An experimental database on the effects of single- and split injection strategies on spray formation and spark discharge in an optical direct-injection spark-ignition engine fuelled with gasoline, iso-octane and alcohols

Abstract

This paper presents results from a comprehensive optical study of a direct-injection spark-ignition research engine running on gasoline, iso-octane, ethanol, n-butanol and E10 fuels injected from a multi-hole injector located centrally in the combustion chamber. The analysis was based on images of spray development, spark discharge and combustion to understand the effect of injection strategy early and late in the cycle on in-cylinder phenomena. Specifically, 'single' injection strategies at different timings from early to late intake stroke, as well as multiple 'split' injection events with triple pulses in the early intake stroke or double pulses in the intake stroke and late compression stroke were investigated. The engine was run at 1500 RPM at part and full-load conditions (0.5 bar and 1.0 bar inlet plenum pressure, respectively). Engine coolant temperatures of 20-90 °C were employed to understand how the fuel's volatility was related to the phenomena observed. The sprays were imaged for a series of cycles primarily by Laser-sheet illumination on one vertical and two horizontal planes to identify three-dimensional aspects of the spray's development and its interactions with the incoming flow, valves, piston and liner. There was a clear trade-off between the earliest single injection timings in the intake stroke that reduced valve interactions but increased piston impingement and the latest single injection timings towards the end of the intake stroke which reduced piston impingement but increased valve interactions and resulted in less time for fuel evaporation and mixing before ignition timing. The triple injection strategy eliminated effects related to impingement. The spark discharge was also imaged for a series of cycles with all injection strategies and clear differences were observed. Selective combustion imaging provided insights into the flame's growth and motion with early and double early-late split injection strategies. The double early-late injection strategy demonstrated the potential for control of the mixture formation and flow field over the early flame development stage of combustion.3

Aspects of Numerical Modelling of Flash-Boiling Fuel Sprays

“…The majority of novel fuels and additives have high volatilities, which can influence the spray formation in a similar way to increasing fuel temperature or reducing ambient pressure. For example, the addition of ethanol to gasoline can have a significant influence on the spray formation due to an increased evaporation rate or flashing phenomena, even at low blending ratios [4]- [9]. With the inevitable increase in global consumption of new fuel blends, it is essential to understand the effect of flash-boiling on spray formation with a range of components.…”

“…The variation in volatility, as well as properties such as surface tension, specific heat and viscosity of each component, can influence the formation of the spray as single components can flash-boil individually and affect the behaviour of droplets. Previous experimental studies [9] whereby both single and multi-component fuels were injected into a quiescent chamber at superheated conditions, have shown that high volatility single components like n-pentane can be used to represent the behavior of gasoline since it is these components within gasoline's multicomponent blend that drive the spray's collapsing mechanism.…”

“…A large database of spray images is available with this injector with various fuels for validation purposes (including gasoline, alcohols, high and medium volatility hydrocarbon components, etc. ), both in quiescent injection chambers and in-cylinder [9], [36]. A constant mass flow rate of 20 g/s was used throughout the test cases of the current study.…”

Flame front analysis of ethanol, butanol, iso-octane and gasoline in a spark-ignition engine using laser tomography and integral length scale measurements

“…capturing one flame image per crank-angle degree at 1500 RPM. Details of the system have been described in previous studies and not repeated here for brevity [13,20,21,[47][48][49][50][51]. The fast shutter speed reduced background noise but also led to low combustion luminosity levels requiring small thresholding values for processing.…”

“…Pressure data were post-processed to calculate the Indicated Mean Effective Pressure (IMEP), amplitude and timing of peak in-cylinder pressure, including mean values and Coefficients of Variation (COV=Mean/RMS). Heat release analysis and calculation of Mass Fraction Burned (MFB) was performed using methods based on [58,59] and followed practices of earlier work by the current authors for consistency [13,20,21,[48][49][50][51]. All uncertainties involved in acquiring and processing in-cylinder pressure data were carefully considered according to [60].…”