Dual Effects of N-Butanol and Magnetite Nanoparticle to Biodiesel-Diesel Fuel Blends as Additives on Emission Pattern and Performance of a Diesel Engine with ANN Validation

This paper investigates impact of magnetite dispersed in butanol and added to two varied blends of palm biodiesel and diesel (B20 and B30). The developed fuel samples were characterized and tested on single cylinder diesel Yanmar engine (L70N) to observe engine behavior for emissions and performance. Results are compared with two reference fuels: YF50 fuel contains 50 ppm magnetite in B20 and B(n)10Y90 contains 10% butanol with 90% B20. Addition of magnetite and butanol depletes emissions levels and improve performance compared to ordinary B20 and B30 however; samples with higher dosage of magnetite (150 ppm) yielded better results in performance and emission compared with lower dosage (75 ppm). The best sample was C10Z90 which entails 150 ppm magnetite in butanol added at 10% to B30. Brake thermal efficiency (BTE) at highest brake power (BP) point for C10Z90 was 37.28% compared to others (32.88%, 35.22% and 35.96%). Additionally, brake specific fuel consumption (BSFC) of C10Z90 was at least 8.29 g/Kw.hr and at most 84.52 g/Kw.hr less than other samples at highest BP point. Results indicated C10Z90 was lower in carbon-monoxide, hydrocarbon and smoke except for oxides of nitrogen. Artificial Neural Network (ANN) model successfully predicted BTE, BSFC and emissions of the dual fuel application.

Automated summary

This study investigates the impact of magnetite dispersed in butanol and added to blends of palm biodiesel and diesel on engine behavior for emissions and performance. Results show that the addition of magnetite and butanol reduces emissions and improves performance compared to ordinary blends, with higher dosage of magnetite yielding better results. The best sample was found to have 150 ppm magnetite in butanol added at 10% to B30, which showed higher brake thermal efficiency and lower fuel consumption than other samples. An artificial neural network model successfully predicted the performance and emissions of the dual fuel application.