Search Results

A means of validating numerical simulations has been developed which utilizes chemiluminescence measurements from an internal combustion engine. By incorporating OH*, CH2O* and CO2* chemiluminescence sub-mechanisms into a detailed n-heptane reaction mechanism, excited species concentration and chemiluminescence light emission were calculated. The modeled line-of-sight chemiluminescence emission allows a direct comparison of simulation results to experimentally measured chemiluminescence images obtained during combustion in an optically accessible compression ignition engine using neat n-heptane fuel. The spray model was calibrated using in-cylinder liquid penetration length Mie scattering measurements taken from the jets of the high-pressure piezo injector.

The increased interest for using alternative fuels in modern diesel engines requires better combustion control to achieve safe and efficient operation with fuels characterized by different physical and chemical properties. Knowing the ignition delay and the cylinder peak pressure will allow adapting the injection strategy, mainly injection timing to maintain good engine efficiency when operating with different alternative fuels. The use of the measured instantaneous crankshaft speed to estimate peak cylinder pressure and ignition delay is very attractive because speed is already a parameter in the ECU of the engine. Based on models using powertrain dynamics, the paper presents the development of several techniques using the measured speed to estimate the main combustion parameters for single cylinder and four cylinder diesel engines.

As a result of recent focus on the control of Low Temperature Combustion (LTC) modes, dual-fuel combustion strategies such as Reactivity Controlled Compression Ignition (RCCI) have been developed. Reactivity stratification of the auto-igniting mixture is thought to be responsible for the increase in allowable engine load compared to other LTC combustion modes such as Homogenous Charge Compression Ignition (HCCI). The current study investigates the effect of ethanol intake fuel injection on in-cylinder formaldehyde formation and stratification within an optically accessible engine operated with n-heptane direct injection using optical measurements and zero-dimensional chemical kinetic models. Images obtained by Planar Laser Induced Fluorescence (PLIF) of formaldehyde using the third harmonic of a pulsed Nd:YAG laser indicate an increase in formaldehyde heterogeneity as measured by the fluorescence signal standard deviation.

Fuel wall impingement commonly occurs in small-bore diesel engines. Particularly during engine starting, when wall temperatures are low, the evaporation rate of fuel film remaining from previous cycles plays a significant role in the autoignition process that is not fully understood. Pre-injection chemiluminescence (PIC), resulting from low-temperature oxidation of evaporating fuel film and residual gases, was measured over 3200 μsec intervals at the end of the compression strokes, but prior to fuel injection during a series of starting sequences in an optical diesel engine. These experiments were conducted to determine the effect of this parameter on combustion phasing and were conducted at initial engine temperatures of 30, 40, 50 and 60°C, at swirl ratios of 2.0 and 4.5 at 1000 RPM. PIC was determined to increase and be highly correlated with combustion phasing during initial cycles of the starting sequence.

Ultraviolet chemiluminescence has been observed in a diesel engine cyclinder during compression, but prior to fuel injection under engine starting conditions. During a portion of the warm-up sequence, the intensity of this emission exhibits a strong correlation to the phasing of the subsequent combustion. Engine exhaust measurements taken from a continuously misfiring, motored engine confirm the generation of formaldehyde (HCHO) in such processes. Fractions of this compound are expected to be recycled as residual to participate in the following combustion cycle. Spectral measurements taken during the compression period prior to fuel injection match the features of Emeleus' cool flame HCHO bands that have been observed during low temperature heat release reactions occurring in lean HCCI combustion. That the signal from the OH* bands is weak implies a buildup of HCHO during compression.

EGR is a proven technology used to reduce NOx formation in both compression and spark ignition engines by reducing the combustion temperature. In order to further increase its efficiency the recirculated gases are subjected to cooling. However, this leads to a higher load on the cooling system of the engine, thus requiring a larger radiator. In the case of turbocharged engines the large variations of the pressures, especially in the exhaust manifold, produce a highly pulsating EGR flow leading to non-steady-state heat transfer in the cooler. The current research presents a method of determining the pulsating flow field and the instantaneous heat transfer in the EGR heat exchanger. The processes are simulated using the CFD code FIRE (AVL) and the results are subjected to validation by comparison with the experimental data obtained on a 2.5 liter, four cylinder, common rail and turbocharged diesel engine.

This paper presents an experimental investigation conducted to determine the parameters that control the behavior of autoignition in a small-bore, single-cylinder, optically-accessible diesel engine. Depending on operating conditions, three types of autoignition are observed: a single ignition, a two-stage process where a low temperature heat release (LTHR) or cool flame precedes the main premixed combustion, and a two-stage process where the LTHR or cool flame is separated from the main heat release by an apparent negative temperature coefficient (NTC) region. Experiments were conducted using commercial grade low-sulfur diesel fuel with a common-rail injection system. An intensified CCD camera was used for ultraviolet imaging and spectroscopy of chemiluminescent autoignition reactions under various operating conditions including fuel injection pressures, engine temperatures and equivalence ratios.

The simulation of I.C. Engines operation, especially during transients, requires a fairly accurate estimation of the internal mechanical losses of the engine. The paper presents generic friction models for the main friction components of the engine (piston-ring-liner assembly, bearings and valve train), considering geometry of the engine parts and peculiarities of the corresponding lubrication processes. Separate models for the mechanical losses introduced by the injection system, oil and water pumps are also developed. All models are implemented as SIMULINK modules in a complex engine simulation code developed in SIMULINK and capable to simulate both steady state and transient operating conditions. Validation is achieved by comparison with measurements made on a four cylinder, common rail diesel engine, on a test bench capable to run controlled transients.

The paper analyses the particularities of the lubricating conditions at the contact between the cam and a flat tappet in the valve train of an internal combustion engine and develops a method for the calculation of the friction force. The existing lubrication models show the predominance of the entraining speed and oil viscosity on the thickness of the oil film entrapped between cam and tappet, predicting a very small value (less than 0.1 μm) of the oil film thickness (OFT). The oil viscosity increases exponentially with pressure in the Hertzian contact, determining non-Newtonian behavior of the oil in the contact zone. Using the model developed by Greenwood and Tripp [11] for the contact of two rough surfaces and the Eyring model [2] for the oil it is shown that non-Newtonian behavior of the oil prevails and that the OFT plays a secondary role on the friction force.

The paper presents a method of determining the Indicated Mean Effective Pressure (IMEP) and the gas pressure torque of a multi-cylinder engine using data obtained from the measurement of the crankshaft's speed variation. At steady state operating conditions a Fourier series describe the gas pressure torque of a cylinder and the resultant torque may be obtained by adding the harmonic components corresponding to all cylinders. Only the major harmonic orders, having the same phase for all cylinders add algebraically appearing with large contributions in the spectrum of the resultant torque. The lowest major component has a low frequency and, at this frequency, the crankshaft behaves dynamically like a rigid body. In this situation it is possible to correlate the amplitude of this harmonic order of the gas pressure torque to the same harmonic order of the crankshaft speed.

A dimensionless relationship that estimates the maximum bearing load of a 4-cylinder 4-stroke in-line engine has been found. This relationship may assist the design engineer in choosing a desired counterweight mass. It has been demonstrated that: 1) the average bearing load increases with engine speed and 2) the maximum bearing load initially decreases with engine speed, reaches a minimum, then increases quickly with engine speed. This minimum refers to a transition speed at which the contribution of the inertia force overcomes the contribution of the maximum pressure force to the maximum bearing load. The transition speed increases with an increase of counterweight mass and is a function of maximum cylinder pressure and the operating parameters of the engine.

The paper analyses the friction in the valve train of an internal combustion engine trying to separate the contribution of the different components to the total friction losses in the valve train. The measurements are performed on a running engine in order to avoid extraneous factors introduced by simulating rigs. The experimental engine is instrumented with strain gauge bridges on the rocker arm, the push rod and the camshaft to measure forces and moments acting on these components. Original techniques are developed to isolate and determine the friction forces between the valve stem and its guide, the friction force in the rocker arm bearing and the combined friction between cam/tappet and tappet/bore. It was found that the friction in the rocker arm bearing never reaches hydrodynamic conditions and that the friction coefficient between cam and tappet reduces with an increase in the engine speed.

Even under steady state operating conditions, the pressure variation in individual cylinders, and the corresponding gas-pressure torque are subjected to small random fluctuations from cycle to cycle. The gas-pressure torque of a cylinder may be expressed as a sum of harmonically variable components, each harmonic being affected by these fluctuations. A probabilistic model of the vector interpreting such a harmonic component is developed and used to determine the statistical parameters of the resultant random vector representing the corresponding harmonic order of the engine torque. At the low frequencies of the lowest harmonic orders of the engine torque the crankshaft behaves like a rigid body. This behavior permits to correlate the statistical parameters of the same harmonic components of the resultant torque and of the measured engine speed. This correlation is proved by experiments and used to identify faulty cylinders.

The non-uniform character of the torque produced by a reciprocating I.C. engine is reflected in the cyclic variation of the crankshaft's speed. Because the crankshaft is an elastic structure, its response to the different harmonic components of the torque is different and changes with engine speed. The lowest harmonic components of the engine torque do not excite torsional vibrations and correlate fairly well with the corresponding harmonic orders of the crankshaft's speed. Based on a random vector model of the harmonic components of the gas-pressure torque, a statistical correlation is obtained between amplitudes and phases of the same harmonic component of the gas-pressure torque and of the crankshaft's speed. The lowest major harmonic order determines the average IMEP of the engine and the half-order detects if a cylinder is a lesser contributor to the total engine output and identifies the deficient cylinder.

The Indicated Mean Effective Pressure (IMEP) is a very important engine parameter, giving significant information about the quality of the cycle that transforms heat into mechanical work. For this reason, modern data acquisition systems display, on line, the cylinder pressure variation together with the corresponding IMEP. The paper presents a very simple algorithm for the calculation of IMEP, based on the correlation between IMEP and the gas pressure torque. It was found that that the IMEP may be calculated by a very simple formula involving only two harmonic components of the cylinder pressure variation. The computation of the two harmonic components is very easily performed because it does not involve the calculation of an average pressure and the cylinder volume variation. The method was experimentally validated showing differences less than 0.2% with respect to the IMEP calculated by the traditional method.

The effects of the start/stop (S/S) transients on the Hybrid Electric Vehicle (HEV) operation and emissions are explored in this study. The frequency with which the engine starts and stops during an urban driving cycle is estimated by using the NREL's Advanced Vehicle Simulator software (ADVISOR). Furthermore, several tests were conducted on single-cylinder and multi-cylinder direct injection diesel engines in order to measure the cycle-resolved mole fractions of the hydrocarbons and nitric oxide exhaust emissions under frequent start/stop mode of operation. The frictional losses in engine in its entirety as well as in its components are also determined. In addition, the dynamic behavior of different high pressure fuel injection systems are investigated under the start and stop mode of operation.

This paper presents a global friction model of a diesel engine. The model accounts for the individual contributions of the main components of the mechanical losses and the influence of specific design and operating parameters on the mechanical losses. The main components considered in the model are: the piston-ring assembly, the valve train, the bearings and auxiliaries (injection pump, oil pump and coolant pump). For each of these components, the model was developed based on geometric parameters, operating conditions and the physics governing the friction. The individual models were assembled in a global friction model of a multicylinder diesel engine, and a computer code was developed to simulate the total mechanical losses of the engine. The experimental validation of the model was obtained by comparing the simulated crankshaft's speed variation with the instantaneous speed measured by a shaft encoder.

The cylinder block of a high-speed marine diesel engine is a complex structure subjected to a complex loading. The design optimization of the cylinder block requires a reliable Finite Element Model (FEM), capable to predict, with a reasonable accuracy, the actual strains and stresses. The experimental investigation presented in the paper is meant to provide the necessary information for a better estimation of the boundary conditions and the validation of the FEM of the cylinder block. In order to obtain an image of the stress field in the cylinder block, a system of 10 strain gauge rosettes have been placed at significant locations on the cylinder block. The temperature at the location of the rosettes was measured with an optical pyrometer and a method has been developed to calculate this temperature using the measured strain. A fairly good agreement was obtained between the measured and the calculated temperatures during the cooling of the engine.

This paper presents a simplified piston ring assembly (PRA) friction model accounting for the piston ring pack and the piston skirt. The ring model considers both mixed and hydrodynamic lubrication; the skirt model considers hydrodynamic lubrication only. The Reynold's equation is used as a governing equation for the hydrodynamic regimes of both models. Simplified assumptions are used for the mixed lubrication in the ring model. The ring model generates unique Stribeck curves for a given ring's geometry; the skirt model generates generic relationships between the friction force and skirt geometry, piston speed, oil viscosity, and assumed boundary conditions. Ring starvation effects are introduced by varying the boundary conditions, as appropriate. The results of the models are compared to measurements made on a motored and fired single cylinder diesel engine; the theoretical calculations provide a reasonable estimate of the measured data.

The variation of the crankshaft's speed is influenced by the action of the cylinders and shall reflect the contribution of each cylinder to the total engine output. At the same time, the speed variation is influenced by the torsional stiffness of the cranks, the mass moments of inertia of the reciprocating mechanisms and the average speed and load of the engine. As the result, the variation of angular motion of the crankshaft is complex, each particular influence changing its importance as speed and load are modified. The diagnostic method presented in the paper is based on the analysis of the amplitudes and phases of the lowest harmonic orders of the measured speed and is capable to determine the average Indicated Mean Effective Pressure (IMEP), to detect nonuniformities in cylinder operation and to identify the faulty cylinder(s).