Experimental and machine learning analysis of injector hole diameter effects on performance and emissions in an ammonia-biodiesel dual-fuel diesel engine
摘要
Unreliability of fossil fuel resources and environmental consequences of its use necessitate finding alternative energy sources, free of carbon emission. In diesel engines, atomization is very important for proper functioning; specifically, injector hole diameter affects the size and penetration length of droplets produced. In this paper, we analyze the use of biodiesel produced from waste cooking oils as the main fuel along with ammonia as the pilot fuel in the dual-fuel diesel engine. The experiments were performed with injector hole diameters of 0.26 to 0.34 mm and five types of fuel mixtures: D100 (100% diesel), A10, A20, A30, and A40 (10 to 40% ammonia mixed with biodiesel). According to data collected at 100 points of engine operation, the best results were recorded for the A20 fuel mixture at an injector hole diameter of 0.26 mm with IP = 6.39 kW and BTE = 35.3%, and NOx = 3 g/kWh. The superior performance is attributed to finer atomization at smaller injector diameters, which enhances air–fuel mixing and combustion efficiency, while the moderate ammonia fraction reduces CO₂ and NOx without compromising ignition stability. To support the experiment, three machine learning algorithms such as Random Forest, Gradient Boosting, and Support Vector Regression were used to train and validate models on the basis of this dataset. Prediction accuracy of these models reached up to 99%. The high correlation of the results proves the validity of findings and indicates promising potential of ammonia-biodiesel blends.