Modeling and Grey Relational Multi-response Optimization of Chemical Additives and Engine Parameters on Performance Efficiency of Diesel Engine

Abstract

Singular optimization of engine conditions for better engine performance have been studied extensively. However, in the practical sense, more than one performance characteristics are essential in the optimization of engine conditions. The current study investigates the effect, optimization, and modeling of engine conditions on multi-characteristics of a single cylinder-dual direct injection-water cooled diesel engine with the help of Taguchi-grey relational and regression analyses. The engine conditions employed are engine load, hydrogen, multi-walled carbon nanotubes (MWCNTs), ignition pressure, and ignition timing, at four different levels. The engine performance characteristics analyzed were brake thermal efficiency (BTE), brake specific fuel consumption (BSFC), hydrocarbons (HC), nitrogen oxide (NOx), carbon monoxide (CO), and carbon dioxide (CO2). The results showed that there was a similar behavioral pattern of the effect of engine conditions on engine performance, except for ignition timing. The optimal settings for better engine performance were obtained at 25% engine load, 20% hydrogen, 50 ppm MWCNTs, 220 bar ignition pressure, and 21 obTDC ignition timing. Interestingly, the discovered optimal did not fall within the considered experimental runs, however, the predicted optimal engine performance was within 95% confidence bounds. It is recommended that the experimental work based on the obtained optimal settings should be conducted to elucidate the efficacy of the confirmation analysis. The analysis of variance showed that the engine load was the most significant factor on the overall engine performance, having a contribution of 71.47%, followed by hydrogen and MWCNTs. Also, the ignition pressure and timing were not significant on the overall engine performance, which showed a need to place more attention on the significant factors for better engine performance. The mathematical and graphical modeling showed the efficacy of the design analysis, while the interaction plots showed broader detailed factor settings for better engine performance.