Search Results

Local measurements of the mean and rms velocities have been obtained by laser Doppler velocimetry in the coolant passages of a transparent model of a Ford 2.5L diesel cylinder head and block at a steady flowrate of 6.83 × 10-4 M3/s. The simulation of the coolant fluid by a mixture of hydrocarbon fluids at a predetermined constant temperature allowed accurate matching of the refractive index to that of the acrylic model, thus providing optical access for LDV measurements of the internal flow in sensitive areas where cooling is essential to prevent metal-fatigue failure. The results were obtained in sufficient detail to allow further validation of CFD coolant flow models.

Laser Doppler velocimetry, phase Doppler anemometry and Mie scattering were applied to a single-cylinder, four-valve, spark-ignition gasoline research engine equipped with a fully transparent liner and piston, to obtain information about the tumble flow and the droplet size and velocity distributions during induction and compression, for lean air/fuel mixture ratios of 17.5 and 24 and with closed-valve and open-valve fuel injection. The mixture distribution obtained with the two injection strategies was correlated with flame images, pressure analysis and exhaust emissions which confirmed the advantages of combining open-valve injection with tumble to allow stable and efficient engine operation at an air/fuel ratio of 24 through charge stratification and faster flame growth.

The ability of certain induction systems to enhance turbulence levels at the time of ignition, through formation of long-lived tumbling vortices on the plane of the valve and cylinder axes, has been investigated in a two-valve spark-ignition engine by rotating the intake port at 90° and 45° to the orientation of production directed ports. Detailed measurements of the three velocity components, obtained by laser velocimetry, revealed that the 90° port generated a pure tumble motion, with a maximum tumbling vortex ratio of 1.5 at 295°CA, zero swirl, and 42% turbulence enhancement relative to the standard configuration, while the 45° port gave rise to a combined tumble/swirl structure with a maximum tumbling vortex ratio of 0.5 at 285°CA, swirl ratio of 1.0 at TDC, and turbulence enhancement of 24%. The implications of the two types of flow structures for combustion are discussed.

The spatial and temporal characteristics of a diesel spray injected into the atmosphere through a multi-hole nozzle used in small DI Diesel engines have been investigated by laser techniques as a function of pump speed and load. The results showed that spray tip penetration and velocity depend on injection frequency rather than injected volume and the spray is asymmetric during the early and main part of the injection period. In the time/space domain different structures have been identified within the injection period, with the early injection period characterized by a well atomized cloud of droplets, the main period by the spray head and a dense core and the late injection period by the disintegrating dense core and the spray tail. IN DIRECT-INJECTION DIESEL ENGINES for passenger cars, fuel is injected through multi-hole nozzles at high pressure to promote mixing with the rapidly swirling air inside the combustion chamber.

Heat release analysis, ion current and laser Doppler velocimetry methods are combined to examine the effect of pre-spark turbulent mixture motion on flame propagation and combustion characteristics in the four-valve pentroof combustion chamber of a spark ignition engine. Two inlet port configurations leading to different strengths of induction-generated tumbling vortices are considered with emphasis on lean mixture operation under partial load conditions. The results show a good correlation between mixture turbulent characteristics at ignition and flame development, flame propagation rate, combustion duration and cyclic variability. It is shown that turbulence enhancement through induction-generated tumble counterbalances the lean mixture ignition delay by enhancing the flame propagation speed, leading to extension of the engine lean operating limits.

A constant-volume combustion chamber was used to examine injection of a small quantity of slightly rich fuel/air mixture towards the spark plug around the time of ignition, in an overall very lean mixture rotating at velocities representative of modern spark-ignition engines. The results show that it is possible to achieve 100% ignitability with overall air-fuel ratios in excess of 50 and much faster burn rates than those with initially homogenous mixtures of the same equivalence ratio with high swirl and turbulence. The advantages of this method of local charge stratification have been demonstrated in terms of both pressure measurements and shadowgraphs of the early flame development while the transient characteristics of the injected rich mixture at the spark plug gap were monitored by a fast flame ionization detector.

Two optical single-cylinder spark-ignition engines equipped with two- and four-valve cylinder heads were used to examine the flow and flame interaction under lean mixture conditions. Images of the developing flame under quiescent, swirl, low tumble and high tumble flow conditions corresponding to a wide range of mean velocity and turbulence levels around the time of ignition were obtained with an image-intensified CCD camera using the light radiated by the flame and the flow in the vicinity of the spark plug was quantified by laser Doppler velocimetry. In the case of the tumbling flow, the flame images were software-processed to allow estimation of the total flame area, the displacement of its centre as a function of crank angle and their correlation with the cylinder pressure.

Measurements are presented of droplet characteristics and air velocity in the cylinder of a 0.36 litre four valve engine, equipped with an sohc VTEC-E valve train and port injection. The results show that injection at crank angles, θinj(s), when the inlet valve is open results in most of the liquid volume flux being in the form of droplets with Sauter mean diameter between 20 and 30 mm which strikes the sleeve up to about 2.5 cm below the exhaust valves, thus generating a locally rich cloud there. The amount of liquid phase gasoline passing through the plane 16 mm below the spark plug gap increases with θinj(s) up to 50 CA after intake TDC and this, together with the crank angle of droplet arrival and vapour generation, controls stratification of the gaseous fuel phase. The optimum injection time is when the fuel-rich cloud is generated so that the tumble vortex convects it to the spark plug at the time of ignition.

Measurements of cylinder pressure and flame travel velocity have been obtained in a single cylinder engine with two arrangements of port geometry and with mixture equivalence ratios from 0.68 to 0.9. They are complemented by photographs of the flame development and measurements of local velocity. The investigation compares the combustion processes in terms of the maximum pressure, flame speed and in-cylinder flow velocity without and with an intake shroud which increased both the tumble and swirl ratios. The extent to which residual burned gas retarded the combustion rate and increased cyclic variability are quantified. The photographic studies confirm the dominant effect of the swirling flow on flame propagation and deviations of the flame kernel from spherical as the air-fuel ratio is increased, with much higher probability of influence of velocity fluctuations.

Previous measurements of the velocity, size and number density of droplets have been reported in one cylinder of a production two-valve engine as a function of position, crank angle, injection timing, rotational speed, load and cooling water temperature. In this paper, similar measurements are reported in two cylinders of the same engine, this time with four cylinders firing, and with two manifolds and injectors. They were obtained with a phase-Doppler velocimeter with measurements ensembled in relation to an optical shaft encoder. The engine was also instrumented to provide air and fuel flow rates and temperatures. The results show that most of the droplets emerge in a comparatively small region of the inlet valve and that the characteristics of the spray are important mainly when injection takes place with the inlet valve open.

The performance of a four-valve engine operating with combustion in all cylinders has been determined in terms of indicated mean-effective pressure, drivability and concentrations of unburned hydrocarbon in the exhaust gases with a stoichiometric mixture of gasoline and air and four injectors including two with air assist. In addition, size and velocity characteristics of the fuel sprays were measured with a phase-Doppler velocimeter outside and inside the engine. With operation at a steady rotational speed of 1200 rpm, the indicated mean- effective cylinder pressure and its covariance were found to be nearly constant with the initiation of injection from 150 to 600 degrees of crank angle after top-dead-centre of intake.

The droplet velocity, size and distributions of iso-octane fuel from single hole and twin jet air-assist injectors have been measured by phase Doppler velocimetry in the pent-roof for two cylinder head designs of firing four-valve engines running at 1500 rpm, together with the airflow during induction and compression. The use of the twin jet air-assist injector together with the introduction of a transfer-passage between the two intake ports of a Honda VTEC-E valve train arrangement resulted in reduction in ISNOx and COV-1mep of the order of half of those with the single hole injector design without a transfer passage. Droplets, for both heads and injectors, having passed the inlet valves, impinged directly onto the sleeve opposite to their entry without striking the exhaust valves and had velocities up to 30 m/s and Sauter mean diameters which varied from 20 to 50pm.

Measured values of velocity and associated turbulence properties have been obtained in a motored single-cylinder Diesel engine and in a plexiglass simulation engine. The quality of the signals obtained from thereal engine and the corresponding useful data rate are quantified and discussed. As a consequence, it is shown that measurements in real engines are likely to relate to crank-angle intervals of the order of 10 degrees. The implications of this conclusion are quantified by analysis and by measurements in the plexiglass engine. The results, from the motored Diesel engine, also show that values of velocity cannot be measured throughout the cylinder cavity.

Ensemble-averaged and in-cycle axial and swirl velocities have been measured by laser Doppler anemometry in the three-dimensional flow field of a four-stroke model engine motored at 200 rpm with a compression ratio of 6.7 and various cylinder head and piston geometries. The inlet configurations comprised an axisymmetric port with a shrouded valve and an off-centre port with two valve and swirl generating vane geometries. The piston configurations comprised flat, cylindrical and re-entrant axisymmetric piston-bowls. The results indicate that with the off-centre port a complex vortical flow pattern is generated during induction, which later either collapses in the absence of induction swirl or is transformed into a single rotating vortex in the transverse plane when swirl is present. The axisymmetric port with the shrouded valve gives rise to a double vortex structure and higher turbulence levels at TDC of compression compared to the off-centre port.

Measurements of three velocity components have been obtained by forward-scatter laser Doppler anemometry in the transparent cylinder of a modified production engine motored in the speed range 300-2000 rpm with a disc-type combustion chamber and a compression ratio of 7.4. The in-cylinder flow development has been examined in detail with and without induction swirl at an engine speed of 1000 rpm. In both cases turbulence was non-isotropic during compression with a tendency towards isotropy at TDC. The swirl velocities at TDC of compression scaled linearly with engine speed while the volume-averaged turbulence intensity varied more than linearly with engine speed for both induction system configurations, increasing from 0.45 to 0.6 times the mean piston speed in the absence of induction swirl and from 0.4 to 0.5 with induction swirl.

The origin and development of swirl center precession in engine flows has been investigated in a steady flow rig, with and without a porous plate simulating a stationary piston, and in a model engine motored at 200rpm; swirl, in all cases, was generated by means of 60° vanes located in the axisymmetric inlet port. The swirl center performs a helical motion that originates as an instability in the forced-vortex core from its interaction with the axial flow at a free stagnation point and develops in the engine from the piston towards the cylinder head; an opposite trend has been observed in the steady flow case with the open-ended cylinder. In the ensemble-averaged measurements, swirl center precession has been identified by the increased tangential velocity fluctuations around the off-centre zero swirl velocity.

The interaction of fuel sprays and airflow in the intake system of a port fuel-injected spark-ignition engine has been examined experimentally in a pulsating-flow rig which comprised the cylinder head and intake manifold of a production engine connected to a large-capacity plenum chamber, with the camshaft of the intake valves driven by an electrical motor at engine speeds between 1000 and 5000 rpm and with air sucked through the system by a suction fan. Static pressure measurements in the intake port showed periodic pulsations with frequencies of 360 and 200 Hz with open and closed valves, respectively, and these corresponded to quarter- and half-waves in the manifold and were independent of engine speed.