APPLIED COMBUSTION RESEARCH

MODELING FUEL-ENGINE INTERACTIONS IN ADVANCED ENGINES

Future advanced internal combustion (IC) engines need to achieve higher efficiency, lower exhaust gas emissions, and--most important--be developed in tandem with new fuel formulations. Requirements for new modes of operations are different from the conventional spark-ignition (SI) and compression-ignition (CI) engines. For next generation SI engines, downsized and highly boosted gasoline direct injection engines can improve vehicle fuel economy; however, increased in-cylinder pressure cannot eliminate abnormal combustion phenomena such as pre-ignition and super-knock, which leads to the destruction of the engine. For next generation CI engines, the flagship program of the Saudi Aramco Fuel Technology Team, Gasoline Compression Ignition (GCI), with low octane fuel, is designed to improve fuel efficiency and well-to-wheel CO2 emissions. To resolve the most critical issues in ongoing fuel technology activities, and to successfully meet the requirements of new fuel formulations, a fundamental understanding of underlying physical and chemical characteristics is a prerequisite. To this end, cutting-edge diagnostics tools, to visualize and accurately model fuel and combustion engine behaviors, are essential to optimize new fuel and engine systems.

OBJECTIVES

The objective of this project is to develop advanced visualization and computational diagnostics to match new fuel formulations with novel internal combustion engine technologies. The following three major tasks are proposed:

  • GCI engine fuel spray mixing and combustion characterization in an optically-accessible combustion vessel and optical internal combustion engine;
  • advanced imaging and analysis of pre-ignition and super-knock phenomena in an optical vessel, shock tube, and pre-ignition engine;
  • quantitative prediction of complex physical and chemical phenomena using high fidelity simulations to help GCI combustion system design and identify the root cause of pre-ignition and super-knock.

The program is divided into three main sub-projects.

More information about Modeling of Fuel-Engine Interactions in Advanced Engines Research and project-specific contact details are available from:

Hong G. Im, Ph. D.
hong.im@kaust.edu.sa
Professor, Mechanical Engineering
Clean Combustion Research Center (CCRC)
King Abdullah University of Science and Tehnology (KAUST)

Junseok Chang,  Ph.D.
junseok.chang@aramco.com
Engine Combustion Team Leader
Fuel Technology Division
Research and Development Center, Saudi Aramco

 

1. SPRAY MIXING AND COMBUSTION CHARACTERIZATION

This task will support ongoing research and development at Saudi Aramco to characterize the behavior of naphtha fuels in commercial injectors, and to provide benchmark data to calibrate high fidelity spray models employed with computational fluid dynamics (CFD) tools for full-cycle engine simulations. Currently, heavy-duty diesel injectors and light-duty GDI injectors are being investigated for their behavior under reacting and non-reacting conditions. The task will center on two experimental facilities: (a) an optically accessible constant volume combustion chamber (CVCC) and (b) optical single cylinder engines.

A constant volume combustion chamber (CVCC) offers an easy and cost-effective way to model engine conditions. To provide high-temperature, high-pressure environments similar to the compression stroke of an internal combustion engine, a constant-volume combustion chamber with optical access will be used to produce engine in-cylinder conditions.

Read more

The combined effect of fuel volatility and reactivity on in-cylinder mixture stratification will be investigated;this will be achieved primarily through the study of optical and metal engines. Effects of evaporative cooling from fuel injection will be analyzed for rich and lean mixtures. Fuel blends displaying various chemical kinetic and volatility behaviors will be studied at different operating conditions, including inlet temperature and pressure, mixture strength, charge density, compression ratio, ignition timing, EGR variations, and swirl or tumble intensity. The fuels to be studied are gasoline-diesel mixtures and light/heavy naphtha mixtures.

Read more

 

2. PRE-IGNITION AND SUPER-KNOCK CHARACTERIZATION

To study pre-ignition and super-knock characteristics, the proposed project will focus on the fundamental processes that govern pre-ignition in SI engines. It is postulated that the initiation of pre-ignition is caused by a variety of factors, such as oil droplets, metal particles and turbulence-induced mixture/temperature fluctuations. Three conditions are considered for the initiation of this type of pre-ignition and flame propagation: a) autoignition delay time of a hot spot, b) critical heat of the initiation of autoignition, and c) critical heat of the initiation of flame propagation.

The plan is to modify and adapt the constant volume combustor (laminar) already available in CCRC for performing combustion at elevated pressures and temperatures, to allow operation at similar conditions in a boosted SI engine. Part of this subtask will be to complete the ANSYS modeling of the vessel to ensure its safe operation at these elevated conditions. Enhanced heaters and insulation will be installed to facilitate high initial temperature operation. The constant volume combustor is equipped with three orthogonal pairs of quartz windows providing excellent optical access to allow photographic (high-speed Schlieren imaging and chemiluminescence) and laser (OH-PLIF) diagnostics, both available at CCRC.

Read more

Conventionally postulated causes of pre-ignition (oil droplets, metal particles, turbulence) are absent in the homogeneous environment of the shock tube. Therefore, it is interesting to systematically observe the pre-ignition events in a shock tube. The driven section of the KAUST shock tube will be modified to provide optical access for high-speed imaging and Schlieren studies. The proposed work will focus on four aspects: (i) how pre-ignition varies with fuel type (n-alkane to iso-alkane to aromatics); (ii) what regimes/conditions (T, P, equivalence ratio) favor pre-ignition; (iii) what factors influence pre-ignition in the homogeneous environment of a shock tube; (iv) how to avoid or minimize pre-ignition (additives, mixing).

Read more

Pre-ignition studies in a metal engine will be studied under this program. Lubricant oil will be introduced to the engine by both controlled and uncontrolled means. In the controlled mode, an oil droplet, or mist, will be introduced from the port fuel injector (PFI). The pre-ignition events that will occur sporadically will be captured, evaluated, and statistically analyzed to determine pre-ignition frequency of the engine under a specific condition. Pre-ignition frequency and the root of the pre-ignition will be identified with the aid of a combustion measurement platform that comprises visualization and thermal imaging (infrared camera and endoscope), and indicating modules. Each fuel and lubricant will be characterized for its pre-ignition frequency and pre-ignition temperature.

Read more


3. PREDICTIVE MODELING AND SIMULATIONS OF SPRAY AND COMBUSTION

Development of advanced computational modeling capabilities is a critical component of this project. The main scope of the modeling development the improved ability to describe spray break-up, penetration, and mixing, and subsequent mixed-mode combustion processes. The task will be undertaken in two concurrent tracks: (a) spray and combustion characterization and validation against experiments through ongoing collaborations using commercial software, CONVERGE, and (b) development of advanced large eddy simulation (LES) capabilities to describe spray, mixing, and combustion. (c) High fidelity simulations will also be conducted to characterize pre-ignition and super-knock phenomena to complement the experimental effort.

The capabilities of CONVERGE will be utilized in simulations of ignition quality testers (IQT) and full-cycle engines. With additional developments in spray and combustion submodels, these simulations will be capable of producing full-cycle engine combustion characteristics for comparison with experimental measurements. The IQT simulations will provide essential spray characteristics which are difficult to measure experimentally, such as spray dispersion and evaporation. The modeling will provide valuable information about physical versus chemical ignition delay processes encountered in the IQT devices, thus serving as a reference data set to independently validate spray models as well as the underlying chemical kinetic models.​

Read more

Next generation high fidelity simulation of engine combustion attempts to provide detailed information about local and transient characteristics such as wall quenching or cycle-to-cycle variability. Large eddy simulation (LES) for turbulent flow and passive scalar mixing has been studied for many decades, but LES of turbulent reacting flows--especially incorporating detailed chemistry--has only been developed in recent years. Particularly in IC engine applications, accurate descriptions of scalar mixing, combustion, liquid spray breakup, atomization, turbulent dispersion and evaporation are critical for predictive simulation of highly transient in-cylinder flow and combustion processes. The proposed project will develop a comprehensive state-of-the-art LES simulation capability which,  for fidelity and computational efficiency, is optimally suited for modern engine applications.

Read more

To complement and cross-validate the experimental effort, high fidelity predictive models will be developed and parametric simulations will be conducted. In particular, the direct numerical simulation (DNS) code developed at KAUST CCRC will be utilized and modified to computationally reproduce a well-defined pre-ignition event. The code is based on high-order finite differencing and explicit/implicit time integration, incorporating detailed/reduced chemical reaction mechanisms and transport properties built on the Chemkin framework. The code has been fully tested for excellent near-linear scalability for massively parallel computation. The DNS, without turbulent submodels, will provide temporally- and spatially-resolved information to unravel fundamental characteristics associated with pre-ignition behavior.

Read more

Reactive simulations of combustion processes require accurate description of the chemical kinetics; however, most fuel chemistry models are highly detailed and consist of hundreds to thousands of species and reactions. This work will focus on two aspects of fuel chemistry: (a) developing surrogate models to represent the real fuels with a few molecules, (b) size reduction of the detailed models for use in CFD simulations.

Read more