Search Results

A system for both ignition and ion current measurement was designed and set up at Istituto Motori. Particular attention was paid to the problem of dissipating the residual energy stored in the ignition coil, reducing the electromagnetic interferences and especially improving the response of the measurement system. In order to assess the capability of the ion current signal to give reliable and accurate information for knock detection, a number of tests were carried out at full load on a commercial PFI four cylinder engine, at various air/fuel ratios and spark timings. Some knock indices based on the ionization signal, both band pass filtered and non-filtered, were introduced, in particular: the Amplitude of the Second Ionization Peak (ASIP), the Mean not filtered Ionization Current signal (MIC), the Maximum Amplitude of Ionization Current signal Oscillation (MAICO), the Integral of Modulus of filtered Ionization Current signal Oscillation (IMICO).

There is an increasing interest on innovative controls based on in-cylinder pressure measurement. However, due to current capabilities of ECUs, only a small number of pressure cycles can be sampled. Therefore suitable algorithms are required to extract from few pressure points the information useful to implement the desired control strategies. In this paper a method for the interpolation of in-cylinder pressure cycle during the constant mass phase of S.I. engines is described. It is based on the estimation of Cumulative Heat Release (CHR) in selected crank angle positions and on a simple algorithm of pressure reconstruction. To test this method, peak pressure position and magnitude have been identified on the basis of a set of experimental pressure cycles measured on a gasoline direct injection engine. In addition, a control scheme for the closed loop control of spark timing have been set up using a numerical engine (WAVE) and a numerical ECU build in the Matlab-Simulink environment.

Several methods have been proposed to use pressure signal for air/fuel ratio estimation, knock detection and optimal spark timing selection. In this paper some of these methods were compared, and their accuracy and effectiveness was checked. In order to avoid the misleading effects of measurement errors, the comparison was performed using a database of test conditions obtained by means of the WAVE code (Ricardo). New correlations physically based were introduced to evaluate the trapped air mass and the Exhaust Gas Recycling (EGR), cylinder per cylinder. These correlations can give a very important contribution to balance the air-fuel ratio in each cylinder and to improve EGR control strategies.

Experimental measurements and numerical computations were made to characterize a spray generated by a high-pressure swirl injector. The Phase Doppler technique was applied to get information on droplet sizes (d10) and axial velocities at defined distances from the injector tip. Global spray visualization was also made. Computations were carried out using a modified version of KIVA 3V. In particular, the break-up length of the sheet and its dimension were computed from a semi-empirical correlation related to the wave instability theory suggested by Dombrowski, including the modifications introduced by Han and Reitz. Two different approaches were used to describe the initial spray conditions. According to the first, discrete particles with a characteristic size equal to the thickness of the sheet are injected. The second approach assumes, that the particles having a SMD computed by a semi-empirical correlation are injected according to a statistical distribution.

The electromechanical devices proposed in technical literature are very flexible. However, their working principle (fixed lift, fixed lifting time and variable valve events) imposes the use of different strategies, such as cylinder or port deactivation to enable work at partial loads. This happens in particular at low loads. The present paper aims to evaluate the effect of the design and of the strategies adopted to vary the load (cylinder or port deactivation etc) on performance and on pollutant formation of a stoichiometric DISI engine. The calculations were performed by a commercial one-dimensional code (Wave produced by Ricardo). This tool was also used to give inputs to the design of the electromechanical actuator. The electromechanical design of the actuator was carried out with the aid of the code Flux2D produced by Cedrat. This code allows the complete simulation of transients and of the electrical losses of the actuator.

A number of Variable Valve Actuators (VVA) has been recently proposed to improve the performances and the part load efficiency of spark ignition engines. Due to their peculiarity, these systems work with different strategies (late or early inlet valve closing, reduced lift etc.). The shape and the timing of the valve lift affect not only the pumping losses, but also air motion inside the cylinder. That influences mixture formation and combustion evolution of Direct Injection Spark Ignition (DISI) engines. The present paper compares the performances of different VVA systems with the aid of a 1D code for the simulation of the inlet and of the exhaust phases, and of a fluid-dynamic 3D code to evaluate mixing phenomena inside the cylinder.

The paper gives a short outline of experiences gained at IM in the design of an electromechanical valve actuator having two magnets and two balanced springs. This device works at fixed lift, fixed lifting time and variable valve event. Some of most relevant topics regarding the design of such a type of actuator have been analyzed. In particular, in the first part of the paper the problems of varying load have been tackled with the aid of a commercial one-dimensional code. Based on these results, the electromechanical design of the actuator has been started. Three axial-symmetric geometries have been compared using an electromagnetic CAD. The requirements of systems to perform the first lift at engine start up and the catching in operation have been analyzed. The evaluation of energy losses connected with these phases has been also done.

An experimental and numerical investigation, using the Laser Doppler Anemometry (LDA) technique and a 3D fluid-dynamic code (KIVA 3V), was carried out in a prototype engine under steady-state conditions. The aim of the present activity was the flow field characterization and the effect of the intake geometry on the in-cylinder tumble flow. A new steady flow test rig designed for capturing the tumble motion within a test cylinder, made by a blower and an engine head, was assembled to simulate the intake flow. The engine head was mounted on an aluminum cylinder, having the same bore as the real engine. The cylinder was provided with optical accesses on the periphery and a flat optical window located at the bottom to a depth equal to the stroke of the engine. The cylinder was also equipped with two cylindrical ducts, used as air outflow ports.

A simple model of combustion and pollutant formation has been set up. It is part of an engine simulator to be used for the study of engine control strategies. The calculation of inlet and exhaust phases is performed by an emptying and filling method, based on the knowledge of mean inlet and exhaust conditions. A single zone thermodynamic model has been utilized for the calculation of the combustion phase. The values of the shape factors of heat release patterns have been modeled to take into account air/fuel ratio, EGR, load and turbulence at ignition starting. Crevice storage of unburned mixture has been considered as the dominant mechanism for unburned HC production. A model for mixing and burning of HC inside the cylinder has been proposed. NO is calculated using the three steps Zeldovich approach. The model produces realistic calculations of combustion pressure and pollutants emission at various speed, load, ignition timing and EGR.

The effect of external EGR on knock was evaluated using a CFR engine. Combustion pressure was sampled on a time basis. A band pass filter in the time domain was applied to the pressure cycles. Five knock indices were calculated for each combustion cycle. The problem to quantify knock intensity was focused. At this extent measurements were carried out on standard isooctane-n-heptane blends in the test conditions used for the determination of the Motor Method Octane Number (MON). Knock intensity was varied acting on compression ratio. For each index, the conditions of absence of knock were determined using motored cycles. The indices were compared and one of them, showing the lowest C.O.V., was selected for further measurements. The effect of EGR on test fuels having different composition was evaluated varying the compression ratio, at fixed ignition timing. In this way, the same level of detonation, obtained in the absence of EGR, was realized with different amount of external EGR.

A fuel matrix consisting of twelve gasolines was tested in presence of Exhaust Gas Recirculation (EGR). The fuels have different percentages of aromatics (20÷35% vol.), olefins (5÷15% vol.) and oxygen (0÷2% wgt). Four different oxygenated compounds (MTBE, ETBE, TAME, DIPE) were chosen as additives. Tests were carried out on a MPI premixed spark ignition engine at steady operating conditions (2000 rpm, 2 bar BMEP, 13.5% EGR) and stoichiometric air/fuel ratio. Regulated and unregulated pollutants were measured upstream the catalytic converter. Cyclic variation of Indicated Mean Effective Pressure (IMEP) in presence of EGR was also evaluated. The adoption of EGR increases PAH and aldehydes emissions, and decreases benzene emissions of unoxygenated fuels. Conversion efficiencies of CO and of total HC are lowered by EGR. An increase of aromatics content in an unoxygenated fuel leads to higher engine out NOx emission. This effect is reduced if MTBE is added.

The present paper describes a calculation procedure aimed to evaluate the evaporative loss of the lubricating oil deposed on the cylinder surface of i.c. engines. The model is based on the simultaneous solution of the diffusion and energy equations referred to the liquid/gas interface. A parametric analysis has been carried out to investigate the influence of some variables on oil evaporative losses. The calculation results show that both cylinder surface temperature and lubricating oil composition are significant parameters. Conversely, the oil film thickness does not seem to play an important role in evaporative loss. Furthermore, it has been ascertained that the lubricating oil evaporation is increased during the intake stroke. In any event, the evaporative loss represents a small percentage (a few percentage points) of total oil consumption if the distillate fractions are not too much light.

3D calculations of homogeneous charge spark ignition engines were carried out using the KIVA III code. A modified wall function was introduced by an approximate solution of the one -dimensional simplified equations of energy and mass balance. The model takes into account the pressure unsteadiness and the mean rate of combustion in boundary layer. Moreover a modified turbulent conductivity law was proposed following the classical Prandtl approach. The predictions of heat transfer model were compared with the mean heat flows calculated by thermodynamic processing of pressure cycles in motored engines. Two engines with different geometry were used. Namely: a CFR engine running 900 rpm and an AVL engine, running at 2200 rpm. The results regarding heat transfer seem very encouraging. The combustion phase was simulated using a Fractal Flame Model (FFM) elsewhere describe. Simulations in firing conditions were compared with measurements carried out on a CFR engine and on an AVL engine.

A commercial four stroke spark ignition engine has been tested at steady conditions, with three different compression ratios, namely: 10, 11.5 and 13. Exhaust Gas Recycle (EGR) has been varied in the range 0% - 20 %. Air/fuel ratio has been maintained at stoichiometric by a closed loop control with Exhaust Gas Oxygen sensor feedback. Significant gains on fuel economy and CO emission index have been achieved at medium and high loads by the simultaneous adoption of EGR and high compression ratios. In these conditions the sum of HC and NOx emission indices attains significant reductions at any load. The tests have shown that EGR allows to avoid knock even at wide open throttle and Maximum Brake Torque timing.

The present paper describes some experiments aiming to point out the link between oil consumption and reverse blowby. Some tests have been carried out on a motored single cylinder diesel engine. The reverse blowby gas mass flow has been evaluated by a thermodynamical model that utilizes both the measured combustion and second land pressures, and the blowby gas mass flow. Oil consumption has been measured in real time using a CO2-tracer method, whereas the blowby has been measured by a fast response orifice meter. The first ring lifting has also been recorded. It has been observed that, under certain engine operating conditions, both blowby and oil consumption assume quite constant levels. On the contrary, under other operating conditions, they vary in a cyclical way. However, in both cases, a relationship between blowby, reverse blowby and oil consumption can be recognized.

Multidimensional computations of homogeneous charge spark ignition engines were made with the KIVA II code. Combustion was simulated using the Fractal Flame Model of Zhao [5]. The original code was modified to obtain better calculations of heat transfer and to take into account the mass flow in the crevices. The predictions were compared with measurements carried out on a CFR engine. The tests were carried out in stoichiometric condition with isooctane. Compression ratio, ignition timing and EGR level were selected as test parameters. The global agreement between calculations and experiments was evaluated on the basis of heat release, indicated pressure patterns and pollutants measurements. For the lower compression ratio (7.7) the predictions of pressure cycle generally were in good agreement with experiments. However the empirical constant used in this condition cannot be used at higher compression ratio to obtain acceptable predictions of the pressure cycle.

Nowadays most passenger cars are equipped with spark ignition engines with a three way catalyst. Thus, the improvement of fuel consumption of this type of engine represents a very attractive goal. In fact, it may cause a reduction of pollutant emission, and simultaneously, it may give a contribution to the lowering of global CO2 production. In this paper, a strategy to control the combustion process of stoichiometric spark ignition engines is described. It is based on the adoption of Exhaust Gas Recycle (EGR) in high compression ratio engines. The tests carried out have shown that EGR can control the knock, even at Wide Open Throttle (WOT), with a compression ratio of about 13.5. Improvements of efficiency higher than 10%, at different loads and speeds, have been achieved by the adoption of this technique. Similar improvements have been obtained for CO, while more substantial reductions have been measured for NOx.

A numerical technique, aimed to the reconstruction of the analog output of an analyzer during continuous exhaust gas analysis, is presented. To this purpose the system composed by sample line and analyzer is described as a discrete Linear Time Invariant system with Finite Impulse Response. This technique has been tested on the reconstruction of the continuous emission measurements of diluted exhaust, obtained during a driving cycle acted on a chassis dynamometer. A comparison with the results obtained with CVS bag analysis has been made. The air/fuel ratio during the test cycle has been evaluated and compared with the signal of an oxygen sensor. An attempt to evaluate the emission indices in the transients has been also made, comparing the results of reconstructed and non reconstructed signals.

In the present paper a mathematical model on ring pack behavior is presented. The program considers the aspect of gas flow into and from the inter-ring volume and the relative ring dynamic. Furthermore a proper mass balance on the oil film has been considered to automatically evaluate both starvation and the oil accumulation in front of the inlet boundary of each ring. The model can give quite accurate predictions of the gas flows and the oil film thicknesses. It may be considered the first step for the simulation of oil mist formation and evaporation that are the most important phenomena for oil consumption prediction. UBRICATING OIL gives a strong contribution to particulate formation in diesel engines. Moreover it influences the unburned hydrocarbon emission of spark ignition engines because of the absorption/desorption phenomenon between the unburned fuel and the lubricating oil films [1, 2].

An evaluation of emission levels and fuel consumption of urban hybrid vehicles has been performed. Both heavy and light duty vehicles have been considered and a comparison on specific consumptions and emissions has been carried out between the hybrid and traditional configurations. The battery behavior during charge and discharge transients has been taken into account because it is one of the most critical elements of hybrid systems. The results of this investigation indicate the possibility to reach significant reductions of consumption and emissions through the adoption of hybrid systems for urban transports.