Search Results

A system has been developed that allows near real time measurements of total, volatile, and nonvolatile particle concentrations in engine exhaust. It consists of a short section of heated catalyst, a cooling coil, and an electrical aerosol analyzer. The performance of this catalytic stripper system has been characterized with nonvolatile (NaCl), volatile sulfate ((NH4)2 SO4), and volatile hydrocarbon (engine oil) particles with diameters ranging from 0.05-0.5 μm. The operating temperature of 300°C gives essentially complete removal of volatile sulfate and hydrocarbon particles, but also leads to removal of 15-25% of solid particles. This system has been used to determine total, volatile, and nonvolatile particle concentrations in the exhaust of a Diesel engine and a spark ignition engine. Volatile volume fractions measured in Diesel exhaust with the catalytic stripper system increased from 19-65% as the equivalence ratio (load) decreased from 0.64-0.13.

A test reactor was designed for a 6.7 kW, 303 cc, single cylinder, air cooled, gasoline fueled engine. The reactor was very efficient at hydrocarbon (HC) and carbon monoxide (CO) reductions - with up to 99.9 and 98.6% removed, respectively. It had no effect on oxides of nitrogen (NOx) emissions. With the reactor, the engine met the California Air Resources Board (ARB) proposed Tier II emission standards. A factorial test was used to determine that A/F ratio and air injection rate significantly affected CO reduction efficiency whereas air injection location, ignition timing, and engine load did not. Relationships were established between CO reduction, air injection rate, and reactor core temperature.

Light absorption and scattering coefficients have been measured for particles emitted by two diesel engines; one direct injection and one indirect injection, operating over a range of speeds and loads. Integrating plate absorption measurements yield a specific absorption coefficient of 9.1 square meters per gram of non-volatile particulate matter at 550 nm wavelength. This absorption coefficient is inversely proportional to wavelength and independent of engine operating conditions or type. The scattering coefficient was simultaneously measured as 1.3 square meters per gram of undifferentiated particulate matter. These experimental results are shown to be in the range predicted by theoretical absorption and scattering calculations, which have been made for elongated carbon-void particles.

The U.S. Bureau of Mines has sponsored research Co determine the particle size distribution and concentration of submicron particles upstream and downstream of a ceramic particle trap mounted in the exhaust stream of a Caterpillar 3304 diesel engine. Particle size distribution and mass were measured with an electrical aerosol analyzer, a diffusion battery-condensation nuclei counter combination, and filters. The engine was operated at 1400 and 1800 RPM and 3 load conditions at each speed-In general, the collection efficiency of the trap was high, ranging between 89 to 96%. Size distribution analysis revealed that the trap was generally more efficient at removing particles smaller than 0.1 µm diameter than larger particles. However, under certain conditions formation of nuclei (less than 0.056 µm diameter) downstream of the trap took place.

Particle mass and size distribution measurements have been made on the exhaust of an Onan prechamber diesel engine. Seven fuels were examined: no. 1 and no. 2 diesel fuel, 40 and 50 cetane number secondary reference fuels, and no. 2 diesel fuel doped with three different concentrations of Lubrizol 565, a barium-based smoke suppressant. The no. 1 and no. 2 diesel fuels and the 50 cetane number reference fuels produced very similar emissions with emission indices in the range 0.3-1.3 mg (gm-fuel)-1 and volume mean diameters between .09 and 0.15 μm. The 40 cetane number reference fuel produced both smaller emission indices, 0.2 to 0.8 mg (gm-fuel)-1, and particle diameters, 0.03 to 0.09 μm. These reductions were apparently related to the longer ignition delay period of the 40 cetane number fuel, which allowed better mixing of the fuel and air prior to combustion.

This paper describes the performance and emissions of a hydrogen-fueled, spark-ignited engine. An electronic control device, designed to provide the engine with a timed injection of the fuel, is shown to give high mean effective pressures and high efficiencies. The oxides of nitrogen from the exhaust gases have been analyzed and the mechanism for their formation is reviewed. The paper further describes an experiment with traces of hydrocarbons added to the hydrogen in an attempt to explain any additional phenomena that may be taking place during the combustion, such as “prompt NO” which is known to occur in hydrocarbon flames only. As it turns out, such additions have a negligible effect on the NOx formation in the region investigated.

Vibration signals from diesel engines were analyzed for the purpose of isolating signals relating to injection or combustion which could be used to time the engines. Nonintrusive sensors, magnetically attached to the engine, were used to obtain these vibration signals. Components believed to be associated with combustion or fuel injection were electronically isolated from the remaining engine noise, and subsequently processed to produce specific timing signals. Digital data acquisition and averaging methods were used, coupled with computerized frequency analysis. The signals were experimentally correlated with the combustion process over a wide range of injection timing. The electronic processing system developed provides a real time digital measure of the timing. Data on the accuracy and correlation of experimental measurements will be presented.

Particle formation, growth, coagulation and combustion in the cylinder of an indirect injection passenger car type diesel engine have been studied using a system which allows the cylinder contents to be rapidly expelled through a blowdown port, diluted, and collected in a sample bag for subsequent analysis. Characteristic blowdown times were about 0.5 ms. Samples were analyzed using a condensation nuclei counter to determine particle number concentrations and an electrical aerosol analyzer to determine particle volume concentrations in the 0.01 to 1.0 μm diameter range. Measurements were made with the engine operating at 1000 rpm and an equivalence ratio of 0.32. Peak particle number concentration in the cylinder 13 times the exhaust level, and peak particle volume (or mass) concentration in the cylinder 3 times the exhaust level were observed. These results suggest that significant particle coagulation and oxidation occur during the expansion stroke.

The performance of a closed-loop electronic fuel injection timing control system for diesel engines has been investigated, both experimentally and analytically. The Electronic Control System (ECS) studied is a version of the “Optimizer,” a peak seeking control which can find the maximum of one variable with respect to another. In this case, it was used to find the timing for maximum brake torque (MBT). The ECS can also be operated in a “biased” mode in which it will hold the timing either advanced or retarded of MBT, but in a fixed relationship to it. Performance and emissions of a medium duty engine equipped with the ECS were measured on an engine dynamometer. The results clearly demonstrate that, for a variety of operating conditions and for two fuels, the ECS can find and hold the timing at MBT or in fixed relationship to it.

A unique system which allows sampling of the entire contents of one of the cylinders of a 5.7-liter V-8 indirect-injection diesel engine has been developed. An explosively actuated cutter ruptures a diaphragm in the combustion chamber and allows the contents of the cylinder to rush out and be subsequently diluted and quenched with cool nitrogen. Particles are collected with a high-volume impactor/filter system. This system has been used to collect a series of particle samples at crankangles ranging from 5 to 40 degrees after top dead center. Particle samples from the exhaust were also obtained. The samples have been extracted to determine the soluble organic fraction. These extracts have been analyzed for five polycyclic aromatic compounds: pyrene, fluoranthene, benz(a)pyrene, benz(k)fluoranthene, and 1-nitropyrene. The results indicate significant removal of the first four between the combustion chamber and the exhaust manifold.

In-cylinder and exhaust soot mass measurements have been made on a single-cylinder conversion of a 4-cylinder, 2.8 1, high-swirl, direct-injection diesel engine using a sampling system which allows dumping, diluting, quenching, and collecting the entire contents of the cylinder on a time scale o£ about 1 ms. Experiments have been performed at engine speeds of 1,000 and 1,500, and equivalence ratios, ϕ, of 0.4 and 0.7. Soot mass first appears shortly after top dead center and reaches a peak between 15 and 30 crankangle degrees after top dead center (CAD ATDC). After reaching its peak value, soot concentration decreases with increasing crankangle and approaches exhaust levels by 40-60 CAD ATDC. The time lag between the start of combustion and the first appearance of soot increases with ϕ and ranges from 0.2 to 1 ms. The initial rate of soot formation ranges from 0.26 to 0.30 mg ms−1 and varies little with speed or ϕ.

Time resolved primary and agglomerate particle size distribution measurements have been made on samples obtained from within the cylinder and from the exhaust of a single-cylinder modification of a 2.8 liter displacement, four-cylinder, naturally-aspirated, high swirl, direct-injection diesel engine. The total cylinder sampling method has been used to sample, quench, and dilute the entire contents of the cylinder in about 1 ms. Experiments have been performed at an equivalence ratio of 0.7 and a speed of 1000 RPM. An electrostatic aerosol sampler and a transmission electron microscope have been used to determine primary and agglomerate particle size distributions for both in-cylinder and exhaust samples. An electrical aerosol analyzer and a diffusion battery followed by a condensation nucleus counter were used to further characterize the agglomerate size distributions of exhaust samples.

An adaptive control system which determines the optimum system parameters based on the engine response to changes in those parameters, has been tested as an ignition timing control system on several gaseous fueled engines. The changes in the MBT timing for speed, load, air-fuel ratio, and fuel type were explored. The ability of the control system to correct the timing for these parameters was demonstrated. An air-fuel ratio control based on the same technique is also discussed.

Diesel particles carry charges ranging from 1-5 units of elementary charge per particle. The charge is bipolar, i.e., there are approximately equal numbers of positively and negatively charged particles. The charge distribution with respect to size follows a high temperature, Boltzmann equilibrium relationship. The first part of this paper describes charge measurements made on diesel particles emitted by three different diesel engines, and postulates a charging mechanism. The second part of the paper is an examination of how this natural charge may be used to collect particles from the exhaust. The charge level produced by combustion is only slightly lower than the charge level produced by the corona discharge in a conventional electrostatic precipitator. Thus, a simple electrostatic precipitator without a corona section will collect diesel particles affectively.

Measurements of particle concentrations in one cylinder of a 1982 5.7 liter GM V-8 diesel engine have been made using a unique total cylinder sampling system. The first part of the paper is devoted to an examination of the performance of the sampling system. The role of blowoff and nucleation in the formation of sample artifacts is discussed. The remainder of the paper is devoted to the results of a study of the formation and removal of carbon particles during diesel engine combustion. Several operating conditions have been examined. The influence of injection timing, load, EGR, and oxygen addition on particle formation and removal has been investigated. The concentrations of volatile and nonvolatile particulate matter have been measured as a function of crankangle position. Particle formation begins 1-5 crankangle degrees (CAD) after the start of combustion.

This study examines the effect of a ceramic particle trap and a manganese fuel additive on the mutagenic activity and chemical composition of diesel exhaust particulate matter from a heavy-duty mining engine. Particles were collected by dilution tunnel sampling from a 4-cylinder, Caterpillar 3304, naturally-aspirated, indirect-injection engine operated at six steady-state conditions. Depending on engine load and speed the ceramic particle trap reduced the following emissions: particulate matter, 80 – 94%; soluble organic fraction (SOF), 83 – 95%; 1-nitropyrene, 94 – 96%; and SOF mutagencity, 72% (cycle-weighted average). When the Mn fuel additive was used without a ceramic particle trap the total cycle mutagenic activity emitted increased 7-fold, in part, due to elevated emissions of 1-nitropyrene.

This paper describes an examination of the origin of the response of a real-time exhaust particle sensor. The sensor works by detecting the net electrical charge carried by diesel exhaust particles emitted during exhaust blow-down. The distribution of charge on these particles has been measured using an electrical mobility analysis system. The results show that the exhaust particles are highly charged and that their charge distributions are nearly symmetrical. The sensor signal results from a slight departure from this symmetry. The results suggest that most of the charge on the exhaust particles results from bipolar charging by flame ions during combustion, but that the net charge detected by the sensor results from surface interactions which some of the larger particles undergo during exhaust blowdown.

Experiments were performed to measure the number-weighted particle size distributions emitted from a gasoline direct injection (GDI) engine. Measurements were made on a late model vehicle equipped with a direct injection spark ignition engine. The vehicle was placed on a chassis dynamometer, which was used to load the engine to road load at five different vehicle speeds ranging from 15 - 100 km/hr. Dilution of the exhaust aerosol was carried out using a two-stage dilution system in which the first stage dilution occurs as a free jet. Particle size distributions were measured using a TSI 3934 scanning mobility particle sizer. Generally speaking, the presence of the additives did not have a strong, consistent influence on the particle emissions from this engine. The polyether amine demonstrated a reduction in particle number concentration as compared to unadditized base fuel.

Nanoparticle formation during exhaust sampling and dilution has been examined using a two-stage micro-dilution system to sample the exhaust from a modern, medium-duty diesel engine. Growth rates of nanoparticles at different exhaust dilution ratios and temperatures have been determined by monitoring the evolution of particle size distributions in the first stage of the dilution system. Two methods, graphical and analytical, are described to determine particle growth rate. Extrapolation of size distribution down to 1 nm in diameter has been demonstrated using the graphical method. The average growth rate of nanoparticles is calculated using the analytical method. The growth rate ranges from 6 nm/sec to 24 nm/sec, except at a dilution ratio of 40 and primary dilution temperature of 48 °C where the growth rate drops to 2 nm /sec. This condition seems to represent a threshold for growth. Observed nucleation and growth patterns are consistent with predictions of a simple physical model.