R.S.G. Baert

Design and Operation of a High Pressure, High Temperature Cell for HD Diesel Spray Diagnostics: Guidelines and Results

This paper first compares strengths and weaknesses of different options for performing optical diagnostics on HD diesel sprays. Then, practical experiences are described with the design and operation of a constant volume test cell over a period of more than five years. In this test rig, pre-combustion of a lean gas mixture is used to generate realistic gas mixture conditions prior to fuel injection. Spray growth, vaporization are studied using Schlieren and Mie scattering experiments. The Schlieren set-up is also used for registration of light emitted by the combustion process; this can also provide information on ignition delay and on soot lift-off length. The paper further describes difficulties encountered with image processing and suggests methods on how to deal with them. Results are presented that illustrate the wide range of capabilities of this test-rig when combined with high speed video registration, in particular its potential for studying issues related to vaporizing fuel spray dynamics.

Survivability of thermographic phosphors (YAG:Dy) in a combustion environment

The feasibility of applying laser-induced phosphorescence in a combustion environment was shown by testing the consistency of the emission‐temperature relations of thermographic phosphor particles (YAG:Dy). The relations were calibrated before and after the phosphor particles had passed a flame front. The calibrations were performed in air and in pure oxygen. The emission‐temperature relation prevails from around 300 K to 1300 K. The difference in emission‐temperature relation for the two different cases is less than the experimental precision (3%).

Direct injection of high pressure gas : scaling properties of pulsed turbulent jets

Existing gasoline DI injection equipment has been modified to generate single hole pulsed gas jets. Injection experiments have been performed at combinations of 3 different pressure ratios (2 of which supercritical) respectively 3 different hole geometries (i.e. length to diameter ratios). Injection was into a pressure chamber with optical access. Injection pressures and injector hole geometry were selected to be representative of current and near-future DI natural gas engines. Each injector hole design has been characterized by measuring its discharge coefficient for different Re-levels. Transient jets produced by these injectors have been visualized using planar laser sheet Mie scattering (PLMS). For this the injected gas was seeded with small oil droplets. The corresponding flow field was measured using particle image velocimetry (PIV) laser diagnostics. From the corresponding measurements, both the jet spreading angle and penetration have been determined according to different definitions and the interrelation between these definitions has been examined. Results show that -beyond the initial transition period and almost up to the tip vortex region - (average) jet angle is almost constant. Furthermore, jet penetration is well predicted by correlations that implicitly assume momentum conservation at constant static pressure. Measurements suggest a different time-averaged velocity profile from that typically assumed in some of these correlations. Tip vortex position and size scale with transient jet penetration.

Uncooled EGR as a means of limiting wall-wetting under early direct injection conditions

Collision of injected fuel spray against the cylinder liner (wall-wetting) is one of the main hurdles that must be overcome in order for early direct injection Premixed Charge Compression Ignition (EDI PCCI) combustion to become a viable alternative for conventional DI diesel combustion. Preferably, the prevention of wall-wetting should be realized in a way of selecting appropriate (most favorable) operating conditions (EGR level, intake temperature, injection timing-strategy etc.) rather than mechanical modification of an engine (combustion chamber shape, injector replacement etc.). This paper presents the effect of external uncooled EGR (different fraction) on wall-wetting issues specified by two parameters, i.e. measured smoke number (experiment) and liquid spray penetration (model). Experiments performed in a dedicated heavy-duty direct injected (HDDI) diesel engine suggest that the elevation of intake temperature caused by delivery of external uncooled exhaust gases led to significant reduction in wall wetting. This is combined with IMEP improvement. In-house spray- and ignition modeling was used to gain insight into the measured trends. Copyright

Towards control-oriented modeling of natural gas-diesel RCCI combustion

For natural gas (NG)-diesel RCCI, a multi-zonal, detailed chemistry modeling approach is presented. This dual fuel combustion process requires further understanding of the ignition and combustion processes to maximize thermal efficiency and minimize (partially) unburned fuel emissions. The introduction of two fuels with different physical and chemical properties makes the combustion process complicated and challenging to model. In this study, a multi-zone approach is applied to NG-diesel RCCI combustion in a heavy-duty engine. Auto-ignition chemistry is believed to be the key process in RCCI. Starting from a multi-zone model that can describe auto-ignition dominated processes, such as HCCI and PCCI, this model is adapted by including reaction mechanisms for natural gas and NOx and by improving the incylinder pressure prediction. The model is validated using NG-diesel RCCI measurements that are performed on a 6 cylinder heavy-duty engine. For three different engine operating points, it is operated at various diesel injection timings and NG-diesel blend ratios. The validation is focused on variables that are relevant for engine control, such as CA50, peak cylinder pressure, and engine-out NOx emissions. The validation shows that the multi-zone method with detailed chemistry reproduces the correct trends for important control parameters. From this validated model, real-time, map-based RCCI models are derived, which are considered to be an important step towards model-based NG-diesel RCCI control development.

Comparing Different Turbulence Modelling Approaches for Square Duct Simulations Using the Kiva Code

In internal-combustion engines (ICE) the fluid dynamics is characterized by strong anisotropy. The standard two equation k–e model is well known to be not appropriate in this case. A detailed study of the numerical modelling properties of the well known Kiva-3V code has been performed for two different approaches to anisotropic turbulence modelling. In the first approach a Smagorinsky-type LES model is evaluated. In the second approach a Time-Dependent RANS model has been adopted, using the Explicit Algebraic Stress Model of Gatski and Speziale [1]. For validation of both approaches numerical simulations of a turbulent flow in a square duct geometry are compared to DNS data. It is concluded from this work that the applied RANS approach is the best available practise to model the anisotropic properties of the fluid flow for ICE simulations as long as the computational resources to perform real LES simulations remain limited.Copyright