Performance of petrodiesel and biodiesel fuelled engines: a fundamental study of physical and chemical effects

Abstract

In this work, biodiesel and petrodiesel combustion is studied under conditions that represent those in an engine at top-dead-centre. The primary focus of this study is on improving the understanding of biodiesel feedstock properties on spray structure, understanding the effect of strain on soot formation in biodiesel and petrodiesel combustion using a kinetics-based soot model, developing a simplified soot model that can model soot formation in both biodiesel and petrodiesel combustion, and applying the model to study soot formation in sprays. The differences in feedstock properties primarily affect the liquid phase penetration. It is shown that liquid penetration is influenced by entrainment rate, vapour pressure, and the average droplet size, in decreasing order of influence. The vapour-phase penetration and mixture fraction distribution in the sprays are not significantly influenced by the changes in feedstock properties. Kinetic mechanisms for the oxidation of surrogate fuels for biodiesel and diesel and for soot formation are employed in the study. A one-dimensional flamelet code is employed to investigate the response of the soot formation to changes in scalar dissipation rate. The soot formation in biodiesel combustion is found to be more sensitive to changes in scalar dissipation rate. This suggests that increasing turbulence in a biodiesel-fuelled engine is likely to have a greater impact on soot emissions than in a petrodiesel-fuelled engine. Through a reaction pathway analysis, it is found that the differences in soot are on account of differences in the concentration of the aromatic species. Critical kinetic pathways and important species responsible for soot formation are identified for the fuels. Having identified the critical species and pathways, a semi-empirical two-equation soot model is developed to model soot in both hydrocarbon diesel and biodiesel combustion. Results from the kinetic soot formation model are employed to calibrate the constants of the semi-empirical model. To the best knowledge of the author, this is the first soot model formulated that can model soot formation in the combustion of both fuels. The semi-empirical model is implemented in an in-house Reynolds-averaged Navier Stokes (RANS) multi-dimensional spray code and employed to predict soot in biodiesel and diesel sprays. The computed spray results are compared with available measurements in the literature. Compared to the performance of another well-validated semi-empirical two-equation soot model, the soot model developed in this work is shown to better predict soot in both biodiesel and diesel sprays.