Search Results

A direct comparison is made of time-resolved measurements of diesel PM emissions obtained using laser-induced incandescence (LII), an electrical low pressure impactor (ELPI), and a tapered element oscillating microbalance (TEOM). The measurements were made on two diesel passenger vehicles, one of which was equipped with a diesel particulate filter. Both LII and the ELPI performed well for both vehicles, whereas the TEOM lacked the sensitivity required for the filtered vehicle. We estimate that the LII system used has a limit of detection better than 0.2 mg/mi.

A new diagnostic technique is described that has the capability of making real-time, in situ measurements of the volatile fraction of diesel particulate matter (PM). LIDELS uses two laser pulses of comparable energy, separated in time by an interval sufficiently short to freeze the flow field, to measure the change in PM volume caused by laser-induced desorption of the volatile fraction. The first laser pulse produces elastic light scattering (ELS) that gives the volume of the total PM, and also deposits the energy to desorb the volatiles. ELS from the second pulse gives the volume of the remaining solid portion of the PM, and the ratio of these two measurements is the quantitative solid volume fraction. Calibration is required for the individual total PM and solid fraction to be quantitative. Applicability of the technique is demonstrated for load and EGR sweeps for a turbocharged, direct-injection diesel engine.

Laser-induced incandescence (LII) is a sensitive diagnostic technique capable of making exhaust particulate-matter measurements during transient operating conditions. This paper presents measurements of LII signals obtained from the exhaust gas of a 1.9-L TDI diesel engine. A scanning mobility particle sizer (SMPS) is used in fixed-size mode to obtain simultaneous number concentration measurements in real-time. The transient studies presented include a cranking-start/idle/shutdown sequence, on/off cycling of EGR, and rapid load changes. The results show superior temporal response of LII compared to the SMPS. Additional advantages of LII are that exhaust dilution and cooling are not required, and that the signal amplitude is directly proportional to the carbon volume fraction and its temporal decay is related to the primary particle size.

A comparison of scanning mobility particle sizer (SMPS) and laser-induced incandescence (LII) measurements of diesel particulate matter (PM) was performed. The results reveal the significance of the aggregate nature of diesel PM on interpretation of size and volume fraction measurements obtained with an SMPS, and the accuracy of primary particle size measurements by LII. Volume fraction calculations based on the mobility diameter measured by the SMPS substantially over-predict the space-filling volume fraction of the PM. Correction algorithms for the SMPS measurements, to account for the fractal nature of the aggregate morphology, result in a substantial reduction in the reported volume. The behavior of the particulate volume fraction, mean and standard deviation of the mobility diameter, and primary particle size are studied as a function of the EGR for a range of steady-state engine speeds and loads for a turbocharged direct-injection diesel engine.

A flame-kernel model is formulated for the analysis of data obtained using a recently developed spark plug equipped with fiber-optic flame-arrival detectors. The detectors measure the elapsed time from ignition to flame arrival at the detector locations for each engine cycle. The model, which assumes a flame kernel of elliptical cross section undergoing growth and convective displacement at constant rates, is used to estimate those rates from cycle-resolved measurements. It is shown that convection-rate estimation, ostensibly an interpolation of detector signals, in some cases involves an extrapolation that is sensitive to model assumptions. Implications concerning experimental procedure and data interpretation are discussed.

A recently developed spark plug equipped with fiber-optic flame-arrival detectors has been used to measure the motion and rate of growth of the early flame kernel. The cylinder pressure and gas velocity in the spark gap were measured simultaneously with the flame kernel measurements, permitting the data to be analyzed on a cycle-by-cycle basis to identify cause-and-effect correlations between the measured parameters. The data were obtained in a homogeneous-charge research engine that could be modified to produce three very different flow fields: (1) high swirl with high turbulence intensity, (2) tumble vortex with moderate turbulence intensity, and (3) negligible bulk motion with low turbulence intensity. The results presented show a moderate correlation between the combustion duration and the rate of growth of the flame kernel, but virtually no correlation with either the magnitude or direction of movement of the flame kernel away from the spark gap.

The in-cylinder flow field of a Schnürle (loop) scavenged two-stroke engine has been examined under conditions simulating both blower and crankcase driven scavenging. Measurements of the radial component of velocity were obtained along the cylinder centerline during fired operation at delivery ratios of 0.4, 0.6, and 0.8. Both mean velocity profiles and root mean square velocity fluctuations near top center show a strong dependence on the scavenging method. Complementary in-cylinder pressure measurements indicate that combustion performance is better under blower driven scavenging for the engine geometry studied. IN THE PAST TEN YEARS the engine research and development community has demonstrated a renewed interest in two-stroke engine technology. Many manufacturers have new engine designs operating on test stands and in prototype vehicles being road tested.

Double-pulse laser-induced desorption with elastic laser scattering (LIDELS) is a diagnostic technique capable of making time-resolved, in situ measurements of the volatile fraction of diesel particulate matter (PM). The technique uses two laser pulses of comparable energy, separated in time by an interval sufficiently short to freeze the flow field, to measure the change in PM volume caused by laser-induced desorption of the volatile fraction. The first laser pulse of a pulse-pair produces elastic laser scattering (ELS) that gives the total PM volume, and also deposits the energy to desorb the volatiles. ELS from the second pulse gives the volume of the remaining solid portion of the PM, and the ratio of these two measurements is the quantitative solid volume fraction. In an earlier study, we used a single laser to make real-time LIDELS measurements during steady-state operation of a diesel engine.

Video imaging has been used to investigate the evolution of liquid fuel films on combustion chamber walls during a simulated cold start of a port fuel-injected engine. The experiments were performed in a single-cylinder research engine with a production, four-valve head and a window in the piston crown. Flood-illuminated laser-induced fluorescence was used to observe the fuel films directly, and color video recording of visible emission from pool fires due to burning fuel films was used as an indirect measure of film location. The imaging techniques were applied to a comparative study of single and dual spray fuel injectors for both open and closed valve injection, for coolant temperatures of 20, 40 and 60°C. In general, for all cases it is shown that fuel films form in the vicinity of the intake valve seats.

Laser-induced incandescence is used to measure time-resolved diesel particulate emissions for two lean NOx trap regeneration strategies that utilize intake throttling and in-cylinder fuel enrichment. The results show that when the main injection event is increased in duration and delayed 13 crank-angle degrees, particulate emissions are very high. For a repetitive pattern of 3 seconds of rich regeneration followed by 27 seconds of NOx-trap loading, we find a monotonic increase in particulate emissions during the loading intervals that approaches twice the initial baseline particulate level after 1000 seconds. In contrast, particulate emissions during the regeneration intervals are constant throughout the test sequence.

A photographic study is presented illustrating the influence of mixture motion on flame propagation in an internal combustion engine. Variation in swirl and turbulence levels was achieved by rotating the orientation of a shroud on the intake valve. Laser Doppler velocimetry was used to characterize the precombustion fluid motion. A flexible shadowgraph system was developed for visualizing in-cylinder events. The results show that cyclic variation is not necessarily decreased by increasing the burn rate. The fastest burn achieved in this study occurred with high swirl, when the flame remained attached to the spark plug. If random detachment of the flame occurred, however, cyclic variation was greatly enhanced.

Measurements are presented showing the effect of swirl level and spark location on burn duration in a homogeneous-charge engine. Laser shadowgraph photographs of the flame structure were used to help interpret the observed results. As expected, without swirl the burn duration was a direct function of flame travel distance, such that central ignition was optimal. When swirl was introduced, off-axis ignition was aided by flame-holder effects that enhanced the flame speed in the circumferential direction. However, only for the highest swirl level studied (swirl number = 8.3) was the burn rate increased by moving the ignition point toward the cylinder wall. For lower swirl levels, central ignition was still preferable.

Laser Doppler velocimeter results are presented for the mean velocity and turbulence intensity measured in a motored research engine. The compression of complex bulk motions created during induction produces turbulence as the piston approaches top dead center. The turbulence field is shown to be isotropic but nonhomogeneous. A zero-dimensional computer simulation based on an averaged k-ϵ model is shown to adequately predict the decay of turbulence at a point in the flow after the production phase is completed. Cylinder pressure measurements were recorded for homogeneous stoichiometric combustion for a range of engine speeds and ignition locations. A two-zone (burned and unburned gases) thermodynamic model accurately predicts the measured pressure histories when the turbulence results determined from the motored tests are used to establish initial conditions for the combustion model.

A hot-wire anemometer was used to study the air motion in a motored i.c. engine. Measurements were made of the mean velocity, turbulence intensity, and integral scales of turbulence. The engine speed was varied from 500 to 2500 rpm, and the hot-wire probe was traversed both across the combustion chamber clearance volume and down into the piston sweep volume. The latter traverse was accomplished by probe-accommodating “wells” built into the piston crown, which were subsequently shown to severely disrupt the flow during the compression and expansion strokes. The results show the mean velocity and turbulence intensity to vary linearly with engine speed, and the turbulence scales to be a function of geometry only. The structure of turbulence was found to be inhomogeneous in the clearance volume and the upper portion of the sweep volume.

Measurements are presented for the turbulence intensities and mean velocities obtained in a research engine in which a grid was used to create a flow field characterized by negligible mean motions and homogeneous and isotropic turbulence at the time of ignition. Pressure measurements for homogeneous stoichiometric combustion indicate a very low level of cyclic variation. The combustion-induced mean flow field is shown to be characteristic of a one-dimensional compression of the unburned gases, which produces a small increase in the bulk turbulent kinetic energy ahead of the flame. Most of the effect of combustion appears to occur locally, as the turbulence in the preflame gases close to the flame front is strongly amplified in the direction of flame propagation. Parallel to the flame surface there is little effect until the flame has propagated nearly all the way across the chamber.

A two-dimensional axisymmetric numerical model is used to illustrate how the presence of walls can significantly influence the shape of a spark-ignited premixed gas flame, even when wall boundary layers are neglected. This in-viscid model is based on tracking the flame interface as it interacts with the combustion-generated flow field. Comparisons made with a quasi-dimensional phenomenological model show that the assumption of a spherical flame surface held centered at the ignition location can lead to a large underprediction of the flame area. This occurs because the spherical flame is prematurely truncated by the chamber walls, and because the spherical assumption constrains the flame surface-to-volume ratio to be the absolute minimum. In contrast, results from the two-dimensional model show that the flame slows down as it approaches a wall since the velocity induced by volume expansion must be zero normal to the boundary.

An optical probe for measuring the motion and rate of growth of the early flame kernel in spark ignition engines is described. The probe consists of a standard spark plug with eight optical fibers installed in a ring at the base of the threaded region of the plug. The fibers collect the light emitted from the flame as it crosses the field of view of the fibers, and transmit the light to photomultiplier tubes. The time from ignition until detection of the flame is used to compute the average flame velocity in the direction of each fiber relative to the spark location. The real-time data acquisition system permits statistical analysis of cycle-by-cycle variations in the combustion rate. Because the probe was built using a standard 14 mm spark plug, it can be used in unmodified production automotive engines.

Ionization probes installed in the head gasket of a spark ignition engine are used to measure the cycle-resolved arrival time of the flame at eight discrete points at the perimeter of the cylinder bore. Simultaneous data acquisition of the ionization probe and cylinder pressure measurements permits the flame burn pattern, the combustion rate, and the cyclic variability of these quantities to be observed on a video monitor m real-time as engine operating parameters are varied. To demonstrate the technique, measurements are presented for uniformly-spaced and clustered arrangements of ionization probes and differing conditions of fluid motion, spark location, spark plug configuration, and equivalence ratio.

An experiment is described for the direct measurement of the turbulent burning velocity during premixed combustion in a spark ignition engine. The gas velocity is measured using a high data rate laser Doppler velocimeter system that resolves the unburned gas motion on an individual cycle basis. The ensemble-averaged flame speed is determined from ionization probe measurements of the time of flame arrival at discrete positions along the path of flame propagation. The difference between the cycle-resolved unburned gas velocity and the ensemble-averaged flame speed gives a direct measurement of the turbulent burning velocity that is unbiased by cyclic variations in the combustion rate. The value of burning velocity obtained is shown to be in close agreement with an empirical model previously determined.

Laser Doppler velocimetry has been used to measure velocity and turbulence intensity profiles in the wall boundary layer of a spark-ignited homogeneous-charge research engine. By using a toroidal contoured engine head it was possible to bring the laser probe volume to within 60 μm of the wall. Two different levels of engine swirl were used to vary the flow Reynolds number. For the high swirl case under motored operation the boundary layer thickness was less than 200 μm, and the turbulence intensity increased as the wall was approached. With low swirl the 700-1000 μm thick boundary layer had a velocity profile that was nearly laminar in shape, and there was no increase in turbulence intensity near the wall. When the engine was fired the boundary layer thickness increased for both levels of swirl.