Search Results

To meet stringent emissions regulations, filtration devices are often used in engine exhaust systems to reduce particulate mass (PM) and particulate number (PN). Diesel particulate filters (DPFs) are a well-established means of reducing PM from diesel engines to meet emissions regulations. New emissions regulations will most likely require a similar technology on gasoline engines with direct injection, gasoline particulate filters (GPFs). Due to differences in the exhaust and particulate characteristics, the design and operation of GPFs and DPFs differ. In a DPF filtration is dominated by the buildup of a soot cake. Whereas in a GPF, much of the soot is trapped inside the porous substrate, or filter wall, where deep-bed filtration is dominant. Thus, an accurate model describing the porous filtration properties of GPF substrates is desired. The pore filtration model (PFM) was developed to more accurately model the deep-bed filtration process that occurs in a GPF.

Using non-viscosity dynamic structure Large Eddy Simulations (LES) turbulence model, spray=induced turbulence is investigated on a number of different Computational Fluid Dynamics (CFD) grids of varying mesh sizes (from 0.5 to 2 mm mesh). Turbulent flow is induced inside a quiescent chamber by liquid fuel spray and then left to decay after end of injection by virtue of its molecular viscosity and turbulent dissipation. Coherent structures (CS) of this turbulent flow are constructed and visualized using λ2 definition. Using CS, analysis is performed on the turbulent flow around the liquid spray jet. These CS from LES are then compared against the results from RANS calculations as well. The visualization of CS helps to explain the mechanism of fuel-air mixing obtained from LES results and its difference with RANS calculations.

Numerical investigation is carried out to explore various strategies of combustion mode switching in a diesel engine operating at high power. Numerical results are compared with high power single cylinder (CAT 3401E) experiments for combustion phasing and emission characteristics. In this study CFD calculations are carried out using the KIVA CFD code with Large Eddy Simulation turbulence model and Direct Chemistry Solver sub-models. The advanced turbulence and combustion sub-models enabled more realistic visualization of the effects of single-cycle mode switching on in-cylinder flow structures, fuel-air mixing behavior and combustion phasing. Two circumstances of mode switch are presented in this study. Mode switches are performed from traditional High Temperature Combustion to early injection PCCI combustion and vice versa. In this investigation several aspects of combustion control are investigated.

A novel 2-zone combustion chamber concept (patent pending) was developed using multi-dimensional modeling. At minimum volume, an axial projection in the piston divides the volume into distinct zones joined by a communication channel. The projection provides a means to control the mixture formation and combustion phasing within each zone. The novel combustion system was applied to reactivity controlled compression ignition (RCCI) combustion in both light-duty and heavy-duty diesel engines. Results from the study of an 8.8 bar BMEP, 2600 RPM operating condition are presented for the light-duty engine. The results from the heavy-duty engine are at an 18.1 bar BMEP, 1200 RPM operating condition. The effect of several major design features were investigated including the volume split between the inner and outer combustion chamber volumes, the clearance (squish) height, and the top ring land (crevice) volume.

A swept-volume method of calculating the volume swept by the flame during each time step is developed and used to improve the calculation of fuel reaction rates. The improved reaction rates have been applied to the ignition model and coupled with the level set G-equation combustion model. In the ignition model, a single initial kernel is formed after which the kernel is convected by the gas flow and its growth rate is determined by the flame speed and thermal expansion due to the energy transfer from the electrical circuit. The predicted ignition kernel size was compared with the available experimental data and good agreements were achieved. Once the ignition kernel reaches a size when the fully turbulent flame is developed, the G-equation model is switched on to track the mean turbulent flame front propagation.

The SCR Filter simultaneously reduces NOx and Particle Matter (PM) in the exhaust and is considered an effective way to meet emission regulations. By combining the function of a Diesel Particulate Filtration (DPF) and a Selective Catalytic Reduction (SCR), the SCR Filter reduces the complexity and cost of aftertreatment systems in diesel vehicles. Moreover, it provides an effective reaction surface and potentially reduces backpressure by combining two devices into one. However, unlike traditional flow through type SCR, the deNOx reactions in the SCR Filter can be affected by the particulate filtration and regeneration process. Additionally, soot oxidation can be affected by the deNOx process. A 1-D kinetic model for integrated DPF and NH3-SCR system over Cu-zeolite catalysts was developed and validated with experimental data in previous work[1].

A modified version of the three dimensional CFD code KIVA-3, which accommodates moving valves and a moving piston crown, has been applied to a heavy-duty, four cycle, dual intake valve, direct injection, diesel engine The fluid domain encompasses an intake runner, two valved ports and a cylinder with a Mexican hat bowl-in-piston configuration In the first part of the study, the modified KIVA-3 code was used to simulate the flow through the port and cylinder of the engine without a piston The intake valves were cycled (300 rpm at the camshaft) Two turbulence models were compared in this part of the study, standard k - ε and the RNG modified k - ε as discussed in [11] The results were compared to particle image velocimetry (PIV) images Large scale flow features of the computer simulations agreed moderately well with the ensemble averaged flow bench results There was very little difference between the results from the two turbulence models Motored engine simulations including a piston were conducted to characterize the in-cylinder gas flow and mixing during the intake and compression strokes Characterization methods were developed which yield insight into the in-cylinder gas motion and mixing during both strokes Intake and residual gases are tracked separately, both large scale convection and turbulent mixing are investigated, flow critical points are examined to provide information about flow topology and turbulence production is correlated with the evolution of flow structures Results show that mixing of the intake and residual gases is very non-uniform Many complex flow structures develop during intake and are destroyed during compression However, several structures survive through compression and contribute to enhanced mixing near top dead center These significant structures have been identified and tracked back to intake The flow field near top dead center exhibits spatial inhomogeneities in temperature and small scale mixing parameters such as turbulence kinetic energy and its dissipation rate

Large-eddy simulation (LES) is a useful approach for the simulation of turbulent flow and combustion processes in internal combustion engines. This study employs the ANSYS Forte CFD package and explores several key and fundamental components of LES, namely, the subgrid-scale (SGS) turbulence models, the numerical schemes used to discretize the transport equations, and the computational mesh. The SGS turbulence models considered include the classic Smagorinsky model and a dynamic structure model. Two numerical schemes for momentum convection, quasi-second-order upwind (QSOU) and central difference (CD), were evaluated. The effects of different computational mesh sizes controlled by both fixed mesh refinement and a solution-adaptive mesh-refinement approach were studied and compared. The LES models are evaluated and validated against several flow configurations that are critical to engine flows, in particular, to fuel injection processes.

Surface film formation is an important phenomenon during spray impingement in a combustion chamber. The film that forms on the chamber walls and piston bowl produces soot post-combustion. While some droplets stick to the wall surface, others splash and interact with the gas present inside the combustion chamber. Accurate prediction of both the film thickness and splashed mass is crucial for surface film model development since it leads to a precise estimation of the amount of soot and other exhaust gases formed. This information could guide future studies aimed at a comprehensive understanding of the combustion process and might enable development of engines with reduced emissions. Dynamic structure Large Eddy Simulation (LES) turbulence model implemented for in-cylinder sprays [1] has shown to predict the flow structure of a spray more accurately than the Reynolds-averaged Navier-Stokes turbulence model.

Diesel particulate samples were collected from a light duty engine operated at a single speed-load point with a range of biodiesel and conventional fuel blends. The oxidation reactivity of the samples was characterized in a laboratory reactor, and BET surface area measurements were made at several points during oxidation of the fixed carbon component of both types of particulate. The fixed carbon component of biodiesel particulate has a significantly higher surface area for the initial stages of oxidation, but the surface areas for the two particulates become similar as fixed carbon oxidation proceeds beyond 40%. When fixed carbon oxidation rates are normalized to total surface area, it is possible to describe the oxidation rates of the fixed carbon portion of both types of particulates with a single set of Arrhenius parameters. The measured surface area evolution during particle oxidation was found to be inconsistent with shrinking sphere oxidation.

While Low Temperature Combustion (LTC) strategies such as Reactivity Controlled Compression Ignition (RCCI) exhibit high thermal efficiency and produce low NOx and soot emissions, low load operation is still a significant challenge due to high unburnt hydrocarbon (UHC) and carbon monoxide (CO) emissions, which occur as a result of poor combustion efficiencies at these operating points. Furthermore, the exhaust gas temperatures are insufficient to light-off the Diesel Oxidation Catalyst (DOC), thereby resulting in poor UHC and CO conversion efficiencies by the aftertreatment system. To achieve exhaust gas temperature values sufficient for DOC light-off, combustion can be appropriately phased by changing the ratio of gasoline to diesel in the cylinder, or by burning additional fuel injected during the expansion stroke through post-injection.

Homogeneous low temperature combustion is believed to be a promising approach to resolve the conflict of goals between high efficiency and low exhaust emissions. Disadvantageously for this kind of combustion, the whole process depends on chemical kinetics and thus is hard to control. Reactivity controlled combustion can help to overcome this difficulty. In the so-called RCCI (reactivity controlled compression ignition) combustion concept a small amount of pilot diesel that is injected directly into the combustion chamber ignites a highly diluted gasoline-air mixture. As the gasoline does not ignite without the diesel, the pilot injection timing and the ratio between diesel and gasoline can be used to control the combustion process. A phenomenological multi-zone model to predict RCCI combustion has been developed and validated against experimental and 3D-CFD data. The model captures the main physics governing ignition and combustion.

Modeling ignition kernel development in spark ignition engines is crucial to capturing the sources of cyclic variability, both with RANS and LES simulations. Appropriate kernel modeling must ensure that energy transfer from the electrodes to the gas phase has the correct timing, rate and locations, until the flame surface is large enough to be represented on the mesh by the G-Equation level-set method. However, in most kernel models, geometric details driving kernel growth are missing: either because it is described as Lagrangian particles, or because its development is simplified, i.e., down to multiple spherical flames. This paper covers the geometric aspects of kernel development, which makes up the core of a Triangulated Lagrangian Ignition Kernel model. One (or multiple, if it restrikes) spark channel is initialized as a one-dimensional Lagrangian particle thread.