Investigation of the fuel film formation from spray-wall interaction of clean and fouled fuel injectors

Abstract

In direct injection spark ignition engine (DISI), a liquid spray interacting with a solid surface (referred to as spray-wall interaction) is considered to be one of the major sources of soot emissions. A consequence of spray-wall interaction is a fuel film on a surface, which often evaporates less quickly than the duration of an engine cycle. As a flame contacts a fuel film a pool fire is produced, which leads to increased soot emission from DISI engines. Late evaporating fuel films formed due to the spray wall interaction need to be explored, as they are a prominent source of soot emissions through the DISI engine. The fuel injector within DISI engines is exposed to high temperatures from compressed gases and flames. These conditions can lead to fuel deposits on the injector surface or within the injector nozzles where liquid fuel is present when the injector is exposed to high temperatures. These deposits can alter the spray pattern formation, which may ultimately impact the spray-wall interaction and hence film formation behaviour that leads to soot production. The effect of the injector deposit, on spray formation in the DISI engine and its relation to the film formation on the plate still needs to be explored.

In this thesis, the film formation behaviour of clean and fouled (i.e., injectors with deposits) injectors are investigated to evaluate the differences in film formation and evaporation behaviour under ambient conditions. The injectors in question are a multi-hole (6-hole) injectors intended for a side-mounted orientation within a DISI engine. The difference in spray formation behaviour of clean and fouled injectors is analysed within the first set of experiments presented in the thesis. The spray pattern is studied with an 80-bar injection pressure and 1ms injection duration using spray shadowgraphy and side illumination imaging. The penetration length (distance from the injector nozzle to the farthest point along the plume axis for fully developed spray plume) of each nozzle hole of clean and fouled injectors is calculated from the images acquired using 3D Pythagoras theorem. Penetration lengths of fouled injector nozzles are shown to be longer than the same nozzles of the clean injector. Spray plumes emitted from the clean injector are shown to be wider and better atomized. In contrast, the fouled injector has thinner spray plumes with evidence of a liquid jet towards the centre, which suggests a spray plume with less atomization. To further understand the liquid mass that is discharged from each nozzle, the momentum flux is measured using a force sensor, which is used to provide a direct measure of the liquid mass (referred to as the impingement meter measurement method). The total mass injected by a clean injector is 6 to 8% higher than fouled injector.

The second half of the experimental investigations presented in this thesis investigates the fuel film formation on the plate by clean and fouled injectors. For these experiments, a well-known Laser Induced Fluorescence (LIF) methodology was used to measure fuel film distributions resulting from the spray impinging onto a quartz glass plate. For these experiments, LIF was performed to excite toluene (1% by volume), which was used as a fuel tracer within a non-fluorescing iso-octane base. A simplified experimental setup was constructed for which the spray from the injector impinged onto a quartz glass plate placed at a 50 mm distance from the injector. The average film thickness for both the injectors lies between 0 to 4 μm with clean injector film occupying a larger area. Mass deposited by each nozzle of clean and fouled injector is measured, and this thesis describes its dependency on multiple factors (i.e., total mass injected by nozzle hole, penetration length, impingement distance (D), and excess penetration length). Fouled injector nozzles form thick films with less area occupied by the films. These films take longer time to evaporate. The remaining mass on the surface for fouled injector fuel films at the end of the temporal sequence (800ms ASOI) is higher than clean injector films due to the faster evaporation of clean injector films.