Investigating the impact of fatty acid profiles on biodiesel lubricity using artificial intelligence techniques

Abstract

Biodiesel lubricity is a crucial factor influencing engine performance and longevity, primarily determined by its fatty acid composition. This study evaluates the tribological properties of biodiesel derived from 15 different feedstocks using High-Frequency Reciprocating Rig (HFRR) tests, 3D-laser microscopy, Scanning Electron Microscopy (SEM), and Energy-Dispersive X-ray Spectroscopy (EDS). The results indicate that biodiesel with higher unsaturation levels, particularly those rich in monounsaturated and polyunsaturated fatty acids, exhibits superior lubricity, characterized by reduced wear scar diameters and enhanced film formation. Conversely, biodiesels with high saturated fatty acid content demonstrate larger wear scar diameters and lower film formation efficiency, leading to increased friction and wear. To further analyze the impact of fatty acid composition on lubricity, an artificial intelligence (AI)-based approach using the Adaptive Boosting (AdaBoost) algorithm was implemented. The AI model effectively predicts wear scar diameter, friction coefficient, and film formation, providing insights into the complex interactions between fatty acid profiles and tribological performance. Feature importance analysis and sensitivity evaluation reveal that polyunsaturated fatty acids significantly enhance lubricity, while an optimal balance between saturated and unsaturated fatty acids is necessary to achieve stable frictional behavior. These findings emphasize the potential of AI-driven predictive modeling as a cost-effective tool for optimizing biodiesel lubricity, reducing the need for extensive experimental trials. The integration of advanced tribological testing and AI analysis offers a deeper understanding of biodiesel's lubrication mechanisms, supporting the development of high-performance, sustainable biofuels.