Search Results

A computational fluid dynamics (CFD) guided design optimization was conducted for the piston bowl geometry for a heavy-duty diesel engine. The optimization goal was to minimize engine-out NOx emissions without sacrificing engine peak power and thermal efficiency. The CFD model was validated with experiments and the combustion system optimization was conducted under three selected operating conditions representing low speed, maximum torque, and rated power. A hundred piston bowl shapes were generated, of which 32 shapes with 3 spray angles for each shape were numerically analyzed and one optimized design of piston bowl geometry with spray angle was selected. On average, the optimized combustion system decreased nitrogen oxide (NOx) emissions by 17% and soot emissions by 41% without compromising maximum engine power and fuel economy.

Homogeneous Charge Induced Ignition (HCII) combustion utilizes a port injection of high-volatile fuel to form a homogeneous charge and a direct injection of high ignitable fuel near the Top Dead Center (TDC) to trigger combustion. Compared to Conventional Diesel Combustion (CDC) with high injection pressures, HCII has the potential to achieve diesel-like thermal efficiency with significant reductions in NOx and PM emissions with relatively low-pressure injections, which would benefit the engine cost saving remarkably. In the first part of current investigation, experiments were conducted at medium load with single diesel injection strategy. HCII exhibited great potential of using low injection pressures to achieve low soot emissions. But the engine load for HCII was limited by high heat release rate. Thus, in the second and third part, experiments were performed at high and low load with double diesel injection strategy.

Plug-in electric vehicle (PHEV) has a promising prospect to reduce greenhouse gas (GHG) emission and optimize engine operating in high-efficiency region. According to the maximum electric power and all-electric range, PHEVs are divided into two categories, including “all-electric PHEV” and “blended PHEV” and the latter provides a potential for more rational energy distribution because engine participates in vehicle driving during aggressive acceleration not just by motor. However, the frequent use of engine may result in severe emissions especially in low state of charge (SOC) and ahead of catalyst light-off. This study quantitatively investigates the impact of oil viscosity and driving mode (hybrid/conventional) on oil dilution and emissions including particle number (PN).

Multi-mode combustion is an ideal combustion strategy to utilize HCCI for internal combustion engines. It combines HCCI combustion mode for low-middle load and traditional SI mode for high load and high speed. By changing the cam profiles from normal overlap for SI mode to the negative valve overlap (NVO) for HCCI mode, as well as the adjustment of direct injection strategy, the combustion mode transition between SI and HCCI was realized in one engine cycle. By two-step cam switch, the throttle action is separated from the cam action, which ensures the stabilization of mode transition. For validating the feasibility of the stepped switch, the influence of throttle position on HCCI combustion was carefully studied. Based on the research, the combustion mode switch was realized in one engine cycle; the whole switch process including throttle action was realized in 10 cycles. The entire process was smooth, rapid and reliable without any abnormal combustion such as knocking and misfiring.

In this paper, a hybrid combustion mode in four-stroke gasoline direct injection engines was studied. Switching cam profiles and injection strategies simultaneously was adopted to obtain a rapid and smooth switch between SI mode and HCCI mode. Based on the continuous pressure traces and corresponding emissions, HCCI steady operation, HCCI transient process (combustion phase adjustment, SI-HCCI, HCCI-SI, HCCI cold start) were studied. In HCCI mode, HCCI combustion phase can be adjusted rapidly by changing the split injection ratio. The HCCI control strategies had been demonstrated in a Chery GDI2.0 engine. The HCCI engine simulation results show that, oxygen and active radicals are stored due to negative valve overlap and split fuel injection under learn burn condition. This reduces the HCCI sensitivity on inlet boundary conditions, such as intake charge and intake temperature. The engine can be run from 1500rpm to 4000rpm in HCCI mode without spark ignition.

This study improved the computational efficiency significantly using parallel computation and reduced mechanisms. A 3-dimensional engine moving mesh of intake port, exhaust port and combustion chamber was established for HCCI engine cycle simulation. To achieve a more accurate analysis, chemical kinetics was implemented into the CFD code to study the intake, spray, ignition, combustion, and pollution formation process in HCCI engine. The simulations were run on a cluster of 16-CPU, parallelized by Message-Passing Interface (MPI) mode. The cases with detailed and reduced reaction mechanisms were calculated using 1, 2, 4, 8, 16 CPUs respectively and the corresponding computational time and speed-up were discussed. Using MPI 8-CPU with reduced mechanism (less than 40 species) is the optimal scheme for CFD/Chemistry calculation of typical HCCI engine.

A conceptual approach to help understand and simulate droplet induced pre-ignition is presented. The complex phenomenon of oil-fuel droplet induced pre-ignition has been decomposed to its elementary processes. This approach helps identify the key fluid properties and engine parameters that affect the pre-ignition phenomenon, and could be used to control LSPI. Based on the conceptual model, a 3D CFD engine simulation has been developed which is able to realistically model all of the elementary processes involved in droplet induced pre-ignition. The simulation was successfully able to predict droplet induced pre-ignition at conditions where the phenomenon has been experimentally observed. The simulation has been able to help explain the observation of pre-ignition advancement relative to injection timing as experimentally observed in a previous study [6].

With increasingly stringent requirements and regulations related to particulate matter(PM) emissions, manufacturers are paying more and more attention to emissions from gasoline direct injection(GDI) engines. The present paper proposes an improved two-step soot model. The model is applied in the Kiva-Chemkin program to simulate the processes of spray impinging, fuel mixture preparation, combustion and soot formation in a typical turbocharged downsized GDI engine. The simulation results show that soot formation in the GDI engine is attributed to non-uniform distribution of the air-fuel mixture and pool fire of wall film in the cylinder. Under homogeneous mode, increasing the injection advance angle can optimize fuel atomization and improve air-fuel mixing, thus reducing soot formation. However, an excessive injection advance angle may cause spray to impinge on the cylinder wall and this will sharply increase the soot emission.

A new combustion mode namely multiple premixed compression ignition (MPCI) for gasoline engines was proposed. The MPCI mode can be realized by two or more times gasoline injections into cylinder with a high pressure around the compression TDC and featured with a premixed combustion after each injection in the cylinder, which is different from the existed gasoline direct injection compression ignition (GDICI) modes such as homogeneous charge compression ignition (HCCI) mode with gasoline injection occurred in intake stroke, and partially premixed compression ignition (PPCI) mode with multiple gasoline injections in intake and compression strokes before the start of combustion (SOC). Therefore the spray and combustion of the MPCI mode are alternatively occurred as "spray-combustion-spray-combustion" near the TDC, rather than "spray-spray-combustion" sequence as traditional PPCI gasoline engines.

The detailed chemical kinetics was implemented into the 0D single-zone model, 1D engine cycle model and 3D CFD model respectively. Both simulation and experiment were carried out on a four-stroke gasoline HCCI engine with direct injection. The 0D model (HRG) was developed to study the isolating effect of HCCI parameters. The 1D engine cycle model with HRG was established to study the gas exchange process and the direct injection strategy in the gasoline HCCI engine with two-stage direct injection (TSDI) and negative valve overlap (NVO). The 3D model with HRG was constructed to study the ignition control mechanism in “two-zone HCCI” combustion via analysis the auto-ignition spots distribution in combustion chamber. “Two-zone HCCI” is defined in this paper as premixed ultra-lean compression ignition combining with premixed ultra-rich compression ignition. The three models were validated by experiments.

In this research, propane flash boiling sprays discharged from a five-hole gasoline direct injector were studied in a constant volume vessel. The fuel temperature (Tfuel) ranged from 30 °C to 90 °C, and the ambient pressure (Pamb) varied from 0.05 bar to 11.0 bar. Different flash boiling spray behavior compared to that under sub-atmospheric conditions was found at high Pamb. Specifically, at the sub-atmospheric pressures, the individual flashing jets merged into one single jet due to the strong spray collapse. In contrast, at Pamb above 3.0 bar and Tfuel above 50 °C, the spray collapse was mitigated and the flashing jets were separated from each other. Further analyses revealed that the mitigation of spray collapse at high Pamb was ascribed to the suppression of jet expansion. In addition, it was found that the spray structure was much different at similar Rp, indicating that Rp lacked the generality in describing the structure of flash boiling sprays.

HCCI combustion was studied in a 4-stroke gasoline engine with a direct injection system. The electronically controlled two-stage gasoline injection and spark ignition system were adopted to control the mixture formation, ignition timing and combustion rate in HCCI engine. The engine could be operated in HCCI combustion mode in a range of load from 1 to 5 bar IMEP and operated in SI combustion mode up to load of 8 bar IMEP. The HCCI combustion characteristics were investigated under different A/F ratios, engine speeds, starts of injection, as well as spark ignition enabled or not. The test results reveal the HCCI combustion features as a high-pressure gradient after ignition and has advantages in high thermal efficiency and low NOx emissions over SI combustion. At the part load of 1400rpm and IMEP of 3.5bar, ISFC in HCCI mode is 25% lower and NOx emissions is 95% lower than that in SI mode.

Compressed Natural Gas (CNG), because of its low cost, high H/C ratio, and high octane number, has great potential in automotive industry, especially for heavy-duty commercial vehicles. However, relative slow flame speed of natural gas leads to long combustion duration and low thermal efficiency and tends to cause knock combustion at high load, which will aggravate engine thermal load and reliability. Enhancing turbulence intensity in combustion chamber is an effective way to accelerate flame propagation speed and improve combustion performance. In this study, the flow simulations of several piston bowls with different inner-convex forms were carried out using three-dimensional computational fluid dynamics (3D-CFD) software CONVERGE. The numerical results showed the piston bowls with inner-convex could disturb the charge swirl motion and enhance turbulence of different intensity. A hexagram geometry bowl was proved to have the best function in strengthening turbulence intensity.

The ignition of lubricating oil droplet has been proved to be the main factor for pre-ignition and the following super-knock in turbocharged gasoline direct injection engine. In this paper, the ignition process of lubricating oil droplet in combustible ambient gaseous mixture was investigated in a rapid compression machine (RCM). The pre-ignition induction by oil droplet of the ambient gaseous mixture was analyzed under different initial droplet volume and effective temperature conditions. The oil droplet was suspended on a tungsten fiber in the combustion chamber and the ignition process was recorded by a high-speed camera through the quartz window mounted at the end of the combustion chamber. The pressure traces were also obtained by a sensor in order to get the ignition delay and analyze the combustion process in detail.

A 1-Dimensional (1-D) model of fluid dynamic and chemistry kinetics following hot spot auto-ignition has been developed to simulate the process from auto-ignition to pressure wave propagation. The role of wall effect on the physical-chemical interaction process is numerically studied. A pressure wave is generated after hot spot auto-ignition and gradually damped as it propagates. The reflection of the wall forms a reflected pressure wave with twice the amplitude of the incident wave near the wall. The superposition of the reflected and forward pressure waves reinforces the intensity of the initial pressure wave. Wall effect is determined by the distance between the hot spot center and the cylinder wall. Hot spot auto-ignition near the wall easily initiates detonation under high-temperature and high-pressure conditions because pressure wave reflection couples with chemical reactions and propagates in the mixture with high reactivity.

Occurrence of sporadic super-knock is the main obstacle to the development of advanced gasoline engines. One of the possible inducements of super-knock, agglomerated soot particle induced pre-ignition, was studied for high boosted gasoline direct injection (GDI) engines. The correlation between soot emissions and super-knock frequency was investigated in a four-cylinder gasoline direct injection production engine. The test results indicate that higher in-cylinder soot emission correlate with more pre-ignition and super-knock cycles in a GDI production engine. To study the soot/carbon particles trigger super-knock, a single-cylinder research engine for super-knock study was developed. The carbon particles with different temperatures and sizes were introduced into the combustion chamber to trigger pre-ignition and super-knock.

The effects of the temporal and spatial distributions of ignition timings of combustion zones on combustion noise in a Direct Injection Compression Ignition (DICI) engine were studied using experimental tests and numerical simulations. The experiments were performed with different fuel injection strategies on a heavy-duty diesel engine. Cylinder pressure was measured with the sampling intervals of 0.1°CA in order to resolve noise components. The simulations were performed using the KIVA-3V code with detailed chemistry to analyze the in-cylinder ignition and combustion processes. The experimental results show that optimal sequential ignition and spatial distribution of combustion zones can be realized by adopting a two-stage injection strategy in which the proportion of the pilot injection fuel and the timings of the injections can be used to control the combustion process, thus resulting in simultaneously higher thermal efficiency and lower noise emissions.

High boost and direct injection hold the potential of enhanced power density and fuel consumption in the development of gasoline engines. However, super-knock with strong destructiveness was widely reported at low-speed and high-load operating regime in turbocharged GDI engines. The objective of this study is to clarify the characteristics of super-knock and to try to find some feasible solutions to suppress super-knock. To fast evaluate super-knock at low-speed and high-load regime, a rapid test procedure including three super-knock test sections of 5000 cycles with 3 idle operations, was proposed. The experimental data indicate that pre-ignition is not the sufficient condition for super-knock. Pre-ignition may lead to super-knock, heavy knock, slight knock, and non-knock. Compared with conventional knock, knock intensity of super-knock is much higher and the maximum amplitude of pressure rise at start of knock is more than one order of magnitude higher.

This paper studied the knock combustion process in gasoline HCCI engines. The complex chemical kinetics was implemented into the three-dimensional CFD code with LES (Large eddy simulation) to study the origin of the knock phenomena in HCCI combustion process. The model was validated using the experimental data from the cylinder pressure measurement. 3D-CFD with LES method gives detailed turbulence, species, temperature and pressure distribution during the gasoline HCCI combustion process. The simulation results indicate that HCCI engine knock originates from the random multipoint auto-ignition in the combustion chamber due to the slight inhomogeneity. It is induced by the significantly different heat release rate of high temperature oxidation (HTO) and low temperature oxidation (LTO) and their interactions.

Low Speed Pre-Ignition (LSPI), also referred to as superknock or mega-knock is an undesirable turbocharged engine combustion phenomenon limiting fuel economy, drivability, emissions and durability performance. Numerous researchers have previously reported that the frequency of Superknock is sensitive to engine oil and fuel composition as well as engine conditions in controlled laboratory and engine-based studies. Recent studies by Toyota and Tsinghua University have demonstrated that controlled induction of particles into the combustion chamber can induce pre-ignition and superknock. Afton and Tsinghua recently developed a multi-physics approach which was able to realistically model all of the elementary processes known to be involved in deposit induced pre-ignition. The approach was able to successfully simulate deposit induced pre-ignition at conditions where the phenomenon has been experimentally observed.