20th Australasian Fluid Mechanics Conference Perth, Australia 5-8 December 2016 Preliminary Investigation of Propane Spray Structure in an Optically Accessible Direct-Injection, Spark-Ignition Engine H.B. Aditiya 1 , J.S. Lacey 1 , F. Poursadegh 1 , M.J. Brear 1 , R.L. Gordon 1 , C. Lakey 2 , S. Ryan 2 and B. Butcher 2 1 Department of Mechanical Engineering University of Melbourne, Victoria 3010, Australia 2 Ford Motor Company of Australia Abstract Liquefied petroleum gas (LPG) has attained a significant posi- tion in various markets throughout the world, as it has numerous benefits as an automotive fuel in spark-ignition (SI) engines. Fuelling a direct injection (DI) engine with LPG, rather than conventional, liquid fuels, has the potential to improve fuel ef- ficiency and emissions output. However, there is little informa- tion pertaining to the operation of LPG in such vehicles. There- fore, this study aims to begin to characterise DI propane sprays, acting as a surrogate for LPG, in a DISI engine at injection tim- ings corresponding to both an early, homogeneous strategy and a late, stratified strategy. Fuel spray structures are imaged us- ing a production DISI engine modified to have optical access through a fused silica cylinder liner. Using Mie-scattering to resolve the liquid phase injection, propane sprays are observed to be heavily flash-boiling at both injection conditions. This is in contrast to heavier, liquid fuels where flash-boiling gener- ally occurs only for chamber conditions well below atmospheric pressure. Introduction Modern SI engine research is focused on the improving fuel ef- ficiency and reducing the emissions output of the engine. In order to achieve these goals a combination of innovations and advanced technologies will be required. DI is an internal com- bustion strategy that offers several known benefits. Compared with port fuel injection (PFI), DI engines improve fuel econ- omy due to a more precise metering of the air-fuel ratio. Ad- ditionally, DI is not subject to intake manifold wall-wetting as is PFI, and DI has a thermodynamically beneficial cooling ef- fect on the in-cylinder charge. This charge cooling effect helps suppress autoignition of the fuel charge, and allows for higher compression ratios, as well as lower heat losses to the cylinder walls [4, 8]. Some DI implementations have as much as 20 % higher fuel economy than PFI equivalents [3]. Moreover, a DI strategy enhances the fuel injection transient response with re- spect to the crank angle position [11]. This transient response and the ability for accurate fuel placement within the combus- tion chamber improve cold-start emissions [2] and CO 2 output is decreased by virtue of the increased fuel efficiency of a DI engine. In contrast with the previous emissions regulations, Euro 6b leg- islates not only the particulate mass production limit, but also the particulate number production. A further emission restric- tion is proposed in Euro 6c effective in September 2017 and this will likely require additional advances in DI technology [1]. One potential solution to comply with this proposed legislation is to use an alternative fuel, such as LPG. LPG is an attractive alternative fuel for SI engines, as it offers a higher octane num- ber and more knock resistance than conventional pump gaso- line. LPG also has the benefit of producing less energy-specific CO 2 emissions, and it has less tendency to produce particu- late emissions compared to heavier, liquid fuels. A thorough study by Krieck et al. [5] supports the idea, as they found that LPG fuel produces a negligible amount of soot and around 33 % lower hydrocarbon emissions compared with traditional liquid fuel (gasoline) in all engine conditions. LPG is also relatively less expensive than gasoline even when accounting for density differences, making the fuel financially attractive to consumers. There are several obstacles that must be overcome in order to realise the potential benefits of DI LPG in production vehicles. One of the challenges in using LPG is that its composition is not standardised and varies in different countries. At this point in time, there is limited literature regarding the performance of LPG in SI engines, and a dearth of information about DI LPG fuel delivery and spray mechanisms. As the implementation of a DI system in an engine is strongly dependent on a compre- hensive understanding of the fuel spray, the lack of knowledge regarding DI LPG currently prevents the commercial viability of this technology. Some of the first optical investigations of DI propane (used as a surrogate for LPG) at engine-like conditions was conducted by Lacey et al. [6] in a constant volume chamber. In this study, a range of chamber pressures and temperatures representative of GDI cylinder conditions was explored using DI propane in an experimental GDI injector. The results indicated that the spray structure of propane (used as an LPG surrogate) exhibits significantly more variability than that of iso-octane (used as a surrogate for gasoline) throughout a range of potential combus- tion chamber conditions. Because of its high vapour pressure, propane fuel sprays are subject to severe flash-boiling through- out the majority of chamber conditions corresponding to the GDI operating range, whereas heavier fuels like iso-octane tend to flash-boil only in a narrow range of chamber pressures well below atmospheric pressure. The similar result patterns are also observed in an earlier study of LPG spray behaviour at different back pressures by Mesman and Veenhuizen [7] from investigat- ing propane spray behaviour in a constant-volume cylindrical chamber. In different engine back pressures, they found that propane spray penetration decreases following the increase the back pressures. Flash-boiling spray developments were also ex- hibited in their results as the back pressure decreases to atmo- spheric pressure, due to the pressure difference between back pressure and rail pressure. Spray angle was also the parameter of interest in their study, and they observed that it is only af- fected by the back pressure; variation in fuel injection pressure does not exhibit any correlation with the spray angle. Regard- less the very few studies that have been performed to charac- terise the DI propane spray, to date none has contributed to the DI propane characterisation study in an actual, practical DI en- gine. Thus, despite the promising benefits of an LPG-fuelled, DISI engine, more study of the fuel delivery mechanisms are required in light of the high degree of structural variability in propane sprays. Therefore, this study is intended as an initial investi- gation of how DI propane spray is expected to behave in the