In order to reduce both polluting emissions and fuel costs, many countries allow mixing ethanol to gasoline either in fixed percentages or in variable percentages. The resulting fuel is labeled E10 or E22, where the number specifies the ethanol percentage. This operation significantly changes way the stoichiometric value, which is the air-to-fuel mass ratio theoretically needed to completely burn the mixture. Ethanol concentration must be correctly estimated by the Engine Management System to optimally control exhaust emissions, fuel economy and engine performance. In fact, correct fuel quality recognition allows estimating the actual stoichiometric value, thus allowing the catalyst system to operate at maximum efficiency in any engine working point. Moreover, also other essential engine control functions should be adapted in real time by taking into account the quality of the fuel that is being used. An example is the Spark Advance management, which may benefit from correct fuel quality recognition for example in terms of knock detection and control. Many possible solutions are currently available in mass production vehicles to evaluate the percentage of ethanol in fuel, all of them ranging from indirect measurements based on the use of standard lambda sensors, to the use of dedicated sensors to be installed in the fuel line. The present work shows an innovative solution to the problem of detecting ethanol percentage in fuel, based on the analysis of the acoustic signal measured by installing a microphone in the engine compartment. The paper initially presents the main criteria to choose the best sensor as a trade-off between cost and performance. Then the signals coming from sensors installed in different positions with respect to the engine will be analysed to demonstrate how to perform the best choice. Finally, it will be shown that starting from signal frequency spectrum analysis it is possible to develop a fuel quality recognition index based on combustion frequency amplitude and its higher-order harmonics. The index is then compared with a threshold depending on the actual working point, to detect whether the engine is running on gasolina (E22) or pure ethanol (E100).
Ethanol to Gasoline Ratio Detection via Time-Frequency Analysis of Engine Acoustic Emission
CAVINA, NICOLO';MORO, DAVIDE;
2012
Abstract
In order to reduce both polluting emissions and fuel costs, many countries allow mixing ethanol to gasoline either in fixed percentages or in variable percentages. The resulting fuel is labeled E10 or E22, where the number specifies the ethanol percentage. This operation significantly changes way the stoichiometric value, which is the air-to-fuel mass ratio theoretically needed to completely burn the mixture. Ethanol concentration must be correctly estimated by the Engine Management System to optimally control exhaust emissions, fuel economy and engine performance. In fact, correct fuel quality recognition allows estimating the actual stoichiometric value, thus allowing the catalyst system to operate at maximum efficiency in any engine working point. Moreover, also other essential engine control functions should be adapted in real time by taking into account the quality of the fuel that is being used. An example is the Spark Advance management, which may benefit from correct fuel quality recognition for example in terms of knock detection and control. Many possible solutions are currently available in mass production vehicles to evaluate the percentage of ethanol in fuel, all of them ranging from indirect measurements based on the use of standard lambda sensors, to the use of dedicated sensors to be installed in the fuel line. The present work shows an innovative solution to the problem of detecting ethanol percentage in fuel, based on the analysis of the acoustic signal measured by installing a microphone in the engine compartment. The paper initially presents the main criteria to choose the best sensor as a trade-off between cost and performance. Then the signals coming from sensors installed in different positions with respect to the engine will be analysed to demonstrate how to perform the best choice. Finally, it will be shown that starting from signal frequency spectrum analysis it is possible to develop a fuel quality recognition index based on combustion frequency amplitude and its higher-order harmonics. The index is then compared with a threshold depending on the actual working point, to detect whether the engine is running on gasolina (E22) or pure ethanol (E100).I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
