Experimental investigation and Artificial Neural Network modelling of combustion pressure parameters of dual spark plug SI engine

Abstract

Shortages of petroleum-based fuels due to their rapid consumption rate and their environmental degradation motivates researchers to develop new technologies for the efficient use of petroleum-based fuels as well as the use of environmentally friendly alternate biofuels. Ethanol is one of the promising alternative biofuels for spark ignition (SI) engines. Currently, the Indian union Government aims to mix 20% ethanol by 2030. Therefore, extensive research studies are needed to determine the combustion stability of E20 blend. The purpose of this study was to investigate the combustion pressure fluctuations is dual plug SI engine using ethanol-gasoline blends and to develop Artificial neural network model using back propagation algorithm for the prediction of these fluctuations. The fluctuations of Indicated mean effective pressure (IMEP), maximum cylinder pressure and maximum rate of pressure rise has been measured by conducting experiment in a dual spark plug SI engine under different compression ratios and spark ignition timings using E5, E10, E15 and E20. Then, these experimental data’s have been used for training and testing of BP model to predict the combustion pressure parameters. The % load, blend ratio, compression ratio and spark timings were used as the input parameters whereas IMEP, maximum cylinder pressure and maximum rate of pressure rise were used as the output parameters of the model. The experimental results showed that E20 achieved maximum average IMEP and minimum COVIMEP at a compression ratio of 10:1 and ignition timings of 24° bTDC. The modelling results showed that BP model is capable of predicting the combustion pressure fluctuations with good accuracy.