Optimizing Palm-Based Bio-ⅼubricant Formulations for Diesel Engine Using Machine Learning and Experiment Techniques

Integrating bio-lubricants into internal combustion engines is crucial for sustainable engineering, driven by the need for renewable and eco-friendly alternatives. However, bio-lubricants in diesel engines often face challenges related to insufficient thermal-oxidative stability. This research introduces a novel approach to enhance the thermal-oxidative stability and lubricity of palm oil-derived bio-lubricants for diesel engine applications. By chemically modifying palm olein through transesterification, epoxidation, and oxirane ring-opening, and leveraging the predictive capabilities of the 'BimolP' machine learning tool, we identified base stocks with improved properties. Our results show significant enhancements in thermal-oxidative stability, with the decomposition onset temperatures (Tonset) of modified palm olein increasing from 280°C to 281°C and 295°C, and those of palm methyl ester derivatives rising from 222°C to 245°C and 268°C. Additionally, the friction coefficient of palm olein derivatives increased from 0.104 to 0.106 and 0.113, and for methyl ester derivatives, it rose from 0.101 to 0.108 and 0.113. The final bio-lubricant formulation combines various modified base stocks to meet SAE-30 and SAE-40 diesel lubricant specifications. This study underscores the effectiveness of integrating chemical modifications, machine learning predictions, and blending palm-based stocks to optimize bio-lubricant formulations without the need for prior synthesis or modification of palm olein and palm methyl ester