Search Results

This paper reports on a turbulent flame propagation model combined with a zero-dimensional two-zone stochastic reactor model (SRM) for efficient predictive SI in-cylinder combustion calculations. The SRM is a probability density function based model utilizing detailed chemistry, which allows for accurate knock prediction. The new model makes it possible to - in addition - study the effects of fuel chemistry on flame propagation, yielding a predictive tool for efficient SI in-cylinder calculations with all benefits of detailed kinetics. The turbulent flame propagation model is based on a recent analytically derived formula by Kolla et al. It was simplified to better suit SI engine modelling, while retaining the features allowing for general application. Parameters which could be assumed constant for a large spectrum of situations were replaced with a small number of user parameters, for which assumed default values were found to provide a good fit to a range of cases.

A model of premixed turbulent combustion is modified for multi-dimensional computations of SI engines. This approach is based on the use of turbulent flame speed in order to suggest a closed balance equation for the mean combustion progress variable. The model includes a single unknown input parameter to be tuned. This model is tested against two sets of experimental data obtained by Bradley et al [17, 18 and 19] and Karpov and Severin [15] in fan-stirred bombs. The model quantitatively predicts the development of the turbulent flame speed, the effects of the initial pressure, temperature, and mixture composition on the turbulent flame speed, and the effects of r.m.s. turbulent velocity and burning mixture composition on the rate of the pressure rise. These results were computed with the same value of the aforementioned unknown input parameter of the model.

It has been known from multi-zone simulations that HCCI combustion can be significantly affected by temperature stratification of the in-cylinder gas. With the same combustion timing (i.e. crank angles at 50% heat release, denoted as CA50), large temperature stratification tends to prolong the combustion duration and lower down the in-cylinder pressure-rise-rate. With low pressure-rise-rate HCCI engines can be operated at high load, therefore it is of practical importance to look into more details about how temperature stratification affects the auto-ignition process. It has been realized that multi-zone simulations can not account for the effects of spatial structures of the stratified temperature field, i.e. how the size of the hot and cold spots in the temperature field could affect the auto-ignition process. This question is investigated in the present work by large eddy simulation (LES) method which is capable of resolving the in-cylinder turbulence field in space and time.

An expanding cylindrical laminar flame kernel affected by random external strain rates and diffusivity is numerically simulated in order to gain insight into the influence of small-scale turbulence on the combustion variability in engines. In the simulations, the kernel is strained, as a whole, by external velocity gradients randomly generated with either Gaussian or log-normal probability density functions. The influence of small-scale turbulent heat and mass transfer is modeled by turbulent diffusivity, the randomness of which is controlled by the fluctuations in the viscous dissipation averaged over the kernel volume. The computed results show that small-scale phenomena can substantially affect the quenching characteristics of a small flame kernel and the kernel growth history rj(t); the scatter of the computed curves of rf(t) being mainly controlled by the scatter of the duration of the initial stage of kernel development.

The ability to predict cyclic variations is certainly useful in studying engine operating regimes, especially under unstable operating conditions where one single cycle may differ from another substantially and a single simulation may give rather misleading results. PDF based models such as Stochastic Reactor Models (SRM) are able to model cyclic variations, but these may be overpredicted if discretization is too coarse. The range of cyclic variations and the dependence of the ability to correctly assess their mean values on the number of cycles simulated were investigated. In most cases, the average values were assessed correctly on the basis of as few as 10 cycles, but assessing the complete range of cyclic variations could require a greater number of cycles. In studying average values, variations due too coarse discretization being employed are smaller than variations originating from changes in physical parameters, such as heat transfer and mixing parameters.

In-cylinder flow measurements, conventional swirl measurements and CFD-simulations have been performed and then compared. The engine studied is a single cylinder version of a Scania D12 that represents a modern heavy-duty truck size engine. Bowditch type optical access and flat piston is used. The cylinder head was also measured in a steady-flow impulse torque swirl meter. From the two-dimensional flow-field, which was measured in the interval from -200° ATDC to 65° ATDC at two different positions from the cylinder head, calculations of the vorticity, turbulence and swirl were made. A maximum in swirl occurs at about 50° before TDC while the maximum vorticity and turbulence occurs somewhat later during the compression stroke. The swirl centre is also seen moving around and it does not coincide with the geometrical centre of the cylinder. The simulated flow-field shows similar behaviour as that seen in the measurements.

A validation study of the numerical model of n-heptane spray combustion based on experimental constant-volume data [1] was done, by comparing auto-ignition delays for different pre - turbulence levels and initial temperatures, flame contours, and soot distributions under Diesel-like conditions. The basic novelty of the methodology developed in [2] - [3] is the implementation of the partially stirred reactor (PaSR) model accounting for detailed chemistry / turbulence interactions. It is based on the assumption that the chemical processes proceed in two successive steps: micro mixing, simulated on a sub - grid scale, is followed by the reaction act. When the all Re number RNG k-ε or LES models are employed, the micro mixing time can be consistently defined giving the combustion model a “well-closed” form incorporated into the KIVA-3V code.

Simulations of a Direct Injection Spark Ignition (DISI) engine have been performed for both early injection with homogeneous charge combustion and for late injection with stratified charge combustion. The purpose has been to study flow characteristics, fuel/air mixing, combustion, and NOx and soot formation. Focus is put on the combustion modeling. Two different full load cases with early injection are simulated, 2000 rpm and 6000 rpm. One load point with late injection is simulated, 2000 rpm and 2.8 bar net MEP. Three different injection timings are simulated at the low load point: 77, 82, and 87 CAD bTDC. The spray simulations are tuned to match measured spray penetrations and droplet size distributions at both atmospheric and elevated pressure. Boundary conditions for the engine simulations are taken from 1-D gas exchange simulations that are tuned to match engine tests.

The pressure drop behavior of catalytic converters has been investigated for a number of different substrates, suitable for high performance IC-engines, regarding cell density, wall thickness and coating. The measurements have been performed on an experimental rig with room-air flow and hot-air flow. The data has been used to develop an empirical model for pressure drop in catalytic converters. The sources of pressure drop, such as viscous and inertial effects, have been separated in the model. The influence of turbulence on the pressure drop has been experimentally investigated. The model agrees well with experimental data and previous literature models and can be applied for 1D predictions as well as 3D CFD calculations.

Presented are results from numerical investigations of buoyancy driven flow in a simplified representation of an engine bay. A main motivation for this study is the necessity for a valid correlation of results from numerical methods and procedures with physical measurements in order to evaluate the accuracy and feasibility of the available numerical tools for prediction of natural convection. This analysis is based on previously performed PIV and temperature measurements in a controlled physical setup, which reproduced thermal soak conditions in the engine compartment as they occur for a vehicle parked in a quiescent ambient after sustaining high thermal loads. Thermal soak is an important phenomenon in the engine bay primarily driven by natural convection and radiation after there had been a high power demand on the engine. With the cooling fan turned off and in quiescent environment, buoyancy driven convection and radiation are the dominating modes of heat transfer.

This paper presents a large eddy simulation study of the liquid spray mixing with hot ambient gas in a constant volume vessel under engine-like conditions with the injection pressure of 1500 bar, ambient density 22.8 kg/m₃, ambient temperature of 900 K and an injector nozzle of 0.09 mm. The simulation results are compared with the experiments carried out by Pickett et al., under similar conditions. Under modern direct injection diesel engine conditions, it has been argued that the liquid core region is small and the droplets after atomization are fine so that the process of spray evaporation and mixing with the air is controlled by the heat and mass transfer between the ambient hot gas and central fuel flow. To examine this hypothesis a simple spray breakup model is tested in the present LES simulation. The simulations are performed using an open source compressible flow solver, in OpenFOAM.

Exterior turbulent flow is an important source of automobile cabin interior noise. The turbulent flow impacts the windows of the cabins to excite the structural vibration that emits the interior noise. Meanwhile, the exterior noise generated from the turbulent flow can also cause the window vibration and generate the interior noise. Side-view mirrors mounted upstream of the windows are one of the predominant body parts inducing the turbulent flow. In this paper, we investigate the interior noise caused by a generic side-view mirror. The interior noise propagates in a cuboid cavity with a rectangular glass window. The exterior flow and the exterior noise are computed using advanced CFD methods: compressible large eddy simulation, compressible detached eddy simulation (DES), incompressible DES, and incompressible DES coupled with an acoustic wave model. The last method is used to simulate the hydrodynamic and acoustic pressure separately.

In this paper, the turbulent flow induced by a production side-view mirror assembled on a full-scale production truck is simulated using a compressible k-ω SST detached eddy simulation (DES) approach -- the improved delayed DES (IDDES). The truck configuration consists of a compartment and a trailer. Due to the large size and geometric complexity of the configuration, some simplifications are applied to the simulation. A purpose of this work is to investigate whether the simplifications are suitable to obtain the reasonable properties of the flow near the side-view mirror. Another objective is to study the aerodynamic performances of the mirror. The configuration is simplified regarding two treatments. The first treatment is to retain the key exterior components of the truck body while removing the small gaps and structures. Furthermore, the trailer is shaped in an apex-truncated square pyramid.

Turbulent premixed natural gas - air flame velocities have been measured in a stationary axi-symmetric burner using LDA. The flame was stabilized by letting the flow retard toward a stagnation plate downstream of the burner exit. Turbulence was generated by letting the flow pass through a plate with drilled holes. Three different hole diameters were used, 3, 6 and 10 mm, in order to achieve different turbulent length scales. Turbulent integral length scales were measured using two-point LDA and the stretching in terms of the Karlovitz number could be estimated from these measurements. The results support previous studies indicating that stretching reduces the flame speed.

The most economical way to convert truck and bus DI-diesel engines to natural gas operation is to replace the injector with a spark plug and modify the combustion chamber in the piston crown for spark ignition operation. The modification of the piston crown should give a geometry well suited for spark ignition operation with the original swirling inlet port. Ten different geometries were tried on a converted VOLVO TD102 engine and a remarkably large difference in the rate of combustion was noted between the chambers. To find an explanation for this difference a cycle resolved measurement of the in-cylinder mean velocity and turbulence was performed with Laser Doppler Velocimetry (LDV). The results show a high correlation between in cylinder turbulence and rate of heat release in the main part of combustion.

To study the effect of different injection timings on the charge inhomogeneity, planar laser-induced fluorescence (PLIF) was applied to an operating engine. Quantitative images of the fuel distribution within the engine were obtained. Since the fuel used, iso-octane, does not fluoresce, a dopant was required. Three-pentanone was found to have vapour pressure characteristics similar to those of iso-octane as well as low absorption and suitable spectral properties. A worst case estimation of the total accuracy from the PLIF images gives a maximum error of 0.03 in equivalence ratio. The results show that an early injection timing gives a higher degree of charge inhomogeneity close to the spark plug. It is also shown that charge inhomogeneity gives a more unstable engine operation. A correlation was noted between the combustion on a cycle to cycle basis and the average fuel concentration within a circular area close to the spark plug center.

This work presents a numerical method for predictions of HC oxidation in the cold turbulent wall jet emerging from the piston top land crevice in an S.I. engine, using a complex chemical reaction model. The method has been applied to an engine model geometry with the aim to predict the HC oxidation rate under engine - relevant conditions. According to the simulation a large amount of HC survives oxidation due to the long ignition delay of the wall jet emitted from the crevice. This ignition delay, in turn depends mainly on chemical composition and temperature of the gas mixture in the crevice and also on the temperature distribution in the cylinder boundary layer.

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.

The different roles played by small and large eddies in engine combustion were studied. Experiments compared natural gas combustion in a converted, single cylinder Volvo TD 102 engine and in a 125 mm cubical cell. Turbulence is used to enhance flame growth, ideally giving better efficiency and reduced cyclic variation. Both engine and test cell results showed that flame growth rate correlated best with the level of high frequency, small eddy turbulence. The more effective, small eddy turbulence also tended to lower cyclic variations. Large scales and bulk flows convected the flame relative to cool surfaces and were most important to the initial flame kernel.

This study is focused on the effect of different valve strategies on the in-cylinder flow and combustion A conventional four-valve pentroof engine was modified to enable optical access to the combustion chamber To get information on the flow, a two-component LDV system was applied The combustion was monitored by the use of cylinder pressure in a one-zone heat release model The results show that the flow in the cylinder with the valves operating in the standard configuration has an expected tumble characteristic In this case the high frequency turbulence is homogeneous and has a peak approximately 20 CAD BTDC With one valve deactivated, the flow shows a swirling pattern The turbulence is then less homogeneous but the level of turbulence is increased When the single inlet valve was phased late against the crankshaft dramatic effects on the flow resulted The late inlet valve opening introduced a low cylinder pressure before the valve opened The high pressure difference across the valve introduced a high-velocity jet into the cylinder Turbulence was increased by a factor of two by this operational mode When two inlet valves were used, a reduction of turbulence resulted from a very late inlet cam phase