Motivation

Emission standards for marine diesel engines have been tightened up in the past few years, and will become even more restrictive in the future. A promising concept for meeting emission limits is to use pilot ignited lean-burn natural gas engines. Apart from the reduced CO2 emissions, the low C/H-ratio of natural gas leads to lower flame temperatures and thus lower NOX emissions. Nevertheless homogeneous charge combustion with lean natural gas mixtures suffers from considerable methane slip and restricts high compression ratios due to the knocking limit.

Direct injection of main fuel and pilot fuel allows more flexible control of the combustion process regarding the mixture formation and fuel burnout, thus fuel slip can be reduced and the compression ratio increased – leading to higher efficiency. Furthermore, this concept is ideally suited for various combinations of gaseous and liquid fuels. This is of great importance when considering the fact that future fuel mixes will become more diverse and will include more biofuels. Up till now there has been little research done on injection, mixing, ignition and fuel burnout of this concept. In the course of this project dual fuel direct injection combustion for flexible fuels will be experimentally and numerically investigated.

Experimental studies

Two different test rigs are used for the experimental investigation. General investigations regarding local mixture formation, ignition and combustion progress are studied using a rapid compression-expansion machine (RCEM). The RCEM is pneumatically driven and the movement of the piston is hydraulically controlled. This test rig allows full optical access to the combustion chamber during one single cycle via the glass piston and the two side windows.

The second apparatus is a periodically chargeable constant volume cell which is employed for stability analysis because it can perform 60 combustion events in a row. Optical observation quality increases and the mechanical complexity is simplified when no moving piston is used. This comes at the cost of a rather complex pre-conditioning, consisting of a large volume high pressure air supply and high-capacity air heaters. Additionally, the charge movement deviates strongly from the one in a real engine, although this is of minor importance due to the fact that the focus is laid on fundamental phenomena rather than real motoring conditions.

Different measurement techniques will be used to gain information about various quantities. Apart from pressure indication, these include Tracer-LIF measurements of the mixing field, Schlieren- and Shadowgraph Imaging for jet penetration and structure, and OH*-chemiluminescence imaging of the flame propagation. The results will lead to a better understanding of the underlying phenomena and provide an experimental validation for the numerical model development.

Numerical studies

The 3D-CFD simulation of diesel injection including spray break-up, ignition and combustion is investigated using existing models. Modeling direct injection of natural gas, however, is very challenging due to high pressure ratios. This leads to high gas velocities and under-expanded nozzle flow. Ignition delay times differ a lot from those of pure diesel combustion depending on the injection strategy. An extension for dual fuel combustion will be developed on the basis of existing ignition and combustion models for diesel engines. Promising base models are the ECFM-3Z and the PVM, both already implemented in the commercial CFD code StarCD. All effects are investigated numerically and experimental data are used to validate the results and to improve the models.