Search Results

A simulation method is presented for the analysis of combustion in spark ignition (SI) engines operated at elevated exhaust gas recirculation (EGR) level and employing multiple spark plug technology. The modeling is based on a zero-dimensional (0D) stochastic reactor model for SI engines (SI-SRM). The model is built on a probability density function (PDF) approach for turbulent reactive flows that enables for detailed chemistry consideration. Calculations were carried out for one, two, and three spark plugs. Capability of the SI-SRM to simulate engines with multiple spark plug (multiple ignitions) systems has been verified by comparison to the results from a three-dimensional (3D) computational fluid dynamics (CFD) model. Numerical simulations were carried for part load operating points with 12.5%, 20%, and 25% of EGR. At high load, the engine was operated at knock limit with 0%, and 20% of EGR and different inlet valve closure timing.

In this work are presented experimental and simulated data from a one-cylinder direct injected Diesel engine fuelled with Diesel, two different biodiesel blends and pure biodiesel at one engine operating point. The modeling approach focuses on testing and rating biodiesel surrogate fuel blends by means of combustion and emission behavior. Detailed kinetic mechanisms are adopted to evaluate the fuel-blends performances under both reactor and diesel engine conditions. In the first part of the paper, the experimental engine setup is presented. Thereafter the choice of the surrogate fuel blends, consisting of n-decane, α-methyl-naphtalene and methyl-decanoate, are verified by the help of experiments from the literature. The direct injection stochastic reactor model (DI-SRM) is employed to simulate combustion and engine exhaust emissions (NOx, HC, CO and CO2), which are compared to the experimental data.

The relatively new combustion concept known as partially premixed combustion (PPC) has high efficiency and low emissions. However, there are still challenges when it comes to fully understanding and implementing PPC. Thus a predictive combustion tool was used to gain further insight into the combustion process in late cycle mixing. The modeling tool is a stochastic reactor model (SRM) based on probability density functions (PDF). The model requires less computational time than a similar study using computational fluid dynamics (CFD). A novel approach with a two-zone SRM was used to capture the behavior of the partially premixed or stratified zones prior to ignition. This study focuses on PPC mixing conditions and the use of an efficient analysis approach.

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 self-calibrating model for Diesel engine simulations is presented. The overall model consists of a zero-dimensional direct injection stochastic reactor model (DI-SRM) for engine in-cylinder processes simulations and a package of optimization algorithms (OPAL) suitable for solving various optimization, automatization and search problems. In the DI-SRM, based on an extensive model parameters study, the mixing time history that affects the level of in-cylinder turbulence was selected as a main calibration parameter. As targets during calibration against the experimental data, in-cylinder pressure history and engine-out emissions, including nitrogen oxides and unburned hydrocarbons were chosen. The calibration task was solved using DI-SRM and OPAL working as an integrated tool. Within OPAL, genetic algorithms (GA) were used to determine model constants necessary for calibrating. Engine-out emissions in DI-SRM were calculated based on the reduced mechanism of n-heptane.

This paper reports on a fast predictive combustion tool employing detailed chemistry. The model is a stochastic reactor based, discretised probability density function model, without spatial resolution. Employing detailed chemistry has the potential of predicting emissions, but generally results in very high CPU costs. Here it is shown that CPU times of a couple of minutes per cycle can be reached when applying detailed chemistry, and CPU times below 10 seconds per cycle can be reached when using reduced chemistry while still catching in-cylinder in-homogeneities. This makes the tool usable for efficient engine performance mapping and optimisation. To meet CPU time requirements, automatically load balancing parallelisation was included in the model. This allowed for an almost linear CPU speed-up with number of cores available.

Partially Premixed Combustion (PPC) engines have demonstrated a potential for high efficiency and low emissions operation. To be able to study the combustion in detail but also to perform parametric studies on the potential of the PPC concept a one dimensional (1D) engine simulation tool was used with 1; a prescribed burn rate 2; predictive combustion tool with reduced chemical model and 3; predictive combustion tool with detailed chemical models. Results indicate that fast executing reduced chemistry work reasonably well in predicting PPC performance and that n-decane is possibly a suitable diesel substitute in PPC modeling while n-heptane is not.

Interactions between in-cylinder combustion and emission aftertreatment need to be understood for optimizing the overall powertrain system. Numerical investigations can aid this process. For this purpose, simple and numerically fast, but still accurate models are needed for in-cylinder combustion and exhaust aftertreatment. The chemical processes must be represented in sufficient detail to predict engine power, fuel consumption, and tailpipe emission levels of NOx, soot, CO and unburned hydrocarbons. This paper reports on a new transient one-dimensional catalyst model. This model makes use of a detailed kinetic mechanism to describe the catalytic reactions. A single-channel or a set of representative channels are used in the presented approach. Each channel is discretized into a number of cells. Each cell is treated as a perfectly stirred reactor (PSR) with a thin film layer for washcoat treatment. Heat and mass transport coefficients are calculated from Nusselt and Sherwood laws.

The solution mapping method Adaptive Polynomial Tabulation (APT) for complex chemistry is presented. The method has the potential of reducing the computational time required for stochastic reactor model simulations of the HCCI combustion process. In this method the solution of the initial value chemical rate equation system is approximated in real-time with zero, first and second order polynomial expressions. These polynomials are algebraic functions of a progress variable, pressure and total enthalpy. The chemical composition space is divided a priori into block-shaped regions (hypercubes) of the same size. Each hypercube may be divided in real-time into adaptive hypercubes of different sizes. During computations, initial conditions are stored in the adaptive hypercubes. Two concentric Ellipsoids of Accuracy (EOA) are drawn around each stored initial condition.

The subject of this work is 3D numerical simulations of combustion and soot emissions for a passenger car diesel engine. The CFD code STAR-CD version 3.26 [1] is used to resolve the flowfield. Soot is modeled using a detailed kinetic soot model described by Mauss [2]. The model includes a detailed description of the formation of polyaromatic hydrocarbons. The coupling between the turbulent flowfield and the soot model is achieved through a flamelet library approach, with transport of the moments of the soot particle size distribution function as outlined by Wenzel et al. [3]. In this work we extended this approach by considering acetylene feedback between the soot model and the combustion model. The model was further improved by using new gas-phase kinetics and new fitting procedures for the flamelet soot library.

This paper explains the principle and advantages of the Ignition Progress Variable Library (IPV-Library) approach and its use in predicting engine related premixed, non-premixed and compression ignited combustion events. The implementation of IPV-Library model in the engine-focused CFD code VECTIS is described. To demonstrate the application of the model in predicting various types of combustion, computational results from a 2-stroke HCCI engine, a premixed spark ignition engine and an HSDI diesel engine are presented, together with some comparisons with engine test data.

Today’s car manufactures inevitably have to focus on the reduction of fuel consumption while maintaining high performance standards. In this respect, the downsized turbocharged DISI (Direct Injection Spark Ignition) engine represents an appealing solution. However, downsizing is limited because of knocking phenomena occurring at high- and full-load conditions due to autoignition of the unburned mixture ahead the flame front. A common way of reducing knock tendencies is provided by Exhaust Gas Recirculation (EGR). However, EGR modifies the chemical composition of the cylinder charge and recirculated species like nitric oxide (NO) or unburned Hydrocarbons (HC) particularly increase the reactivity of the unburned mixture. In other words, the EGR influences the Octane Number (ON) of the in-cylinder gases.

The formation of exothermic centers was modeled with a Stochastic Reactor Model (SRM) to investigate their impact on HCCI combustion. By varying the exhaust valve temperature, and thus assigning more realistic wall temperatures, the formation of exothermic centers and the ignition timing was shifted in time. To be able to study the exothermic centers, their formation and their distribution, Scatter plots, standard deviation plots and Probability Density Function (PDF) plots were constructed on the basis of the data the SRM calculations provided. The standard deviation for the particle temperatures was found to be an useful indicator of the degree of homogeneity within the combustion chamber, and thus of how efficient the combustion process was. It was observed that when the standard deviation of the temperature was higher, the emissions of CO and of hydrocarbons present at the end of the closed cycle were higher.

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.

An investigation on reducing the set of modeling parameters for engine cycle simulation is presented. The investigation considers a detailed kinetic model for combustion and emissions predictions coupled to a complete cycle simulation tool applied to a modern Diesel engine. The analysis is based on a previously developed method that combines a 1-D gas dynamics model with a stochastic reactor model for direct injection engines (SRM-DI). Initially, the global and instantaneous performance parameters of a Diesel engine were simulated at different operating conditions. The model was validated and the simulated results were compared to experimental data to assess the quality of the model. Afterwards, the influence of the chosen modeling parameters on engine performance, such as in-cylinder pressure, emissions and global performances, were analyzed. The mixing time proved to be the most important modeling parameter for the stochastic reactor model.

In one-dimensional engine simulation programs the simulation of engine performance is mostly done by parameter fitting in order to match simulations with experimental data. The extensive fitting procedure is especially needed for emissions formation - CO, HC, NO, soot - simulations. An alternative to this approach is, to calculate the emissions based on detailed kinetic models. This however demands that the in-cylinder combustion-flow interaction can be modeled accurately, and that the CPU time needed for the model is still acceptable. PDF based stochastic reactor models offer one possible solution. They usually introduce only one (time dependent) parameter - the mixing time - to model the influence of flow on the chemistry. They offer the prediction of the heat release, together with all emission formation, if the optimum mixing time is given.

A transient interactive flamelet model and a transient flamelet library based model are used to model a medium-duty diesel fueled engine operating in PCCI mode. The simulations are performed with and without the source term accounting for evaporation in the mixture fraction variance equation. Reasonable agreement is found with the experiments with both models. The effect of the evaporation source term in the mixture fraction variance equation is different for the different transient flamelet approaches. For the transient interactive flamelet model the ignition onset is delayed as a consequence of the higher mixture fraction variance, which leads to a higher scalar dissipation rate. The evaporation source term does not affect the global characteristics of the ignition event for the transient flamelet progress variable model, but locally the initial combustion is occurring differently.

Concentrations of hydroxyl radicals and formaldehyde were calculated using homogeneous (HRM) and stochastic reactor models (SRM), and the result was compared to LIF-measurements from an optically accessed iso-octane / n-heptane fuelled homogeneous charge compression ignition (HCCI) engine. The comparison was at first conducted from averaged total concentrations / signal strengths over the entire combustion volume, which showed a good qualitative agreement between experiments and calculations. Time- and the calculation inlet temperature resolved concentrations of formaldehyde and hydroxyl radicals obtained through HRM are presented. Probability density plots (PDPs) through SRM calculations and LIF-measurements are presented and compared, showing a very good agreement considering their delicate and sensitive nature.

We numerically simulate a Homogeneous Charge Compression Ignition (HCCI) engine fuelled with a blend of ethanol and diethyl ether by means of a stochastic reactor model (SRM). A 1D CFD code is employed to calculate gas flow through the engine, whilst the SRM accounts for combustion and convective heat transfer. The results of our simulations are compared to experimental measurements obtained using a Caterpillar CAT3401 single-cylinder Diesel engine modified for HCCI operation. We consider emissions of CO, CO2 and unburnt hydrocarbons as functions of the crank angle at 50% heat release. In addition, we establish the dependence of ignition timing, combustion duration, and emissions on the mixture ratio of the two fuel components. Good qualitative agreement is found between our computations and the available experimental data.

By dividing the combustion process into several phases with phase optimized skeletal mechanisms (POSM), gains in calculation speed were realized with virtually no loss in accuracy. A skeletal mechanism is a reduced mechanism where only the significant species, determined through a set of parameters (one for each species), remain with respect to a detailed mechanism. The parameter is based on a combination of sensitivity and flow analysis. Within the POSM method machine learning algorithms are used to automatically determine and recognize the major phases. Reduction is achieved by keeping only the significant species with respect to each phase. Each phase has a different mechanism, derived from the original and each is smaller than the original.