Alcohol

Fuel efficiency

Engineering

Materials science

Chromatography

Compression ratio

Internal combustion engine

Mean effective pressure

Response surface methodology

Volumetric efficiency

Organic chemistry

Gasoline

Thermal efficiency

Combustion

Automotive engineering

Chemistry

Brake specific fuel consumption

Spark-ignition engine

Fuel

In this study, effects of i-amyl alcohol/gasoline fuel blends on spark ignition (SI) engine performance and emissions were investigated experimentally, predicted by Artificial Neural Network (ANN) and optimized with Response Surface Methodology (RSM). Test engine was operated with pure gasoline and gasoline-isoamyl alcohol (isopentanol) fuel mixtures with different proportion at different engine speeds and various compression ratios (CR). With respect to obtained data from experiments, an ANN model, which is an Artificial Intelligence (AI) application, has been developed to estimate outputs such as brake mean effective pressure (BMEP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE), nitrogen oxides (NOx), hydrocarbon emission (HC) and carbon monoxide (CO) according to CR, fuel blending ratio (by vol.%) and engine speed (rpm). In addition, RSM was applied to find suitable engine operating conditions. According to results, the ANN model can estimate performance and emission parameters of engine by correlation coefficient (R2) between 0.94 and 0.99. It is seen that the max. mean relative error (MRE) is less than 7% compared with outcomes obtained from tests. The RSM study demonstrated that, i-AA ratio of 15% at 8.31 CR and 2957.58 rpm engine speed are the optimal engine operating parameters. In this way, the ANN model with RSM support was found to be an effective tool for predicting and optimizing engine outputs with minimum test.

Performance and Exhaust Emission Prediction of a SI Engine Fueled with I-amyl Alcohol-Gasoline Blends: An ANN Coupled RSM Based Optimization