A 2-D gas-particle two-phase flow model has been developed to study the flow characteristics in a single channel of a honeycomb ceramic diesel particulate filter. A particle source in cell (PSIC) algorithm is used to calculate the gas-particle two-phase flow. Firstly, the gas-phase flow field alone (without taking into account of the particle-phase) is solved for estimation of gas velocity and pressure fields in the Euler coordinate. Secondly, the particle-phase is added in and particles tracked down in the Lagrange coordinate. Thirdly, the particle source which acts on the gas-phase cell is calculated and added to the gas-phase equations. Fourthly, the gas-phase equations with the particle source are solved again. Lastly, the above process is iterated until the flow field is convergent. Taking the above-mentioned approach, the gas-particle two-phase flow characteristic is simulated using FLUENT. The simulation results are in good agreement with experiment data. After the models being calibrated, the influences on particle movement traces of the exhaust gas entry velocity, particulate matter concentration and porous wall thickness are studied and presented. The work conducted provides an important reference to further studies on gas-particle two-phase flow and combustion characteristics during the cake filtration process and the regeneration process.

A multidimensional numerical simulation method was developed to analyze air/fuel premixing, stratified combustion and NOx emission formation in a gasoline direct injection (GDI) engine. Firstly, many sub models were integrated into one Computational Fluid Dynamics (CFD) code: ICFD-CN, such as Sarre nozzle flow, Kelvin–Helmholtz (KH) dynamic jet model, Taylor-Analogy Breakup (TAB) model, Rayleigh–Taylor (RT) droplet breakup model, Lefebvre fuel vaporization model, Liu droplet drag & distortion model, Gosman turbulence & droplet dispersion model, O’rourke wall film model, O’rourke and Bracco droplet impinging & coalescence model, Stanton spray/wall impinging model, the Discrete Particle Ignition Kernel(DPIK)ignition model, the single step combustion and the patulous Zeldovich model for NOx generation mechanism. The integrated CFD code was then calibrated against experimental data in a gasoline direct injection engine for several engine operating conditions. Afterwards it was applied to investigate the influences on air/fuel premixing, stratified combustion as well as NOx emission formation of various combustion chamber designs Simulation results indicate that the distribution of air/fuel mixture becomes more and more uniform before ignition occurs, because of the elevated air/fuel mixture velocity and enhanced swirl level, in the range of combustion chamber bowl offsets changing from 0.0 cm to 1.7 cm with a step of about 0.25 cm. This, in turn, leads to shorter ignition delay, slower combustion, lower in-cylinder peak temperature and stronger heat transfer between the burnt and un-burnt zones. It is recommend that the best combustion chamber bowl offset should be from 0.5 cm to 1.0 cm, based on the compromise between NOx emission and un-burnt fuel (evaporated and liquid gasoline) mass. The 3D simulation model developed in this study is capable to provide detailed insights of all stages of the in-cylinder processes, including air motion, air/fuel mixing, combustion and emission formation. It is proven that the simulation approach and results very helpful in providing design guidelines and optimizing for the combustion chamber shape design.

This paper describes an improved mathematical model to study the emission conversion effectiveness of a three-way catalytic converter, which employed detailed chemical reaction mechanism. The model also accounts for adsorption/release of oxygen in the catalyst monolith under non-stoichiometric A/F conditions. A commercial CFD code FLUENT was utilized to solve the governing equations for flow and pressure drop and to simulate the transient process in a three-way catalytic converter in a multi-dimensional manner. A comparison between simulation results and experimental data for a three-way catalyst was conducted and a good agreement was observed. Based on the improved model, some geometric parameters were studied for an elliptic monolith catalyst, which are widely used in today’s converter systems because of its advantages in packaging. Simulation results show that, in general, decreasing the ellipse ratio, i.e. the ratio of the major over the minor axis length, has favorable effect on catalyst conversion efficiency. However the significance of improvement depends on the ellipse ratio itself: if the ellipse ratio is large, the effect is strong when reducing it; as the ratio is approaching to 1, the improvement becomes less and less significant. On the other hand, increasing the converter length also improves the catalyst conversion performance, so does the increase in the cross-sectional area. For a fixed converter volume: the performance of a short catalyst with large cross-sectional area is generally not as good as the one with longer length but smaller cross-section area. For a fixed void fraction of the catalyst, a catalyst with higher cell density but thinner wall thickness, because of its increased surface area, is generally more effective than the one with lower cell density but thicker wall thickness. The improved CFD model and simulation results are proven capable of providing guidelines for practical design improvements to meet the increasingly stringent emission legislations.