Machine learning predictions for enhancing engine performance and emission using aluminum oxide nano additives in castor biodiesel

Nachippan, Murugu and Pathmanabhan, P. and Nagappan, Beemkumar and Upadhye, Vijay J. and Kaliappan, Nandagopal and Balaji, V. and Kamakshi Priya, K. (2025) Machine learning predictions for enhancing engine performance and emission using aluminum oxide nano additives in castor biodiesel. Scientific Reports, 15 (1). ISSN 2045-2322

[thumbnail of Ft 02.pdf] Text
Ft 02.pdf

Download (4MB)

Abstract

This study investigates the role of aluminum oxide nano-additives in enhancing the performance and reducing the environmental footprint of a B30 castor biodiesel blend in a compression ignition (CI) engine, emphasizing principles of green and sustainable chemistry. With global concerns over
emissions and the depletion of fossil fuels, there is an urgent need for cleaner and more efficient alternative fuels. Biodiesel derived from non-edible sources, such as castor oil, presents a sustainable solution, but improvements in its combustion efficiency and emissions profile are crucial for widespread adoption. In this research, aluminum oxide nanoparticles were incorporated into a B30 biodiesel blend to enhance combustion properties, reduce ignition delay, and significantly mitigate
harmful emissions, including carbon monoxide (CO), hydrocarbons (HC), and nitrogen oxides (NOx). The biodiesel was synthesized through the transesterification of castor oil, which is rich in ricinoleic acid, and tested on a Kirloskar diesel engine operating at a constant speed of 1500 rpm under varying load conditions. Results demonstrated that the nano-additive-infused biodiesel blend outperformed
conventional diesel in terms of fuel economy, atomization, vaporization, and overall combustion efficiency. Additionally, the use of aluminum oxide nanoparticles reduced brake-specific fuel consumption (BSFC) and pollutant emissions. To further optimize engine performance and minimize emissions, a machine learning framework was applied, comparing algorithms such as Random Forest and XGBoost. The analysis identified XGBoost as the most accurate predictive tool, offering valuable insights for optimizing engine parameters. This work highlights the application of green and sustainable chemistry through the integration of nano-additives and advanced data analytics to
develop cleaner, more efficient biodiesel fuels, ontributing to the global shift toward environmentally friendly energy solutions.

Item Type: Article
Subjects: Mechanical Engineering > MECH COMPOSITE
Divisions: Engineering > Mechanical Engineering
Depositing User: Unnamed user with email techsupport@mosys.org
Date Deposited: 04 Feb 2026 10:45
Last Modified: 04 Feb 2026 10:45
URI: https://ir.dsce.ac.in/id/eprint/97

Actions (login required)

View Item
View Item