Experimental Investigation and Artificial Intelligence-Based Modeling of Novel Biodiesel Fuels Containing Hybrid Nanoparticle Additives

Abstract

This work investigates the influence of hybrid NiO-SiO2 nanoparticles on the engine behavior of biodiesel derived from waste sunflower oil and evaluates the experimental outcomes using a data-driven modeling approach. Biodiesel was produced via transesterification and doped with nanoparticles at concentrations of 50, 75, and 100 ppm. Performance and emission tests were conducted on a single-cylinder diesel engine operating at constant speed under varying loads. Specific fuel consumption, brake thermal efficiency, CO, HC, NOx, smoke opacity, and exhaust gas temperature were recorded and analyzed. The incorporation of nanoparticles improved combustion quality and contributed to substantial reductions in harmful emissions. The WSOB20 blend containing 100 ppm NiO-SiO2 provided the most balanced results, decreasing CO, HC, and smoke emissions by 39.50%, 39.40%, and 35.20%, respectively, relative to diesel fuel, while preserving competitive thermal efficiency. A linear regression model developed for CO prediction produced a low mean squared error (1.08 & times; 10(-5)), indicating strong predictive capability. The findings confirm that hybrid nanoparticle additives can enhance biodiesel performance while supporting accurate emission forecasting.