Search Results

To achieve global climate goals, greenhouse gas emissions must be drastically reduced. The energy and transportation sectors are responsible for about one third of the greenhouse gases emitted worldwide, and they often use internal combustion engines (ICE). One effective way to decarbonize ICEs may be to replace carbon-containing fossil fuels such as natural gas entirely, or at least partially, with hydrogen. Cost-effective development of sustainable combustion concepts for hydrogen and natural gas/hydrogen mixtures in ICEs requires the intensive use of fast and robust simulation tools for prediction. The key challenge is appropriate modeling of flame front propagation. This paper evaluates and applies different approaches to modeling laminar flame speeds from the literature. Both appropriate models and reaction kinetic calculations are considered.

Precise prediction of combustion parameters such as peak firing pressure (PFP) or crank angle of 50% burned mass fraction (MFB50) is essential for optimal engine control. These quantities are commonly determined from in-cylinder pressure sensor signals and are crucial to reach high efficiencies and low emissions. Highly accurate in-cylinder pressure sensors are only applied to test rig engines due to their high cost, limited durability and special installation conditions. Therefore, alternative approaches which employ virtual sensing based on signals from non-intrusive sensors retrieved from common knock sensors are of great interest. This paper presents a comprehensive comparison of selected approaches from literature, as well as adjusted or further developed methods to determine engine combustion parameters based on knock sensor signals. All methods are evaluated on three different engines and two different sensor positions.

Virtual sensing refers to the processing of desired physical data based on measured values. Virtual sensors can be applied not only to obtain physical quantities which cannot be measured or can only be measured at an unreasonable expense but also to reduce the number of physical sensors and thus lower costs. In the field of spark ignited internal combustion engines, the virtual sensing approach may be used to predict the cylinder pressure signal (or characteristic pressure values) based on the acceleration signal of a knock sensor. This paper presents a method for obtaining the cylinder pressure signal in the high-pressure phase of an internal combustion engine based on the measured acceleration signal of a knock sensor. The approach employs a partial differential equation to represent the physical transfer function between the measured signal and the desired pressure. A procedure to fit the modeling constants is described using the example of a large gas engine.

In-cylinder flows in internal combustion engines are highly turbulent in nature. An important property of turbulence that plays a key role in mixture formation is anisotropy; it also influences ignition, combustion and emission formation. Thus, understanding the turbulence anisotropy of in-cylinder flows is critical. Since the most widely used two-equation Reynolds-averaged Navier-Stokes (RANS) turbulence models assume isotropic turbulence, they are not suitable for correctly capturing the anisotropic behavior of turbulence. However, large eddy simulation (LES) can account for the anisotropic behavior of turbulence. In this paper, the Reynolds stress tensor (RST) is analyzed to assess the predictive capability of RANS and LES with regard to turbulence anisotropy. The influence of mesh size on turbulence anisotropy is also looked into for multi-cycle LES.

Turbulence anisotropy has a great influence on mixture formation and flame propagation in internal combustion engines. However, the visualization of turbulence in simulations is not straightforward; traditional methods lack the ability to display the anisotropic properties in the engine geometry. Instead, they use invariant maps, and important information about the locality of the turbulence anisotropy is lost. This paper overcomes this shortcoming by visualizing the anisotropy directly in the physical domain. Componentality contours are applied to directly visualize the anisotropic properties of turbulence in the three-dimensional engine geometry. Using an RGB (red, green, blue) color map, the three limiting states of turbulence (one-component, axisymmetric two-component and isotropic turbulence) are displayed in the three-dimensional physical domain.

Lean burn combustion concepts with high mean effective pressures are being pursued for large gas engines in order to meet future stringent emission limits while maintaining high engine efficiencies. Since severe boundary conditions for the ignition process are encountered with these combustion concepts, the processes of spark ignition and flame initiation are important topics of applied research, which aims to avoid misfiring and to keep cycle-to-cycle combustion variability within reasonable limits. This paper focuses on the fundamental investigation of early flame kernel development using different ignition system settings. The investigations are carried out on a rapid compression-expansion machine in which the spark ignition process can be observed under engine-like pressure and excess air ratio conditions while low flow velocities are maintained.

Using natural gas as a fuel in internal combustion engines is a promising way to obtain efficient power generation with relatively low environmental impact. Dual fuel operation is especially interesting because it can combine the safety and reliability of the basic diesel concept with fuel flexibility. To deal with the greater number of degrees of freedom caused by the interaction of two fuels and combining different combustion regimes, it is imperative to use simulation methods in the development process to gain a better understanding of the combustion behavior. This paper presents current research into ignition and combustion of a premixed natural gas/air charge with a diesel pilot spray in a large bore diesel ignited gas engine with a focus on 3D-CFD simulation. Special attention was paid to injection and combustion. The highly transient behavior of the diesel injector especially at small injection quantities poses challenges to the numerical simulation of the spray.

It is critical for gas and dual fuel engines to have improved transient characteristics in order that they can successfully compete with diesel engines. Testing of transient behavior as well as of different control strategies for the multi-cylinder engine (MCE) should already be done on the single cylinder engine (SCE) test bed during the development process. This paper presents tools and algorithms that simulate transient MCE behavior on a SCE test bed. A methodology that includes both simulation and measurements is developed for a large two-stage turbocharged gas engine. Simple and fast models and algorithms are created that are able to provide the boundary conditions (e.g., boost pressure and exhaust back pressure) of a multi-cylinder engine in transient operation in real-time for use on the SCE test bed. The main models of the methodology are discussed in detail.

To achieve high power output and good efficiency and to comply with increasingly stricter emission standards, modern combustion engines require a more complex engine design, which results in a higher number of control parameters. As the measurement effort and the number of sensors for engine development at the test bed continue to increase, it is becoming nearly impossible for the test bed engineer to manually check measurement data quality. As a result, automated methods for analysis and plausibility checks of measurement data are necessary in order to find faults as soon as they occur and to obtain test results of the highest possible quality. This paper presents a methodology for automated fault diagnosis on engine test beds. The methodology allows reliable detection of measurement faults as well as the identification of the root cause of faults.

As emission limits become increasingly stringent and the price of gaseous fuels decreases, more emphasis is being placed on promoting gas engines. In the field of large engines for power generation, dual fuel combustion concepts that run on diesel/natural gas are particularly attractive. Knock in diesel/natural gas dual fuel engines is a well known yet not fully understood complex phenomenon that requires consideration in any attempt to increase load and efficiency. Thus combustion concept development requires a reliable yet robust methodology for detecting knock in order to ensure knock-free engine operation. Operating parameters such as rail pressure, start of injection and amount of diesel injected are the factors that influence oscillations in the in-cylinder pressure trace after the start of combustion. Oscillations in the pre-mixed combustion phase, or ringing, are caused by the rapid conversion of large parts of the injected diesel.

Optimization of engine warm-up behavior has traditionally made use of experimental investigations. However, thermal engine models are a more cost-effective alternative and allow evaluation of the fuel saving potential of thermal management measures in different driving cycles. To simulate the thermal behavior of engines in general and engine warm-up in particular, knowledge of heat distribution throughout all engine components is essential. To this end, gas-side heat transfer inside the combustion chamber and in the exhaust port must be modeled as accurately as possible. Up to now, map-based models have been used to simulate heat transfer and fuel consumption; these two values are calculated as a function of engine speed and load. To extend the scope of these models, it is increasingly desirable to calculate gas-side heat transfer and fuel consumption as a function of engine operating parameters in order to evaluate different ECU databases.

When developing and later using simulation models for combustion prediction in internal combustion engines, it is first of all necessary to determine the model constants. This paper describes the development of a method for the automated determination of model parameters which can be applied to any internal combustion simulation model. The work is not aimed at developing a new optimizing algorithm but at adjusting and adapting an existing optimizer to the special needs and convergence problems, which occur when applied to combustion models. Consequently, the paper describes the set-up of the objective function and several methods for improving the convergence. Finally, an outline for a strategy which uses the optimizing tool for model development is presented.

In this paper a zero-dimensional simulation methodology for efficient pre-optimization of the combustion process in DI diesel engines is presented. A new model for the calculation of the rate of heat release is unveiled. It is based on the separate description of both the primary processes closely related to the fuel jet as well as the following combustion of the fuel mass remaining after the end of injection. The modeling of fuel mass distribution between premixed and diffusion combustion as well as a model for the fuel preparation time are explained. Furthermore, models for the calculation of ignition delay and premixed combustion based on an extended Arrhenius formulation are discussed, as well as turbulent combustion on the basis of a Magnussen model. The new features of the heat release model prove to be necessary to describe the effects of modern high-pressure fuel injection systems on the combustion process regarding the strong influence of the injection rate on the burn rate.