Performance characteristics and exhaust emissions of a single-cylinder diesel engine for different fuels: Experimental investigation and artificial intelligence network

In the current work, as the first phase, the main emphasis deals with the experimental study of the performance characteristics (effective power P_eff and exhaust gas temperature T_exh) and exhaust emissions (including CO, CO₂, HC, and NOx) of a 4-stroke single-cylinder diesel (SCD) engine. The data are extracted from the engine tested for two inlet air temperatures (T_air), and three coolant temperatures (T_coolant) at five different speeds (n_m). To make the investigations even more comprehensive, all the tests are repeated for three different fuels including diesel (D100), diesel–kerosene blend (D95K5), and diesel–water blend (D90W10). In the second phase, an artificial intelligence network (AIN) is trained for expanding the outputs of the experiments. Analyzing the experiments’ outputs revealed that increasing T_air leads to three significant improvements including: reduction of emissions, shortening the ignition delay, and prevention of the quenching phenomenon. This is while, T_coolant has a slight effect. Also, results illustrated that the superiority of D90W10 over the two other investigated fuels in the improvement of engine performance and exhaust emissions. So that, almost 1.67% and 0.97% more power is available on average using by D90W10 blend compared to the D100 fuel and D95K5 blend, respectively. The findings of the AIN showed that the developed model is capable to estimate the engine performance and diesel exhaust emissions (DEE) with a correlation coefficient of 0.99921, 0.99952, 0.93959, 0.96980, 0.95826, and 0.99746 for P_eff, T_exh, CO, CO₂, HC, and NOx, respectively.