LES of spray transients: start and end of injection phenomena

Journal of Visualization, 2019

Fuel injection timing for early injection mode at higher loads in a gasoline direct-injection engine is critical for mixture formation, combustion process and emissions. In the present work, the effect of fuel injection timing on cycle-to-cycle spray variations and macroscopic spray characteristics is quantitatively analyzed at wide-open throttle condition with stoichiometric air-fuel ratio. A Mie scattering technique was employed to visualize the liquid phase of the fuel dispersion with early injection timings at 180°before top dead center (BTDC), 210°BTDC and 240°BTDC of compression with fuel pressure of 5 MPa. A quantitative analysis using proper orthogonal decomposition revealed that cycle-to-cycle spray variations were not significant. Comparing different injection timings, results, however, showed that spray variations were slightly higher at injection timing of 240°BTDC compared to injection timings at 210°BTDC and 180°B TDC. Concerning macroscopic characteristics, it was found that the spray tip penetration length was longer at 210°BTDC compared to injection timings at 240°BTDC and 180°BTDC. It was also observed that the spray areas were comparable for different injection timings until about 0.9 ms after the start of injection, but it was enhanced at later time stamps for injection at 240°BTDC compared to injections at 210°BTDC and 180°BTDC.

2009

Atomization and Sprays, 2009

The spray characteristics of double fuel injection in multihole injectors for direct-injection gasoline engines have been evaluated in a constant-volume chamber using iso-octane as fuel. Measurements of droplets' mean and rms velocity and their diameters were obtained using a 2D phase Doppler anemometer (PDA) at injection pressures up to 120 bar, atmospheric chamber pressures, and ambient temperatures up to 115° C Complementary spray visualization made use of a pulsed light, a CCD camera, and a high-speed video camera synchronized with the injection process. Spray images with double injection revealed that there are delay times in needle opening and closing on the order of 0.6 ms and 0.3 ms, respectively, and that dwell times less than 0.5 ms introduced a prespray prior to the second injection event, while for dwell times more than 0.5 ms, the images showed that the overall jet-spray structure in the cases of the first and second injection event remained the same as those of single injection. PDA results near the injector exit quantified the relatively strong prespray at 0.3 ms dwell time and revealed that its strength reduced with increasing dwell time so that at 1 ms there was no sign of the prespray. The effect of ambient chamber temperature on the spray velocity for double injection was found to be significant at temperatures above 90° C; results at higher temperatures, 115° C, showed a consistent reduction in the mean droplets' velocity up to 10% within the core of the first and second sprays, with a corresponding increase in the rms velocity fluctuations due to the increased vaporization rates and the droplets' momentum loss. The droplet size distribution of the double-injection spray was, in general, similar to that of single injection except during the dwell time when the droplets arithmetic mean diameter with prespray was found to be smaller than those without the prespray because of the different type of droplets and the much higher data rates in the case of the prespray.

Gasoline sprays from high-pressure multi-hole nozzles for direct injection spark-ignition engines have been simulated using a CFD spray model. Model validation takes place against experimental data available for injection into a constant volume chamber. These include CCD spray images and droplet velocity and diameter measurements obtained with a two-component phase Doppler anemometry (PDA) system at injection pressures up to 200bar and chamber pressures varying from atmospheric to 12bar. An 1-D FIE (Fuel Injection Equipment) hydraulic flow simulation provides the injection rate, while a 3-D and two-phase CFD nozzle flow model predicts the injection velocity increase due to cavitation. Finally, a cavitation-induced atomisation model predicts the droplet size distribution very near the nozzle exit. The subsequent spray development is predicted using the Eulerian-Lagrangian methodology, including for liquid droplet aerodynamic break-up, turbulent dispersion, droplet vaporisation and droplet-to-droplet interaction. Overall, both model predictions and measurements confirmed the advantages of high-pressure multi-hole injectors for gasoline direct-injection engines relative to swirl pressure atomisers, in terms of spray structure independency from chamber thermodynamic and injection operating conditions.

SAE technical paper series, 2002

Experimental Thermal and Fluid Science, 2017

Gasoline direct injection (GDI) sprays are complex multiphase flows. When compared to multi-hole diesel sprays, the plumes are closely spaced, and the sprays are more likely to interact. The effects of multi-jet interaction on entrainment and spray targeting can be influenced by small variations in the mass fluxes from the holes, which in turn depend on transients in the needle movement and small-scale details of the internal geometry. In this paper, we present a comprehensive overview of a multi-institutional effort to experimentally characterize the internal geometry and near-nozzle flow of the Engine Combustion Network (ECN) Spray G gasoline injector. In order to develop a complete picture of the near-nozzle flow, a standardized setup was shared between facilities. A wide range of techniques were employed, including both x-ray and visible-light diagnostics. The novel aspects of this work include both new experimental measurements, and a comparison of the results across different techniques and facilities. The breadth and depth of the data reveal phenomena which were not apparent from analysis of the individual data sets. We show that plume-to-plume variations in the mass fluxes from the holes can cause large-scale asymmetries in the entrainment field and spray structure. Both internal flow transients and small-scale geometric features can have an effect on the external flow. The sharp turning angle of the flow into the holes also causes an inward vectoring of the plumes relative to the hole drill angle, which increases with time due to entrainment of gas into a low-pressure region between the plumes. These factors increase the likelihood of spray collapse with longer injection durations.

An experimental study of the mixture preparation characteristics of a direct injection gasoline (G-DI) engine is presented. The airflow and fuel spray features of a top-entry, lean-burn, direct injection strategy are described by the application of advanced LASER-based experimental tools and techniques. Two comprehensive in-cylinder airflow studies were undertaken to show the contrast between the temporal and spatial distributions of the mean and turbulent gas motions, in the mid-cylinder tumble (and cross-tumble) plane, of a conventional, side-entry, port injected (MPI) engine with flat-top piston and a top-entry G-DI engine with a bowled piston. Airflow measurements were performed using LASER Doppler Anemometry (LDA) in the axial (MPI) and axial and radial (G-DI) directions in a motored, single cylinder, research engine over consecutive engine cycles. Optical access to the combustion chambers was achieved by three different methods. In the MPI engine, a modified piston that incorporated a circular window was used in conjunction with an inclined mirror, fixed below the piston. Experimental measurements in the G-DI were performed along the axis of the spark plug and into the piston bowl and within the upper cylinder and the piston topland region using a spark plug window and glass annular section, sandwiched between the cylinder liner and cylinder head. In the MPI engine, the principal air motion can be described as a forward tumble pattern, established late in the intake stroke and which breaks down into turbulence close to the end of the compression stroke. In the G-DI engine, a strong, reverse tumble motion is rapidly formed early in the intake stroke. This motion persisted throughout the compression stroke and up to TDC without an increase in turbulence intensity. The magnitude of the tumble plane velocity was far greater than that measured in the cross-tumble plane. Observed levels of RMS turbulence intensity were significantly greater in the conventional cylinder head towards the end of the compression stroke indicating the conservation of the reverse tumble motion in the G-DI system. During the early injection phase the airflow exhibited the requirements for a homogeneous charge mixture. For late injection, the persistent coherent tumble structure produced flow stability for a stratified mixture approach. The longitudinal integral length scale of turbulence was determined at two locations near the spark plug gap at ignition timings and TDC conditions. The indirect method and Taylor's hypothesis was used to determine the length scale from the temporal autocorrelation function estimate. The criteria for Taylor's hypothesis for stationary turbulent flow were validated for the coherent, tumbling vortex present in the late compression stroke. Phase Doppler Anemometry (PDA) measurements were taken of a high pressure, swirl type gasoline fuel injector. Droplet size, velocity, RMS velocity and validated data rate measurements were used to quantify the spray atomisation quality and geometry under atmospheric, quiescent conditions and within a motored engine for early and late injection scenarios. High-speed LASER light sheet photography was used for visualisation. In-cylinder PDA measurements were performed at the same locations across the mid-cylinder plane, as employed for the LDA air motion study. The PDA measurements and high-speed photography showed that the fuel spray exhibited a swirling, hollow cone structure during early injection. A quasi-steady state inner cone angle for the spray was estimated. For early injection conditions, a pre-injection, swirless 'slug' of fuel was observed, predominantly along the spray axis, in the time-resolved experiments, that preceded the main injection and asymmetrical cone development and that contained relatively large diameter droplets with high axial velocities. For late injection, the mean droplet diameters were larger than those of the early injection and static tests and the fuel spray formed a narrow jet. The difference in the mean properties was attributed to the nature of the in-cylinder gas properties; An ensemble-averaged, volume-weighted mean velocity was used to describe the momentum exchanged between the gas and liquid phases. It was concluded from the comparative studies of early and late fuel injection, that the droplet size distribution was dominated by secondary atomisation through interaction with the gaseous flowfield when the fuel spray was injected into high gas velocities in the spray-wise direction. The fuel spray injection axis was deflected downwards towards the piston by the momentum of the incoming air charge. The uppermost periphery of the fuel spray was subjected to flapping instabilities and partial detachment. Observations for a suitable mixture preparation strategy are highlighted for a direct injection gasoline engine, based upon the conclusions drawn from a comprehensive experimental study. The effect of in-cylinder airflow on the fuel spray characteristics is quantified for both early and late injection operations. The stability of stratification under motored conditions is discussed in terms of the in-cylinder air motion, turbulence intensity, integral length scale of turbulence and its ability to control the distribution of the mixture and preparation of suitable conditions for initial flame kernel growth. It was concluded that models for mixture preparation based upon pre-conceptions derived from manifold injection, tumble combustion systems and high-pressure Diesel fuel spray analyses are not readily applicable to a strategy that is based upon the direct injection of gasoline into the engine combustion chamber under varying pressure and airflow conditions.

An experimental study of the mixture preparation characteristics of a direct injection gasoline (G-DI) engine is presented. The airflow and fuel spray features of a top-entry, lean-burn, direct injection strategy are described by the application of advanced LASER-based experimental tools and techniques. Two comprehensive in-cylinder airflow studies were undertaken to show the contrast between the temporal and spatial distributions of the mean and turbulent gas motions, in the mid-cylinder tumble (and cross-tumble) plane, of a conventional, side-entry, port injected (MPI) engine with flat-top piston and a top-entry G-DI engine with a bowled piston. Airflow measurements were performed using LASER Doppler Anemometry (LDA) in the axial (MPI) and axial and radial (G-DI) directions in a motored, single cylinder, research engine over consecutive engine cycles. Optical access to the combustion chambers was achieved by three different methods. In the MPI engine, a modified piston that incorporated a circular window was used in conjunction with an inclined mirror, fixed below the piston. Experimental measurements in the G-DI were performed along the axis of the spark plug and into the piston bowl and within the upper cylinder and the piston topland region using a spark plug window and glass annular section, sandwiched between the cylinder liner and cylinder head. In the MPI engine, the principal air motion can be described as a forward tumble pattern, established late in the intake stroke and which breaks down into turbulence close to the end of the compression stroke. In the G-DI engine, a strong, reverse tumble motion is rapidly formed early in the intake stroke. This motion persisted throughout the compression stroke and up to TDC without an increase in turbulence intensity. The magnitude of the tumble plane velocity was far greater than that measured in the cross-tumble plane. Observed levels of RMS turbulence intensity were significantly greater in the conventional cylinder head towards the end of the compression stroke indicating the conservation of the reverse tumble motion in the G-DI system. During the early injection phase the airflow exhibited the requirements for a homogeneous charge mixture. For late injection, the persistent coherent tumble structure produced flow stability for a stratified mixture approach. The longitudinal integral length scale of turbulence was determined at two locations near the spark plug gap at ignition timings and TDC conditions. The indirect method and Taylor's hypothesis was used to determine the length scale from the temporal autocorrelation function estimate. The criteria for Taylor's hypothesis for stationary turbulent flow were validated for the coherent, tumbling vortex present in the late compression stroke. Phase Doppler Anemometry (PDA) measurements were taken of a high pressure, swirl type gasoline fuel injector. Droplet size, velocity, RMS velocity and validated data rate measurements were used to quantify the spray atomisation quality and geometry under atmospheric, quiescent conditions and within a motored engine for early and late injection scenarios. High-speed LASER light sheet photography was used for visualisation. In-cylinder PDA measurements were performed at the same locations across the mid-cylinder plane, as employed for the LDA air motion study. The PDA measurements and high-speed photography showed that the fuel spray exhibited a swirling, hollow cone structure during early injection. A quasi-steady state inner cone angle for the spray was estimated. For early injection conditions, a pre-injection, swirless 'slug' of fuel was observed, predominantly along the spray axis, in the time-resolved experiments, that preceded the main injection and asymmetrical cone development and that contained relatively large diameter droplets with high axial velocities. For late injection, the mean droplet diameters were larger than those of the early injection and static tests and the fuel spray formed a narrow jet. The difference in the mean properties was attributed to the nature of the in-cylinder gas properties; An ensemble-averaged, volume-weighted mean velocity was used to describe the momentum exchanged between the gas and liquid phases. It was concluded from the comparative studies of early and late fuel injection, that the droplet size distribution was dominated by secondary atomisation through interaction with the gaseous flowfield when the fuel spray was injected into high gas velocities in the spray-wise direction. The fuel spray injection axis was deflected downwards towards the piston by the momentum of the incoming air charge. The uppermost periphery of the fuel spray was subjected to flapping instabilities and partial detachment. Observations for a suitable mixture preparation strategy are highlighted for a direct injection gasoline engine, based upon the conclusions drawn from a comprehensive experimental study. The effect of in-cylinder airflow on the fuel spray characteristics is quantified for both early and late injection operations. The stability of stratification under motored conditions is discussed in terms of the in-cylinder air motion, turbulence intensity, integral length scale of turbulence and its ability to control the distribution of the mixture and preparation of suitable conditions for initial flame kernel growth. It was concluded that models for mixture preparation based upon pre-conceptions derived from manifold injection, tumble combustion systems and high-pressure Diesel fuel spray analyses are not readily applicable to a strategy that is based upon the direct injection of gasoline into the engine combustion chamber under varying pressure and airflow conditions.

2018

This paper presents a CFD study of Engine Combustion Network (ECN) Spray G, focusing on the transient characteristics of spray at start of injection. The Large Eddy Simulation (LES) coupled with Volume of Fluid (VOF) method is used to model the turbulent two-phase flow. A moving needle boundary condition is applied to capture the internal flow condition accurately. The injector geometry was measured with micron-level resolution using full spectrum x-ray tomographic imaging at Advanced Photon Source (APS) at Argonne National Labora-tory, providing detailed machining error from manufacturer and realistic rough surface. For comparison, a nominal geometry is also used for the simulation. Spray characteristics such as Sauter Mean Diameter (SMD), droplet volume and surface area are extracted by post-processing the CFD outputs. It is seen that compared to the nominal geometry, the use of the high resolution real geometry predicts about 11% lower SMD. The rough surface along with manufactur...

Fuel, 2014

Reliable prediction of spray penetration and spray break-up is required to achieve increases in fuel efficiency and reduction of emissions in diesel engines. Of particular interest is the early transientflow regime. In the current work, diesel fuel spray development was studied using high-speed imaging of a high-pressure diesel common-rail fuel injector mounted in a spherical constant volume combustion chamber. The fuel injector nozzle had four holes aligned on a radial plane with diameters of 90, 110, 130, and 150 µm. Fuel was injected into a room temperature T = 298 K (±1.5%), nitrogen environment at chamber densities of 17.5, 24.2, and 32.7 kg/m 3 (±3%) and for fuel-rail pressures of 1000, 1500, and 2000 bar (±1.5%). Images of the backlit fuel injection were captured at 100,000 frames per second. Image processing algorithms were used to determine fuel spray penetration distance and maximum penetration rate as a function of time. The experimental results for maximum penetration rate and transition time are compared with various quasi-one-dimensional fuel-spray models. The experimental results show departure from the model predictions at higher chamber densities and injection pressures at early times in the spray development. Furthermore, the spray penetration data show 2-dimensional spray geometry changes at early times. A fuel spray tip tracking algorithm was developed to show the maximum penetration distance does not occur along the jet center-line during the transient period of injection and to quantify the angular location of the maximum penetration distance. The data provide valuable insights into transient fuel spray behavior and guide the development of the next generation of spray theory and models.