Search Results

The objective of this paper is to investigate how the combustion chamber design will influence combustion parameters and emissions in a natural gas SI engine. Ten different geometries were tried on a converted Volvo TD102 engine. For the different combustion chambers emissions and the pressure in the cylinder have been measured. The pressure in the cylinder was then used in a one-zone heat-release model to get different combustion parameters. The engine was operated unthrottled at 1200 rpm with different values of air/fuel ratio and EGR. The air/fuel ratio was varied from stoichiometric to lean limit. EGR values from 0 to 30% at stoichiometric air/fuel ratio were used. The results show a remarkably large difference in the rate of combustion between the chambers. The cycle-to-cycle variations are fairly independent of combustion chamber design as long as there is some squish area and the air and the natural gas are well mixed.

The cylinder to cylinder and cycle to cycle variations were measured in a production type Volvo natural gas engine. Cylinder pressure was measured in all six cylinders. Emission measurements were performed individually after all cylinders, and commonly after the turbocharger. Measurements (ECE R49 13-mode) were performed with different spark gap and two different locations for fuel injection, one before the throttle and one before the turbocharger. Heat-release and lambda calculations show substantial cylinder to cylinder variations, due to lambda variations between the cylinders. The slow burn combustion chamber, with low turbulence, results in high cycle to cycle variations (> 100% COV imep) for some of the load cases.

An ionization sensor has been used to study the combustion process in a six-cylinder lean burn, truck-sized engine fueled with natural gas and optimized for low emissions of nitric oxides. The final goal of the investigations is to study the prospects of using the ionization sensor for finding the optimal operating position with respect to low NOx emission and stable engine operation. The results indicate that unstable combustion can be detected by analyzing the coefficient of variation (CoV) of the detector current amplitude. Close relationships between this measure and the CoV of the indicated mean effective pressure have been found during an air-fuel ratio scan with fixed ignition advance.

The present paper aims at discussing the flow/flame interaction in a lean burn spark ignition engine. The mean velocity and cycle resolved turbulence are measured with laser Doppler velocimetry. The cylinder pressure is recorded and a one-zone heat release calculation performed. The very early part of flame propagation is measured using two orthogonal Schlieren systems, each capturing one image of the progressing flame at a given time after spark onset. The two resulting 2D images are then, after preprocessing, used to reconstruct the three-dimensional flame. The volume of the true flame is estimated by simulating, using Markov Chain Monte Carlo techniques, a number of possible flames that are consistent with the projections on the images. The uncertainty of the estimated flame volume is given by the variation of the volume estimates. In the calculations, the volume of that part of the spark plug that is inside the flame is subtracted.

One way to extend the lean burn limit of a natural gas engine is by addition of hydrogen to the primary fuel. This paper presents measurements made on a one cylinder 1.6 liter natural gas engine. Two combustion chambers, one slow and one fast burning, were tested with various amounts of hydrogen (0, 5, 10 and 15 %-vol) added to natural gas. Three operating points were investigated for each combustion chamber and each hydrogen content level; idle, part load (5 bar IMEP) and 13 bar IMEP (simulated turbocharging). Air/fuel ratio was varied between stoichiometric and the lean limit. For each operating point, a range of ignition timings were tested to find maximum brake torque (MBT) and/or knock. Heat-release rate calculations were made in order to assess the influence of hydrogen addition on burn rate. Addition of hydrogen showed an increase in burn rate for both combustion chambers, resulting in more stable combustion close to the lean limit.

Exhaust Emissions from lean burn natural gas engines may not always be as low as the potential permits, especially engines with open loop lambda control. These engines can produce much higher emissions than a comparable diesel engine without exhaust gas after treatment. Even if the engine has closed loop lambda control, emissions are often unacceptably high for future emission regulations. A three way catalyst is, today, the best way to reduce hazardous emissions. The drawback is that the engine has to operate with a stoichiometric mixture and this leads to; higher heat losses, higher pumping work at low to medium loads, higher thermal stress on the engine and higher knock tendency (requiring lower compression ratio, and thus lower brake efficiency). One way to reduce these drawbacks is to dilute the stoichiometric mixture with EGR. This paper compares lean burn operation with operation at stoichiometric conditions diluted with EGR, and using a three way catalyst.

2D-LDV-measurements were made on the flow from one transfer channel into the cylinder in a small two-stroke SI engine. The LDV measuring volume was located just outside the transfer port. The engine was a carburetted piston-ported crankcase compression chainsaw engine and it was run with wide open throttle at 9000 RPM. The muffler was removed to enable access into the cylinder. No additional seeding was used; the fuel and/or oil was not entirely vaporized as it entered the cylinder. Very high velocities (-275 m/s) were detected in the beginning of the scavenging phase. The horizontal velocity was, during the whole scavenging phase, higher than the vertical.

This paper discusses the effects of cooled EGR on a turbo charged multi cylinder HCCI engine. A six cylinder, 12 liter, Scania D12 truck engine is modified for HCCI operation. It is fitted with port fuel injection of ethanol and n-heptane and cylinder pressure sensors for closed loop combustion control. The effects of EGR are studied in different operating regimes of the engine. During idle, low speed and no load, the focus is on the effects on combustion efficiency, emissions of unburned hydrocarbons and CO. At intermediate load, run without turbocharging to achieve a well defined experiment, combustion efficiency and emissions from incomplete combustion are still of interest. However the effect on NOx and the thermodynamic effect on thermal efficiency, from a different gas composition, are studied as well. At high load and boost pressure the main focus is NOx emissions and the ability to run high mean effective pressure without exceeding the physical constraints of the engine.

2-D LDV measurements were performed on two different cylinder designs in a fired two-stroke engine running with wide-open throttle at 9000 rpm. The cylinders examined were one with open transfer channels and one with cup handle transfer channels. Optical access to the cylinder was achieved by removing the silencer and thereby gain optical access through the exhaust port. No addition of seeding was made, since the fuel droplets were not entirely vaporized as they entered the cylinder and thus served as seeding. Results show that the loop-scavenging effect was poor with open transfer channels, but clearly detectable with cup handle channels. The RMS-value, “turbulence”, was low close to the transfer ports in both cylinders, but increased rapidly in the middle of the cylinder. The seeding density was used to obtain information about the fuel concentration in the cylinder during scavenging.

Heavy trucks contribute significantly to the overall air pollution, especially NOx and PM emissions. Models to predict the emissions from heavy trucks in real world on road conditions are therefore of great interest. Most such models are based on data achieved from stationary measurements, i.e. engine maps. This type of “quasi stationary” models could also be of interest in other applications where emission models of low complexity are desired, such as engine control and simulation and control of exhaust aftertreatment systems. In this paper, results from quasi stationary calculations of fuel consumption, CO, HC, NOx and PM emissions are compared with time resolved measurements of the corresponding quantities. Measurement data from three Euro 3-class engines is used. The differences are discussed in terms of the conditions during transients and correction models for quasi stationary calculations are presented. Simply using engine maps without transient correction is not sufficient.