Experimental investigation and AI-based prediction of engine performance and emissions using CeO2-enhanced biodiesel blends

Ganesh K; D K Ramesha; Kapilan Natesan; Norkhairunnisa Mazlan; Sadashiva Prabhu S

doi:10.1038/s41598-025-28932-4

Experimental investigation and AI-based prediction of engine performance and emissions using CeO₂-enhanced biodiesel blends

Sci Rep. 2025 Dec 13;15(1):43861. doi: 10.1038/s41598-025-28932-4.

Authors

Ganesh K¹, D K Ramesha², Kapilan Natesan¹, Norkhairunnisa Mazlan³, Sadashiva Prabhu S⁴

Affiliations

¹ NITTE (Deemed to be University), Nitte Meenakshi Institute of Technology, Yelahanka, Bengaluru, Karnataka, 560064, India.
² Department of Mechanical Engineering, University of Visvesvaraya College of Engineering, K R Circle, Bengaluru, Karnataka, 560001, India.
³ Department of Aerospace Engineering, Faculty of Engineering, Universiti Putra Malaysia (UPM), 43400, Serdang, Selangor, Malaysia.
⁴ Department of Mechanical and Industrial Engineering, Manipal Institute of Technology, Manipal Academy of Higher Education, Manipal, Karnataka, 576104, India. ss.prabhu@manipal.edu.

Abstract

The global depletion of fossil fuel resources and growing environmental concerns have driven the pursuit of sustainable alternative energy sources, with biodiesel emerging as a promising option. In the present work, experimental and AI prediction studies examine the combustion, performance, and emissions of a B20-blended fuel engine. The biodiesel was derived from used temple oil, and CeO₂ nanoparticles (NPs) were used as additives. The experiments conducted at a constant engine speed with various loads reveal that CeO₂ NPs at a 100-ppm concentration significantly improve brake thermal efficiency (BTE). The specific fuel consumption decreases with increasing CeO₂ NPs concentration, indicating enhanced fuel efficiency. The higher cylinder pressure and net heat release (NHR) further highlight the improved combustion characteristics. The emission analysis reveals a notable decline in hydrocarbons, carbon monoxide, and nitrogen oxides levels, with CeO₂ NPs at a 100-ppm concentration achieving the lowest emissions due to superior fuel atomization, improved combustion efficiency, and enhanced thermal management facilitated by the presence of CeO₂ NPs. The B20 blend with CeO₂ NPs at a 100-ppm concentration demonstrated superior performance metrics at maximum load, as evidenced by a 1.57% increase in BTE, a 2.83% rise in cylinder pressure, and a 5.95% increase in NHR compared to conventional diesel. A significant reduction in pollutant emissions, such as carbon monoxide, decreased by 66.67%, hydrocarbons by 13.51%, and nitrogen oxides by 6.04% relative to the B20 blend. The validation of experimental data using a random forest regressor model demonstrates strong predictive accuracy, characterized by low mean squared errors and high R² scores across various performance metrics. Overall, the findings of this work confirm that the B20 blend with CeO₂ NPs is a sustainable and efficient replacement for traditional diesel fuel.

Keywords: Biodiesel; Engine tests; Nanoparticles; Prediction; Random forest regressor model.