Search Results

This paper describes cyclic Air/Fuel ratio (AFR) measurements carried out with a novel device (fUEGO) based on a production Universal Exhaust Gas Oxygen sensor, but modified to give an improved frequency response. The results are compared to AFR calculated from a fast CO/CO2 analyser and a fast response flame ionization detector (FID). The direct comparison of the two different methods for determining the cyclic AFR reveals that the electrochemical device is in reasonable agreement with the more complex carbon balance method and can provide reliable cyclic AFR measurements with a reduced requirement for equipment and data post processing. The fUEGO however is sensitive to elevated levels of uHC's (unburned hydrocarbons) during misfires or partial burns and readings during such situations usually show deviations compared to the carbon balance method.

It is well known that accurate EGR control is paramount to controlling engine out emissions during steady state and transient operation of a diesel engine. The direct measurement of EGR is however non-trivial and especially difficult in engines with no external EGR control where the intake manifold CO2 levels can be measured more readily. This work studies the EGR behaviour in a medium duty diesel engine with a passive EGR rebreathing strategy for steady state and transient operation. High speed (response time ∼1ms) in-cylinder sampling using modified GDI valves is coupled with high frequency response analysers to measure the cyclic in-cylinder CO2, from which the EGR rate is deduced. It was found that controlling the EGR using the passive rebreathing strategy during certain combined speed and load transients is challenging, causing high smoke and NO emissions.

A fast response NO detector has been developed to study fast transient emissions from internal combustion engines. The device combines the standard ChemiLuminescence Detector (CLD) measurement technique used in conventional NO detectors with the rapid sampling system of an existing Fast Flame Ionisation Detector (FFID) hydrocarbon detector. The 10-90% response time of the fast NO detector is approximately 3 milliseconds and enables resolution of transient NO concentration within individual engine cycles. Both the fast NO and fast HC detectors were fitted in the exhaust port of a firing SI engine. With the probe tips at the same position, simultaneous fast transient NO and HC concentration data have been recorded during steady state and transient engine load conditions. Cycle-by-cycle NO concentration, HC concentration, and cylinder pressure are compared and features of the transient NO and HC concentration are discussed.

The measured time-history of the cylinder pressure is the principal diagnostic in the analysis of processes within the combustion chamber. This paper defines, implements and tests a pressure analysis algorithm for a Formula One racing engine in MATLAB1. Evaluation of the software on real data is presented. The sensitivity of the model to the variability of burn parameter estimates is also discussed.

Flying Hot-wire anemometer measurements have revealed the effect of the inlet port angle, and valve lift on the spatial structure of turbulence in a model internal combustion engine. The present experiments are carried out at the BDC, intake stroke at a low piston speed in a dynamically similar flow. The data are presented using both cycle resolved and ensemble averaging techniques. The experiments are carried out at three different measurement positions across the engine cylinder. The inhomogeneity of the intake turbulence is found to be considerable. The role of the inlet jet flow is shown to be crucial.

The Fast-response Flame Ionisation Detector is used to measure levels of un-burnt hydrocarbons in the cylinder and exhaust of engines. Its fast response allows uHC emission processes to be resolved within an engine cycle. This paper describes a method for obtaining even greater detail by post-processing the output of the device using a Finite-Impulse Response (FIR) digital filter. The specification of the filter can be obtained by understanding the flow regimes within the sampling system. Examples from in-cylinder and exhaust sampling are presented with suggestions for implementation and further improvement.

Fast Response Flame Ionisation Detectors can be used to observe events within the combustion chamber of engines to help understand the processes by which hydrocarbons escape combustion. In order to understand the timing of events it is necessary to establish the transit time of the measuring system, in particular the times associated with gas transport through the sampling capillaries. To this end, a quasi steady state flow model has been developed to evaluate the transit time for all parts of the cycle. These predictions can deviate significantly from those using steady-state assumptions. Experimental validation is presented.

A fast-FID sampling technique has been developed to study top-land crevice out-gassing from the moving piston of an SI engine. A sampling probe, housed in the piston crown, delivered gas to the FID head via a flexible transfer tube. Comparisons of the HC concentrations at the top-land location and the bulk gas above the piston crown confirm that HC material is out-gassed from the top-land region during the expansion stroke and is followed by more rapid out-gassing after EVO. The removal of wall HCs has been detected during the exhaust stroke from the probable scrolling effect produced by the rising piston scraping unburned material from the cylinder wall.

The action of crevices in an operating SI engine has been studied with a fast-FID. A single-cylinder Ricardo E6 research engine was fuelled with propane and operated at 1300 RPM. FID measurements in the exhaust port have shown that advancing the ignition timing from 30°BTDC (MBT) to 60°BTDC raises the HC concentration by 25% during the first 120°CA of the exhaust stroke and by 20% for the remainder of the stroke. A static “artificial” crevice of known volume, mounted inside the engine cylinder was used to study the differing HC outgassing characteristics at the two ignition timings. When sampling in-cylinder at the mouth of this crevice, the opposite effect of a 50% reduction in outgas HC concentration occurred when the ignition was advanced to 60°BTDC. It is argued that advancing the ignition causes earlier enflamement of the static crevice and induces burned as well as unburned gas to enter the crevice thereby diluting the HCs from this source.

A fast-response FID has been used to study the concentration of hydrocarbon material at four different locations in a firing SI engine. These were: on the flat surface of the cylinder head, at the exhaust valve seat crevice, just downstream of the exhaust seat in the exhaust port and 20mm downstream from the valve stem in the exhaust manifold. A close-fitting sleeve arrangement enabled the sample tube to be positioned accurately flush with the head face and also to be slid away from the wall into the bulk gases whilst maintaining a gas-tight seal. In this way, wall effects could be noted by moving the probe position without stopping the engine and directly comparing with hydrocarbon levels in the bulk gas. Using propane in a fully-warmed up engine, results showed the presence of HCs residing in a 2mm layer adjacent to the wall after EVO and during the exhaust stroke. These could be detected flowing over the valve seat after EVO and were also observed at the manifold location.

Knock in a spark-ignition engine was characterized in terms of its occurrence and magnitude or intensity. Cylinder pressure data from 90 consecutive individual cycles were generated from a single-cylinder engine of disc chamber design at about 72kHz sampling rate over a range of operating conditions between no knock and 100% of the cycles knocking. Mean values and distribution of following parameters were analysed: knock occurrence crank angle, knock intensity, combustion rate and the end-gas thermodynamic state. The effects of fuel octane number and inlet air temperature on these parameters were studied. The thermal imaging technique has been applied to record two-dimensional surface temperatures of cylinder head and piston simultaneously. The change in surface temperatures during knocking and non-knocking cycles was thus studied. As expected, increase in the inlet air temperature or decrease in the fuel octane number caused the knock onset to occur at less advanced spark timing.

The warm-up characteristics of a spark-ignition engine significantly affect fuel consumption and emissions from cars. A thermal imaging technique has been applied to measure the cylinder head surface temperature and piston surface temperature of an internal combustion engine simultaneously. The two-dimensional thermal images of the cylinder head surface temperature were viewed through an infra-red transmitting window mounted in the piston. The piston surface temperature was measured by painting black two small areas of the window's top surface. The similar thermal characteristics of the window material (silicon) to those of a normal piston and good heat transfer between the window and the piston provided realistic operation conditions. The mean and extreme values of the inlet valve, exhaust valve, two other areas of the cylinder head surface and window surface temperatures were measured from the thermal images during the first two minutes of the engine start.

Increasing pressure on lowering vehicle exhaust emissions to meet stringent California and Federal 1993/1994 TLEV emission standards of 0.125 gpm NMOG, 3.4 gpm CO and 0.4 gpm NOx and future ULEV emission standards of 0.04 gpm NMOG, 1.7 gpm CO and 0.2 gpm NOx has focused specific attention on the cold start characteristics of the vehicle's emission system, especially the catalytic converter. From test data it is evident that the major portion of the total HC and CO emissions occur within the first two minutes of the driving cycle while the catalyst is heating up to operating temperature. The use of an electrically heated catalyst (EHC) has been proposed to alleviate this problem but the cost and weight penalties are high and the durability has yet to be fully demonstrated (1)*. This paper describes a method of reducing the light-off time of the catalytic converter to less than 20 seconds by means of an afterburner.

Hydrocarbon wall quenching has been studied using a 19mm diameter, 1m long combustion tube, open at one end. Mixtures of propane, heptane, iso-octane and gasoline, initially quiescent, were burnt with the ignition source at the closed end. The post-flame HC levels were measured at a series of axial locations using a fast FID. The results indicate that the effective quench layer thickness increases significantly as the molecular weight of the fuel is increased. The diffusion/mixing time constant of the quench layer was found to be approximately 0.1s for propane, 0.4s for iso-octane and 1.0s for gasoline. The axial variation of residual HC levels suggests that flame stretch is a factor influencing the extent of the quench layer.

An investigation has been carried out to measure the spatial structure of the turbulence in a model IC engine at near to BDC, inlet stroke. Flying hot wire anemometer measurements have revealed details of the effect of the inlet port angle and valve lift on the spatial turbulence structure. Cycle resolved turbulence intensity and integral length scale are presented at three cutoff frequencies of 10 Hz, 100 Hz, and 1000 Hz. Although the situation is transient, conventional analysis methods are shown to be useful. The inhomogeneity of the intake turbulence is found to be considerable. It is also shown that high turbulence energy content occurs at lower frequencies. The flow also exhibits high cycle-to-cycle variation in the mean velocity although this variation was not sensitive to the cutoff frequency. At low cutoff frequency the integral length scale of BDC turbulence was found to be comparable with the valve lift. The rôle of the inlet jet flow is shown to be crucial.

A thermal imaging system has been developed for viewing and recording the cylinder head surface temperatures of an internal combustion engine. The system consists of an I.R scanner, associated calibration and image processing equipment and an infra-red transmitting window mounted in the piston. The infrared window material used (silicon) has thermal characteristics close to those of a normal piston. The two dimensional temperature distribution of a cylinder head surface has been measured during start-up. The imaging results from the camera were checked against the readings from the thermocouples fitted into the cylinder head. The agreement was very good, and gives confidence in the system.

Two kinds of self-tuning control algorithms have been applied using a novel smoke sensor as part of a diesel smoke control system. Firstly, pole assignment self-tuning control was used to achieve the desired engine speed response, with a smoke predictor used to solve the problems caused by the time delay between the fuel input and smoke output; secondly, a Pl self-tuning controller was used to solve the problem caused by the variation of the coefficients in the transfer functions with engine conditions. The experimental results and the problems and the ideas for future work are discussed.

A novel smoke sensor was used to measure the smoke response to the fuel rack on a diesel engine. The conventional modelling methods used for engine control were investigated. The synchronization technique and Recursive Least Square method were applied to engine modelling and two models for controller design were derived.

A novel smoke sensor was used to realize smoke feedback control on a diesel engine. The controller design based on a combination of PI control algorithm and the engine performance optimization is described. Experimental results demonstrate how this control system behave to meet both of the speed and smoke requirements during engine transients.

Using a novel, high frequency response FID unit, hydrocarbon measurements in the spark plug gap of a firing gasoline engine have been made. These measurements have been correlated with the pressure development, and a significant correlation was found. The method described can be used on any engine fitted with a modified sparking plug.