Search

Real gas effects on combustion and emissions in internal combustion engines are investigated using three-dimensional computational fluid dynamics. The Peng–Robinson equation of state is implemented to describe pressure–volume–temperature relationships and to calculate thermodynamic properties and relevant partial derivatives. Four facilities are modeled, including non-reacting compression in a motoring engine, combustion in a conventional diesel combustion engine and in a reactivity controlled compression ignition engine, as well as for a non-reacting reflected wave in a shock tube. It is found that the real gas effects of gas mixtures in practical internal combustion engine operation are sensitive to the operating load and the amount of premixed fuel. Excellent agreement against experiments was found for engine simulations with the Peng–Robinson equation of state in terms of cylinder pressure and apparent heat release rate. However, discrepancies with predictions from the ideal gas law grow with increased load and larger amounts of premixed fuel. In particular, the predicted emissions of soot, NOx, CO and unburnt hydrocarbons show increasing sensitivity to real gas effects as a result of changes in combustion phasing. Fuel condensation is also modeled using a vapor–liquid phase equilibrium solver and significant dependency on the equation of state used is found. Therefore, it is recommended to include real gas effects in internal combustion engine modeling to capture combustion and emissions characteristics accurately. Additionally, the results emphasize the role of real gas effects on reaction rates. Shock tube simulations are used to demonstrate the importance of using the real gas equation of state in the interpretation of chemical kinetic measurements. Significantly different compressed gas temperatures behind the reflected shock are predicted when real gas effects are considered. This needs to be realized when developing chemical kinetic models and rate constants for engine applications from shock tube data.

A methodology for modeling gaseous injection has been refined and applied to recent experimental data from the literature. This approach uses a discrete phase analogy to handle gaseous injection, allowing for addition of gaseous injection to a CFD grid without needing to resolve the injector nozzle. This paper focuses on model testing to provide the basis for simulation of hydrogen direct injected internal combustion engines. The model has been updated to be more applicable to full engine simulations, and shows good agreement with experiments for jet penetration and time-dependent axial mass fraction, while available – radial mass fraction data is less well predicted.

There is a need to reduce the computational expense of practical multidimensional combustion simulations. Simulation of Homogeneous Charge Compression Ignition (HCCI) engine processes requires consideration of detailed chemistry in order to capture the ignition and combustion characteristics. Even with relatively coarse numerical meshes and reduced chemistry mechanisms, calculation times are still unacceptably long. For the simulation of Direct Injection (DI) engines, fine meshes are needed to achieve the resolution required by the spray and mixing models, and they are computationally expensive even with reduced chemistry. In addition, the increasing application of CFD for engine design optimization is pushing the demand to reduce computational time. In current design optimizations, depending on the size of the parametric space, hundreds of individual simulations are needed. This work presents an efficient Adaptive Multi-grid Chemistry (AMC) model that can be used in engine CFD codes for simulations of HC...

Modelling the premixed charge compression ignition (PCCI) engine requires a balanced approach that captures both fluid motion as well as low- and high-temperature fuel oxidation. A fully integrated computational fluid dynamics (CFD) and chemistry scheme (i.e. detailed chemical kinetics solved in every cell of the CFD grid) would be the ideal PCCI modelling approach, but is computationally very expensive. As a result, modelling assumptions are required in order to develop tools that are computationally efficient, yet maintain an acceptable degree of accuracy. Multi-zone models have been previously shown accurately to capture geometry-dependent processes in homogeneous charge compression ignition (HCCI) engines. In the presented work, KIVA-3V is fully coupled with a multi-zone model with detailed chemical kinetics. Computational efficiency is achieved by utilizing a low-resolution discretization to solve detailed chemical kinetics in the multi-zone model compared with a relatively high-resolution CFD solution. The multi-zone model communicates with KIVA-3V at each computational timestep, as in the ideal fully integrated case. The composition of the cells, however, is mapped back and forth between KTVA-3V and the multi-zone model, introducing significant computational time savings. The methodology uses a novel re-mapping technique that can account for both temperature and composition non-uniformities in the cylinder. Validation cases were developed by solving the detailed chemistry in every cell of a KIVA-3V grid. The new methodology shows very good agreement with the detailed solutions in terms of ignition timing, burn duration, and emissions.

Intake flow for a dual intake valve diesel engine is modeled using moving valves and realistic geometries. The objectives are to show the importance of intake calculations instead of using assumed initial conditions, to supply spray and combustion models with more accurate initial conditions, and to demonstrate the use of multidimensional modeling to studying intake processes. Global simulation parameters are compared with experimental results and show good agreement. The intake process shows a 30% difference in mass flows and average swirl in opposite directions across the two intake valves. The effect of the intake process on the flowfield at the end of compression is examined. Modeling the intake flow results in swirl and turbulence characteristics that are quite different from those obtained by conventional methods in which compression stroke initial conditions are assumed.

A computational mesh is required when performing CFD-combustion modeling of internal combustion engines. For combustion chambers with moving pistons and valves, like those in typical cars and trucks, the combustion chamber shape changes continually in response to piston and valve motion. The combustion chamber mesh must then also change at each time step to reflect that change in geometry. The method of changing the mesh from one computational time step to the next is called rezoning. This paper introduces a new method of mesh rezoning for the KIVA3V CFD-combustion program. The standard KIVA3V code from Los Alamos National Laboratory comes with standard rezoners that very nicely handle mesh motion for combustion chambers whose mesh does not include valves and for those with flat heads employing vertical valves. For pent-roof and wedge-roof designs KIVA3V offers three rezoners to choose from, the choice depending on how similar a combustion chamber is to the sample combustion chambers that come with KIVA3V. Often, the rezoners must be modified for meshes of new combustion chamber geometries to allow the mesh to successfully capture change in geometry during the full engine cycle without errors. There is no formal way to approach these modifications; typically this requires a long trial and error process to get a mesh to work for a full engine cycle. The benefit of the new rezoner is that it replaces the three existing rezoners for canted valve configurations with a single rezoner and has much greater stability, so the need for ad hoc modifications of the rezoner is greatly reduced. This paper explains how the new rezoner works and gives examples of its use.Copyright © 2010 by ASME

Multi-zone CFD simulations with detailed kinetics were used to model iso-octane HCCI experiments performed on a single-cylinder research engine. The modeling goals were to validate the method (multi-zone combustion modeling) and the reaction mechanism (LLNL 857 species iso-octane) by comparing model results to detailed exhaust speciation data, which was obtained with gas chromatography. The model is compared to experiments run at 1200 RPM and 1.35 bar boost pressure over an equivalence ratio range from 0.08 to 0.28. Fuel was introduced far upstream to ensure fuel and air homogeneity prior to entering the 13.8:1 compression ratio, shallow-bowl combustion chamber of this 4-stroke engine. The CFD grid incorporated a very detailed representation of the crevices, including the top-land ring crevice and headgasket crevice. The ring crevice is resolved all the way into the ring pocket volume. The detailed grid was required to capture regions where emission species are formed and retained. Results show that combustion is well characterized, as demonstrated by good agreement between calculated and measured pressure traces. In addition, excellent quantitative agreement between the model and experiment is achieved for specific exhaust species components, such as unburned fuel, formaldehyde, and many other intermediate hydrocarbon species. Some calculated trace intermediate hydrocarbon species do not agree as well with measurements, highlighting areas needing further investigation for understanding fundamental chemistry processes in HCCI engines.

The growing interest in gaseous fuels (hydrogen and natural gas) for internal combustion engines calls for the development of computer models for simulation of gaseous fuel injection, air entrainment and the ensuing combustion. This paper introduces a new method for modeling the injection and air entrainment processes for gaseous fuels. The model uses a gaseous sphere injection methodology, similar to liquid droplet in injection techniques used for liquid fuel injection. In this paper, the model concept is introduced and model results are compared with correctly- and under-expanded experimental data.

This paper illustrates the applicability of a sequential fluid mechanics, multi-zone chemical kinetics model to analyze HCCI experimental data for two combustion chamber geometries with different levels of turbulence: a low turbulence disc geometry (flat top piston), and a high turbulence square geometry (piston with a square bowl). The model uses a fluid mechanics code to determine temperature histories in the engine as a function of crank angle. These temperature histories are then fed into a chemical kinetic solver, which determines combustion characteristics for a relatively small number of zones (40). The model makes the assumption that there is no direct linking between turbulence and combustion. The results show that the multi-zone model yields good results for both the disc and the square geometries. The model makes good predictions of pressure traces and heat release rates. The experimental results indicate that the high turbulence square geometry has longer burn duration than the low turbulence disc geometry. This difference can be explained by the sequential multi-zone model, which indicates that the cylinder with the square bowl has a thicker boundary layer that results in a broader temperature distribution. This broader temperature distribution tends to lengthen the combustion, as cold mass withinmore » the cylinder takes longer to reach ignition temperature when compressed by the expansion of the first burned gases. The multi-zone model, which makes the basic assumption that HCCI combustion is controlled by chemical kinetics, is therefore capable of explaining the experimental results obtained for different levels of turbulence, without considering a direct interaction between turbulence and combustion. A direct connection between turbulence and HCCI combustion may still exists, but it seems to play a relatively minor role in determining burn duration at the conditions analyzed in this paper.« less

A numerical simulation and optimization study was conducted for a medium speed direct injection diesel engine. The engine's operating characteristics were first matched to available experimental data to test the validity of the numerical model. The KIVA-3V ERC CFD code was then modified to allow independent spray events from two rows of nozzle holes. The angular alignment, nozzle hole size, and injection pressure of each set of nozzle holes were optimized using a micro-genetic algorithm. The design fitness criteria were based on a multi-variable merit function with inputs of emissions of soot, NO x , unburned hydrocarbons, and fuel consumption targets. Penalties to the merit function value were used to limit the maximum in-cylinder pressure and the burned gas temperature at exhaust valve opening. The optimization produced a 28.4% decrease in NO x and a 40% decrease in soot from the baseline case, while giving a 3.1% improvement in fuel economy. The improvements were found to be due to the formation of circulatory flows caused by the interaction of adjacent optimally timed injections. The resulting spinning flow field greatly enhances mixing and combustion rates, and is called "Spin-Spray Combustion*". The resulting fast mixing allowed the use of retarded injection timings, thereby lowering NO x production without increasing soot beyond the target values.

We have conducted a detailed numerical analysis of HCCI engine operation at low loads to investigate the sources of HC and CO emissions and the associated combustion inefficiencies. Engine performance and emissions are evaluated as fueling is reduced from typical HCCI conditions, with an equivalence ratio f = 0.26 to very low loads (f = 0.04). Calculations are conducted using a segregated multi-zone methodology and a detailed chemical kinetic mechanism for iso-octane with 859 chemical species. The computational results agree very well with recent experimental results. Pressure traces, heat release rates, burn duration, combustion efficiency and emissions of hydrocarbon, oxygenated hydrocarbon, and carbon monoxide are generally well predicted for the whole range of equivalence ratios. The computational model also shows where the pollutants originate within the combustion chamber, thereby explaining the changes in the HC and CO emissions as a function of equivalence ratio. The results of this paper contribute to the understanding of the high emission behavior of HCCI engines at low equivalence ratios and are important for characterizing this previously little explored, yet important range of operation.

Recent advances in computer CPU speed and engine model development make the use of multidimensional modeling increasingly attractive for engine designers and researchers seeking methods to reduce pollutant emissions, without sacrificing engine performance. Much progress has been made in engine submodel development in recent years, and modeling has been used to help explain aspects of engine performance and pollutant emissions trends. Factors including injection timing, injection pressure, injection rate-shape, combustion chamber design, boost pressure and EGR can now be explored computationally. A new development is the use of detailed models to help optimize engine design. With such a large number of engine parameters to investigate, it has been necessary to introduce efficient computational methodologies for the analysis of engine design configurations. Genetic-Algorithm-based (GA) search techniques have been demonstrated to be able to successfully follow a family of designs through a large search space to an optimum design. The GA method is applied together with multidimensional modeling in the present study to specify optimum injection rate-shape profiles for multiple injections in a heavy-duty diesel engine over a wide range of speeds and loads. In addition, related optimization studies are reviewed, including optimization of multiple injections with variations in EGR and boost pressure, optimization of intake flow swirl and tumble parameters for variable valve actuation applications, and optimization of the piston bowl geometry for a HSDI diesel engine.

This research investigates how the handling of mixing and heat transfer in a multi-zone kinetic solver affects the prediction of carbon monoxide and hydrocarbon emissions for simulations of HCCI engine combustion. A detailed kinetics multi-zone model is now more closely coordinated with the KIVA3V computational fluid dynamics code for simulation of the compression and expansion processes. The fluid mechanics is solved with high spatial and temporal resolution (40,000 cells). The chemistry is simulated with high temporal resolution, but low spatial resolution (20 computational zones). This paper presents comparison of simulation results using this enhanced multi-zone model to experimental data from an isooctane HCCI engine. The chemical kinetics part of the simulation is handled using the multi-zone segregated solver method developed previously, but now KIVA3V is used to handle the fluid dynamics (convection, mass diffusion and heat transfer) for the entire compression and expansion processes. The results show that carbon monoxide and hydrocarbon emissions may be greatly influenced by the mixing and heat transfer during expansion. The prediction of HC and CO is significantly improved by inclusion of these effects in the simulation.

A multi-zone model has been developed that accurately predicts HCCI combustion and emissions. The multizone methodology is based on the observation that turbulence does not play a direct role on HCCI combustion. Instead, chemical kinetics dominates the process, with hotter zones reacting first, and then colder zones reacting in rapid succession. Here, the multi-zone model has been applied to analyze the effect of piston crevice geometry on HCCI combustion and emissions. Three different pistons of varying crevice size were analyzed. Crevice sizes were 0.26, 1.3 and 2.1 mm, while a constant compression ratio was maintained (17:1). The results show that the multi-zone model can predict pressure traces and heat release rates with good accuracy. Combustion efficiency is also predicted with good accuracy for all cases, with a maximum difference of 5% between experimental and numerical results. Carbon monoxide emissions are underpredicted, but the results are better than those obtained in previous publications. The improvement is attributed to the use of a 40-zone model, while previous publications used a 10-zone model. Hydrocarbon emissions are well predicted. For cylinders with wide crevices (1.3 and 2.1 mm), HC emissions do not decrease monotonically as the relative air/fuel ratio ({lambda}) increases. Instead, maximum HC more » emissions are obtained for an intermediate value of {lambda}. The model predicts this relative air/fuel ratio for maximum HC emissions with very good accuracy. The results show that the multi-zone model can successfully predict the effect of crevice geometry on HCCI combustion, and therefore it has applicability to the design of HCCI engines with optimum characteristics for high efficiency, low emissions and low peak cylinder pressure. « less

We have developed a methodology for predicting combustion and emissions in a Homogeneous Charge Compression Ignition (HCCI) Engine. This methodology combines a detailed fluid mechanics code with a detailed chemical kinetics code. Instead of directly linking the two codes, which would require an extremely long computational time, the methodology consists of first running the fluid mechanics code to obtain temperature profiles as a function of time. These temperature profiles are then used as input to a multi-zone chemical kinetics code. The advantage of this procedure is that a small number of zones (10) is enough to obtain accurate results. This procedure achieves the benefits of linking the fluid mechanics and the chemical kinetics codes with a great reduction in the computational effort, to a level that can be handled with current computers. The success of this procedure is in large part a consequence of the fact that for much of the compression stroke the chemistry is inactive and thus has little influence on fluid mechanics and heat transfer. Then, when chemistry is active, combustion is rather sudden, leaving little time for interaction between chemistry and fluid mixing and heat transfer. This sequential methodology has been capable of explaining the mainmore » characteristics of HCCI combustion that have been observed in experiments. In this paper, we use our model to explore an HCCI engine running on propane. The paper compares experimental and numerical pressure traces, heat release rates, and hydrocarbon and carbon monoxide emissions. The results show an excellent agreement, even in parameters that are difficult to predict, such as chemical heat release rates. Carbon monoxide emissions are reasonably well predicted, even though it is intrinsically difficult to make good predictions of CO emissions in HCCI engines. The paper includes a sensitivity study on the effect of the heat transfer correlation on the results of the analysis. Importantly, the paper also shows a numerical study on how parameters such as swirl rate, crevices and ceramic walls could help in reducing HC and CO emissions from HCCI engines.« less

Intake flow for a dual intake valved diesel engine is modeled using moving valves and realistic geometries. The objectives are to obtain accurate initial conditions for combustion calculations and to provide a tool for studying intake processes. Global simulation parameters are compared with experimental results and show good agreement. The intake process shows a 30 percent difference in mass flows and average swirl in opposite directions across the two intake valves. The effect of the intake process on the flow field at the end of compression is examined. Modeling the intake flow results in swirl and turbulence characteristics that are quite different from those obtained by conventional methods in which compression stroke initial conditions are assumed.