Response Surface Methodology and Artificial Neural Network Methods for Optimization of Performance and Emission Characteristics of Diesel Engine Using Diesel-n-butanol Blend

Abstract

Due to the growing demand for more sustainable and environmentally friendly energy sources, extensive research has been done to reduce emissions and boost the efficiency of CI engines. These engines are known to contribute to two main environmental contaminants: PM and NOx. Among the other alternative fuels that have been researched, n-butanol has shown great potential due to its high oxygen content, considerable latent heat of vaporization, and rapid flame propagation rate, all of which contribute to more efficient and clean combustion. This study combines ANN and RSM techniques to optimize the emissions and performance characteristics of a CI engine running on diesel-n-butanol blends. Pure diesel (B0) and blends including 10%, 20%, and 30% n-butanol (B10, B20, and B30) were used in the experimental testing. An ultrasonicator was used to achieve homogeneous fuel preparation. B10 demonstrated the most balanced performance of all the fuels tested, greatly lowering CO2 and NOx while keeping braking torque (BT) and brake power (BP) on level with regular diesel. Higher blends produced incomplete combustion, which increased emissions of CO and HC, but they also CO2 and NOx. By maximizing BT and BP and minimizing 𝐵𝑆𝐹𝐶 and emissions, 𝑅𝑆𝑀 optimization-based central composite design was employed to determine the ideal operating parameters. BT of 4.18 Nm, BP of 1.52 kW, BSFC of 0.379 𝑘𝑔/𝑘𝑊ℎ, CO2 of 3.4277%𝑣𝑜𝑙, CO of 0.0318 %𝑣𝑜𝑙, NOx of 148 𝑝𝑚𝑚 , and HC of 26 𝑝𝑚𝑚 were the results of the ideal conditions, which were reached at 49.79% engine load and 10% n-butanol ratio. As demonstrated by better R2 and lower error measures (MSE and RMSE), comparative research showed that ANN offered more prediction accuracy than RSM. The nonlinear interactions between engine inputs and responses were successfully represented using ANN. In general, the study showed that using statistical modeling coupled with n-butanol can improve the performance of CI engines and lower emissions. This integrated strategy contributes to the development of sustainable alternative fuels and environmental conservation by providing an approach toward cleaner and more efficient engines.