Objectives: Direct Injection Spark Ignited (DISI) engines are superior to other SI engines in terms of fuel efficiency and CO2 emissions. The topic of DISI engine hydrocarbon requires a new consideration of the problem of droplet wall interaction. Wall films and pools of un-burnt fuel formed by spray or droplet wall impingement is a major source of hydrocarbon emissions. To allow the effective design of DISI engines with respect to mixture formation, CFD tools are indispensable. These CFD codes commonly make use of Eulerian-Lagrangian methods for spray simulation and require appropriate droplet wall interaction models. It is the goal of this project to derive a general droplet wall interaction model applicable to the entire range of conditions present during operation of DISI engines. Description of the Work: Eulerian-Lagrangian methods track individual droplets along their trajectory until the droplets either evaporate, leave the simulation domain or impinge on a wall boundary. In the latter case correlations are required that link the pre and post-impingement conditions of a droplet and secondary droplets respectively. The special conditions in a DISI engine cylinder, i.e. small droplets, high temperatures and a wide range of pressures, do not allow for simple extrapolation of the results of numerous earlier investigations. Since a complete set of information is required for CFD simulations simultaneous measurements of droplet number, size, velocity vector and temperature in the hemisphere above the impact location, plus the resulting wall heat flux, are necessary. The project covers single droplet impacts, droplet chains, monosized sprays and finally polydisperse sprays and considers in particular the relevance of wall temperature and gas phase pressure. This measurement task will be distributed between several project partners according to the appropriateness of their test facilities. With the correlations evolving from these measurements a general droplet wall interaction model for single droplet impact is derived and then extended to encompass high droplet impact frequencies and number densities. Finally the developed interaction model is tested on real polydisperse spray applications under DISI conditions and validated with experimental measurements on injections into a constant volume chamber or optical DISI engine. Expected Results and Exploitation Plans: Besides a significant increase in understanding of the physical processes present in droplet wall impingement, a broad and consistent database is collected. With this data a general statistical model of droplet wall interactions for single and multiple droplet wall interaction can be constructed, accounting for all major influencing parameters. The development of theoretical models will also be supported by this information. DWDIE results enhance the capability and precision of CFD codes to serve as design tools in the development of efficient and clean DISI engines.
From the investigations of droplet-wall-interaction within DWDIE project two major results emerged: a comprehensive data base and models for droplet impact on walls accounting for different conditions. The data base is comprehensive in the way, that numerous experiments have been performed covering wide ranges of parameters. The impact of single droplets, droplet chains, and mono-disperse sprays on walls have been investigated varying the walls's temperature and roughness, back pressure, droplet fluid, impact angle, incoming droplet size, velocity, and frequency/density, where the latter parameter is significant for the distinction between wet and dry walls. Visualisations via images and high speed films show very detailed droplet impact phenomenology. Within the investigated multi-dimensional parameter space the droplet-wall interaction regimes "deposition", "rebound", and "secondary atomisation" (i.e. breakup) are identified, where the latter is divided in further subregimes. Using phase-Doppler-anemometry detailed information on secondary droplet velocity and size is provided.
The experimental data base also includes investigations on spray impingement for different conditions as well as measurements in optical accessible engines. Furthermore, simulations using a volume-of-fluids method have been performed, which help to increase the understanding of physical processes during droplet impingement. The data provided by DWDIE experiments was used to derive models for droplet-wall interaction accounting for the large range of investigated parameters. In this context firstly, existing models from literature were implemented in CFD codes and compared to the new experimental results. It was shown that due to their derivation from smaller data bases, they are not adequate enough to cover a larger range of parameters e.g. as it appears in real GDI sprays. The new models, based on DWDIE results, literature, and basic correlations are an enhancement due good agreements with the experiments for larger parameter ranges. For the implementation a general code-independent structure is provided.
Funding SchemeCSC - Cost-sharing contracts
412 96 Goeteborg