Spark ignition engine performance prediction using artificial neural network fuelled with gasoline ethanol blends

Abstract

Fossil fuel its shortfall and its impact on environment has forced researchers to search for a renewable fuel which is friendly with environment. It has created the interest in the study and evaluation of different alternative fuels like alcohols. These fuels can be easily produced by the process of fermentation and distillation from biomass. This study based on experiments investigates feasibility of ethanol as a fuel for off-road SI engine applications. The analysis of ethanol-gasoline proportionate mixture (E0, E10, E20 and E30) were studied as well as compared for various Loads and CR (6, 8 and 10). The study was carried out on gasoline engine generator setup using electric dynamometer. It was observed that BSFC increases as ethanol content increases and varies inversely with an increase in compression ratio. Similar trends were observedfor different compression ratios. Artificial Neural Network (ANN) modeling was prepared. It was observed that ANN model is able to analyze the performance of engine. Result shows that ANN model was a useful tool in the prediction of BSFC.