Numerical optimization of design and fuel factors and development of a statistical model for the emission control of DI CI engine

Abstract

In the present work an attempt was made to determine the most influencing factor through numerical optimization and also to develop a statistical model for the NOx and smoke control of a DI CI engine with less compromise on engine performance. Seven numeric factors (retarded injection timing, injection pressure, percentage of exhaust gas recirculation, percentage of biodiesel, Butanol, Proponol and Pentanol) and one categorical factor (piston geometry) were selected as the factors for the investigation. Experiments were designed through Taguchi orthogonal array and a single cylinder stationary engine was chosen for conducting the experiments. Results of the experiments were fed into the input for the statistical analysis. NOx, smoke and brake specific fuel consumption (BSFC) were chosen as the responses for the present work. Analysis of variance (ANOVA) and Half Normal plot methods were performed to investigate the factor effects on the chosen responses. A coded regression equation was also developed to predict the responses. From the analysis, it was observed that piston geometry and percentage of biodiesel in the blend have major effect when compared with the other factors tested in this work. Numerical optimization and Fit statistical results were compared with experimental results and toroidal piston geometry at 20 percentage of biodiesel in the blend were found as the best combination of factor levels for this work.