high particle emissions. Therefore, lean operation requires a fundamentally new understanding of in-cylinder processes for developing the next generation of direct-injection (DI) SI engines. In this paper, combustion was investigated in an optically accessible DISI single cylinder research engine fueled with gasoline. Stoichiometric and lean operations were studied in detail through a combined thermodynamic and optical approach. The engine was operated at a fixed rotational speed (1000 rpm), with a wide open throttle, and at the start of the injection during the intake stroke. The excess air ratio was raised from 1 to values close to the flammability limit, and spark timing was adopted according to the maximum brake torque setting for each case. Cycle resolved digital imaging and spectroscopy were applied; the optical data were correlated to in-cylinder pressure traces and exhaust gas emission measurements. Flame front propagation speed, flame morphology parameters, and centroid motion were evaluated through image processing. Chemical kinetics were characterized based on spectroscopy data. Lean burn operation demonstrated increased flame distortion and center movement from the location of the spark plug compared to the stoichiometric case; engine stability decreased as the lean flammability limit was approached. With respect to gasoline engines, the development of direct injection (DI) technologies supported by downsizing, variable valves timing, and lift has allowed the industry to improve fuel economy and engine performance [6] [7] [8] [9] . In order to further enable a reduction in fuel consumption and CO 2 emissions of gasoline direct-injection (GDI) engines, lean burn strategies are often tested [10] . Low fuel consumption with high power output can be simultaneously attained by means of the GDI engine technology, since a lean overall fuel-air mixture is possible with precisely controlled fuel injection directly into the cylinder, which leads to highly efficient combustion. The significant reduction in fuel consumption is achieved as a result of the reduced pumping loss during the intake stroke; furthermore, when the engine is operating under low loads, cooling losses are diminished because of lower combustion temperature. Meanwhile, when the engine is operating under high loads, the volumetric efficiency improves by the latent heat of the fuel and the possibility of knocking is reduced [11] . However, direct fuel injection into the cylinder of a GDI engine leads to the formation of particulate matter (PM), similar to in a diesel engine. The size distributions are strongly affected by the degree of fuel-air mixing because particles are mainly formed in the fuel-rich mixture region as a result of the poor spray evaporation and wall wetting [12, 13] . Several solutions for reducing PM emissions, mainly based on alternative fuel injection and ignition strategies, as well as the control of charge motion, have been proposed taking into account the dependence from the engine operating parameters [14] [15] [16] [17] [18] . However, most of the studies were focused on stoichiometric mixture combustion and not on lean burn GDI operation. The effect of air dilution on in-cylinder particle formation is controversial; it was demonstrated that the PM concentration at a relatively high load decreases only for optimized ignition conditions. The degree of lean operation in GDI engines is principally limited by increasingly unstable combustion, particularly for homogeneous approaches. Under highly dilute conditions, flame speeds decrease, resulting in poor fuel conversion efficiency due to prolonged combustion. The upper stability limit is generally recognized as 5% coefficient of variation (COV) of the indicated mean effective pressure (IMEP) [19, 20] . Moreover, cycle-to-cycle variation plays a key role in combustion since it can skew the output power even up to 10% in GDI engines [21] . Innovative ignition systems [22, 23] and injection strategies [24] have been developed in order to extend this lean limit and exploit the benefits of homogeneous lean (λ > 1.4) combustion in terms of fuel consumption and nitrogen oxide (NO x ) emissions. Deeper insight into the effects of excess air ratio on in-cylinder processes can be a useful support in the development of advanced solutions for the optimization of GDI engines. To this aim, the combustion process in a wall guided DISI engine operating in lean burn homogenous charge conditions was investigated. The experiments were performed at fixed engine speed and wide open throttle (WOT) in an optically accessible single-cylinder engine fueled with commercial gasoline. The air-fuel ratio (λ) varied from stoichiometric to values close to the lean flammability limit. High spatial resolution cycle resolved digital imaging was used to characterize the flame front propagation by macroscopic flame parameters; natural emission spectroscopy in the UV-visible range was applied to follow the evolution of chemical species (such as OH) and soot formation. Optical results were correlated with thermodynamic data and exhaust emission measurements. The main goal of the work is to contribute to the understanding of fundamental in-cylinder processes and support computational fluid dynamics (CFD) modeling.
doi:10.3390/en10091337 fatcat:wperinykgrhidgxjj7ij4nvozu