Search Results

The performance of multi-hole injectors designed for use in second-generation direct-injection gasoline engines has been characterised in a constant-volume chamber. Two types of multi-hole injector have been used: the first has 11 holes, with one hole on the axis of the injector and the rest around the axis at 30 degrees apart, and the second has 6 asymmetric holes located around the nozzle axis. Measurements of droplet axial and radial velocity components and their diameter were obtained using a 2-D phase Doppler anemometer (PDA) at injection pressures up to 120 bar, chamber pressures from atmospheric to 8 bar, and ambient temperatures. Complementary spray visualisation made use of a pulsed light and a CCD camera synchronised with the injection process.

A transparent enlarged model of a six-hole injector used in the development of emerging gasoline direct-injection engines was manufactured with full optical access. The working fluid was water circulating through the injector nozzle under steady-state flow conditions at different flow rates, pressures and needle positions. Simultaneous matching of the Reynolds and cavitation numbers has allowed direct comparison between the cavitation regimes present in real-size and enlarged nozzles. The experimental results from the model injector, as part of a research programme into second-generation direct-injection spark-ignition engines, are presented and discussed. The main objective of this investigation was to characterise the cavitation process in the sac volume and nozzle holes under different operating conditions. This has been achieved by visualizing the nozzle cavitation structures in two planes simultaneously using two synchronised high-speed cameras.

A dense-particle Eulerian-Lagrangian stochastic methodology, able to resolve the dense spray formed at the nozzle exit has been applied to the simulation of evaporating diesel sprays. Local grid refinement at the area where the spray evolves allows use of cells having sizes from 0.6 down to 0.075mm. Mass, momentum and energy source terms between the two phases are spatially distributed to cells found within a distance from the droplet centre; this has allowed for grid-independent interaction between the Eulerian and the Lagrangian phases to be reached. Additionally, various models simulating the physical processes taking place during the development of sprays are considered. The cavitating nozzle flow is used to estimate the injection velocity of the liquid while its effect on the spray formation is considered through an atomisation model predicting the initial droplet size.

The effect of multiple-injection strategy on nozzle hole cavitation has been investigated both experimentally and numerically. A common-rail Diesel injection system, used by Toyota in passenger car engines, has been employed together with a double-shutter CCD camera in order to visualise cavitation inside a submerged and optically accessible (in one out of the six holes) real-size VCO nozzle. Initially the cavitation development was investigated in single injection events followed by flow images obtained during multiple injections consisting of a pilot and a main injection pulse. In order to identify the effect of pilot injection on cavitation development during the main injection, the dwell time between the injection events was varied between 1.5-5ms for different pilot injection quantities. The extensive test matrix included injection pressures of 400 and 800bar and back pressures ranging from 2.4 up to 41bar.

The numerical methodology used to predict the flow inside pressure-swirl atomizers used with gasoline direct injection engines and the subsequent spray development is presented. Validation of the two-phase CFD models used takes place against film thickness measurements obtained from high resolution CCD-based images taken inside the discharge hole of a pressure swirl atomizer modified to incorporate a transparent hole extension. The transient evolution of the film thickness and its mean axial and swirl velocity components as it emerges from the nozzle hole is then used as input to a spray CFD model predicting the development of both non-evaporating and evaporating sprays under a variety of back pressure and temperature conditions. Model predictions are compared with phase Doppler anemometry measurements of the temporal and spatial variation of the droplet size and velocity as well as CCD spray images.

A comprehensive set of computational and experimental results for high- pressure diesel sprays are presented and discussed. The test cases investigated include injection of diesel into air under both atmospheric and high pressure/temperature chamber conditions, injection against pressurized and cross-flowing CF6 simulating respectively the density and flow conditions of a diesel engine at the time of injection, as well as injection into the piston bowl of both research and production turbocharged high-speed DI diesel engines. A variety of high-pressure injection systems and injector nozzles have been used including mechanical and electronic high-pressure pumps as well as common-rail systems connected to nozzles incorporating a varying number of holes with diameters ranging from conventional to micro-size.

A combined two-phase CFD nozzle model and 1-D fuel injection system model is used to predict the flow development inside the discharge hole of a pressure-swirl atomizer connected to a common-rail based fuel injection system for DISI engines. The fuel injection model accounts for the transient pressure pulses developing inside the common-rail and the injector upstream of the nozzle tip and predicts the fuel injection rate through the nozzle. This is then used as input to a 3-D single-phase CFD model estimating the transient development of the swirl velocity inside the pressure-swirl atomizer, as a function of the geometric characteristics of nozzle.

A one-dimensional, transient and compressible flow model was used in order to simulate the flow and pressure distribution in advanced high-pressure fuel injection systems; these include electronic distributor-type pumps with either axial or radial plungers and a common-rail system. Experimental data for the line pressure, needle lift, injection rate and total fuel injection quantity obtained over a wide range of operating conditions (from idle to high speed/full load) were used to validate the model. The FIE system used for validation comprised an electronic high-pressure pump connected to two-stage injectors of different type including 6-hole vertical and 5-hole inclined conical-sac and VCO nozzles.

An enlarged transparent model of a six-hole vertical diesel injector has been manufactured in order to allow flow measurements inside the sac volume and the injection holes to be obtained using a combination of laser Doppler velocimetry (LDV) and the refractive index matching technique under steady state conditions. The measurement points were concentrated in the sac volume close to the entrance of the injection holes as well as inside them on a vertical plane passing through the axis of two injection holes for two different needle lifts. The velocity flow field was characterized in terms of the mean velocity and the turbulent intensity. The results revealed that, under certain conditions, cavitation may occur in the recirculation zone formed at the entrance to the hole since the pressure in this region can reach the value of the vapor pressure of the flowing liquid; this was found to strongly depend on the needle lift and eccentricity.

A computer model simulating the flow in fuel injection systems has been used in order to investigate the fuel injection processes in an in-line pump-based fuel injection system for direct-injection diesel engines. The model is one-dimensional and it is based on the mass and momentum conservation equations for the simulation of the fuel flow and on the equilibrium of forces for the simulation of the mechanical movements of the valves present in the system. The fuel injection system tested comprised an in-line pump whose characteristics were examined by using as input the measured line pressure signal and by modeling the pump operation itself as well as the fuel flow through single- and two-stage injectors. For the validation of the model, extensive comparison with experimental data has been performed for a wide range of pump operating conditions.

A computer model solving the 1-D flow in a typical fuel injection system for direct-injection diesel engines is presented. A Bosch distributor - type VE pump connected to four Stanadyne pencil - type nozzles has been used to validate the computer model over a wide range of operating conditions. Validation of the developed computer code has been performed for eight representative test cases. The predicted values which were compared with the experimental ones include the pumping chamber pressure, the line pressure, the needle lift and the injection rate. Results using as input the measured pumping chamber pressure are also presented in order to identify the error in the injection rate signal attributed to the difference between the simulated and the experimental pumping chamber pressure. In addition, the total fuel injection quantity for pump speeds between 500 and 2000 rpm and lever positions between 20% to 100% was calculated and compared with measurements.

Measurements of flow, spray, combustion and performance characteristics are reported for a Hydra direct-injection diesel, based on the Ford 2.5 L, engine and equipped with a variable-swirl port, a unit fuel injector and optical access through the liner and piston. The results provide links between the pre-combustion and combustion flow and, at the same time, between purpose-built single-cylinder optical engines and multi-cylinder production engines of nearly identical combustion chamber geometry. In particular, the spray penetration was found to depend on engine speed, rather than load, with velocities up to around 260 m/s at atmospheric pressure and temperature which are reduced by a factor of 2.5 under operating conditions and seem to be unaffected by swirl. The duration of combustion was reduced with increasing swirl and ignition delay increased linearly with engine speed.