Search Results

Results of laboratory studies of the static characteristics of several different commercially available heated exhaust gas oxygen sensors are described. In these studies, the emf of the sensors was measured as a function of temperature and of the composition of calibrated gas mixtures. Several different binary gas mixtures (H2/N2, CO/N2, C3H6/N2, C3H8/N2, and CH4/N2) were used together with a variable amount of O2. In addition to laboratory studies, the same sensors were also studied in the exhaust gas of an engine. Whereas at high temperatures thermodynamic equilibrium appears to prevail, clear departures from thermodynamic equilibrium are observed at some lower temperatures (the value of which depends on the specific sensor and the specific gas mixture used). This behavior is manifested by shifts of the emf step away from stoichiometry, broadening of the step, abnormally high emf values in excess oxygen mixtures, and abnormally low emf values in reducing gas mixtures.

Quantitative measurements were made of NOx feedgas emissions during transient engine operation as the response time of an EGR system was progressively-degraded. For a simple acceleration-cruise-deceleration engine speed/torque versus time trajectory, it was found that the NOx emissions were higher during acceleration and lower during deceleration than corresponding values predicted from steady-state mapping data. The magnitude of the differences, as well as the total mass of NOx integrated over the speed/torque trajectory, all increased as the EGR response time was increased. Using a simple dynamic EGR model, NOx feedgas emissions were predicted for engine operation with a production EGR system over a 128 second portion of the FTP CVS cycle. The NOx feedgas predictions were shown to be in excellent agreement with actual emission measurements.

Vehicle chassis dynamometer tests were performed to compare predicted and measured total feedgas emissions and fuel economy for dynamic operation of an engine. In general, these tests showed that predictions based on steady-state mapping data agreed well with measured values except for NOx emissions. Subsequent engine-dynamometer tests indicated that the discrepancy between predicted and measured NOx emissions was due to competing effects of combustion chamber wall temperature and dynamic EGR time response. A technique was developed which utilized the results of a simple transient test to improve the accuracy of predicting NOx emissions when EGR time response was not a factor. The effect of degraded EGR time response on both instantaneous and total NOx emissions was also explored.

An engine-dynamometer study was performed to quantify the air-fuel ratio (A/F) offset between the window of a three-way catalyst (TWC) and the closed-loop control point of an exhaust gas oxygen (EGO) sensor. In this study, the effects of rpm, torque, EGR, and A/F modulation were explored along with the age of the TWC and EGO sensor. In general, it was determined that the closed-loop EGO sensor control point shifts lean as a function of increasing feedgas NOx concentration, thus causing the engine A/F to move away from the high NOx conversion efficiency regime of the TWC.

The transient response of ZrO2 and TiO2 EGO sensors has been investigated under actual engine operating conditions. The results of this study show that the response of an EGO sensor is dependent upon the characteristics of the engine and feedback control system with which it is used. Specifically, sensor response time is a function of the magnitude and frequency of the A/F changes and the initial and final values of A/F to which the sensor is exposed. ZrO2 and TiO2 sensors show similar transient behavior and have practically equivalent response times.

Results of three-way catalyst (TWC) conversion efficiency measurements are presented which demonstrate the effects of dynamic operation. In one series of measurements, the time-averaged conversion efficiencies were determined for the case in which the TWC was subjected to a series of complex air-fuel ratio (A/F) waveforms. The measured efficiencies are compared to predicted efficiencies determined from the superposition of the TWC response to the individual sinusoidal components comprising the complex waveform. This comparison shows discrepancies for CO in the lean region, and for HC and NOx in the rich region. In a second series of measurements, the instantaneous conversion efficiencies were determined for a square wave A/F waveform using a timed-sampling technique.

An engine-dynamometer test procedure for evaluating the transient A/F characteristics of fuel metering systems is described. Actual transient A/F data obtained with a prototype multi-point electronic fuel injection (EFI) system are presented, and an electronic compensation scheme for reducing the A/F transient excursions is discussed.

A procedure which uses modulated engine exhaust to characterize the amplitude vs. frequency response of air-fuel ratio (A/F) measuring equipment is described, and typical results obtained when the procedure is applied to a standard Lamdascan A/F Analyzer are given. A technique for extending the frequency response of the Lamdascan instrument by a factor of approximately two is discussed, and the measured characteristics which result when the improvements are incorporated into the instrument are presented.

A brief physical description of ZrO2 and TiO2 exhaust gas oxygen sensors is presented, followed by a comparison of the two sensors operating under actual open-loop and closed-loop conditions on an engine/dynamometer setup. The sensitivity of the ZrO2 sensor to engine load and cylinder-to-cylinder maldistribution is explored, and the compatibility of the ZrO2 switch point and a three-way catalyst window is considered.

A prototype vaporized gasoline metering system is described which utilizes engine exhaust heat to vaporize liquid gasoline prior to being combined with inlet air. It is shown that the system (1) exhibits minimal time-fluctuations in air-fuel ratio, (2) essentially eliminates the transient variations in air-fuel ratio due to load changes, and (3) provides a very uniform cylinder-to-cylinder distribution of air-fuel ratio. The use of the vapor system at very lean air-fuel ratios is considered, and a CVS cycle prediction of the lean-limit operation is presented.