The investigation of optimum working parameters of spark ignition engines using ethyl alcohol-gasoline blend

THE INVESTIGATION OF OPTIMUM WORKING PARAMETERS OFSPARK IGNITION ENGINES USING ETHYL ALCOHOL-GASOLINEBLEND(Ph.D. Thesis)Tolga TOPGÜLGAZI UNIVERSITYINSTUTE OF SCIENCE AND TECHNOLOGYJune 2006ABSTRACTAs a renewable energy source for the spark ignition engines, ethyl alcohol gainsimportance recently. In this study, the effects of using unleaded gasoline (E0)and unleaded gasoline-ethanol blends (E10, E20, E40 and E60) on engineperformance, exhaust emissions, heat losses and cylinder pressures haveinvestigated experimentally. The experiments were conducted on a Hydra singlecylinder, four-stroke, variable compression ratio, spark ignition and fuelinjection engine. The tests were performed by varying the compression ratio,ignition timing, excess air coefficient and inlet air temperature at differentengine speeds and loads. Maximum brake torque spark timing of the engineobtained with blends, showed no significant change relative to unleaded gasolineunder low compression ratios. The engine performance improved depending onthe ethanol percentage especially at high compression ratios and low enginespeeds. At maximum brake torque spark timing, the effects of air/fuel ratio andinlet air temperature on the variations of engine performance and exhaustemissions are similar with all fuels. Heat losses decreased used with ethanol-gasoline blends a compared to unleaded gasoline. The highest cylinder pressurewas obtained using E60, which has the best anti-knock characteristics amongtest fuels. An artificial neural network (ANN) is used for prediction of engineperformance. The training data for ANN is obtained from experimentalmeasurements. It was found the R2 values are 0,999569 and 0,999717 for theengine torque and BSFC respectively. The results show that ANN providesacceptable predictions in the parameters of the engine.Science Code : 708.3.026Key Words : Unleaded gasoline-ethanol blends, alternative fuel, sparkignition engine, engine performance, exhaust emissionsPage Number : 195Adviser : Assoc. Prof. Dr. H. Serdar YÜCESU