Numerical Studies of In-Cylinder Combustion and Soot Emission Characteristics of Biodiesel Fuels from Different Feedstock

Abstract

Multi-dimensional computation of in-cylinder biodiesel combustion and soot emission characteristics is performed using a commercial Computational Fluid Dynamics (CFD) software, ANSYS FLUENT 13, which is coupled to a chemical kinetic model via a plug-in chemistry solver, CHEMKIN-CFD. To represent the oxygenated straight chain hydrocarbon (HC) of biodiesel, a combined mechanism is utilised by employing the reduced mechanisms of two surrogate fuels, namely methyl butanoate (MB) and nheptane. The biodiesel fuel types considered in this study include the methyl esters of soy (SME), palm (PME) and coconut (CME), which are modelled by employing the respective thermophysical properties of the fuel and through variation of ratio between MB and n-heptane for accurate representation of its oxygen content. Diesel combustion is also simulated to serve as a benchmark for the study. The computational results indicate that diesel has better combustion efficiency at low engine power with the highest peak of premixed combustion (PMC) and the earliest end of combustion (EOC). As engine power increases, the heat released during PMC is lowered by 30% and the combustion duration are lengthened comparatively. In contrast, biodiesel combustion improves at high power, with CME having the most significant effect whereby the peak heat released rate during PMC improves and the EOC advances to be the earliest among all test fuels. The short HC chain length of CME exerts a more significant effect on the combustion. Comparisons between SME and PME which have similar ignition delays highlight that a high level of unsaturation improves the combustion at low engine power. This effect lessens with increased engine power. At high engine power, the end of PMC and the EOC for SME and PME are similar. The highest level of soot emission is observed for diesel at all levels of engine power followed by SME, PME and CME accordingly.