The flexible fuel vehicles (FFV) are definitely a reality in Brazil. Most of the cars currently produced in Brazil are flexible fuel with fuel range from gasohol (E22) to hydrated ethanol (E100). All of these Brazilian applications are fuel sensor less. It means that the fuel inference is based on lambda sensor only. This fuel inference approach takes the intake air mass over the calculated mass of fuel delivered through the injectors in account to calculate the air fuel ratio (AFR) which points out the fuel blend in tank. One of the disadvantages of this inference strategy is that all the fuel considered for AFR calculation must come from injectors. Considering that the cold start and warm-up phases should operate in a richer mixture (choke operation), part of this fuel is delivered to the lubricant which is know as fuel dilution phenomenon in lubricant. According to the fuel properties and the lubricant temperature the diluted fuel evaporates and is mixed in intake air through blow-by system. This extra fuel source is considered in AFR calculation and can, in some conditions, distort the fuel blend inference. In order to prevent this blow-by vapour negative impact on flexible fuel inference logic a mathematical model was developed and calibrated for a 1.6 liter engine FFV application. Many test results with different blends and conditions are discussed. This study confirmed that this mathematical model presented positive results and some improvement opportunities were commented as well.

This content is only available via PDF.
You do not currently have access to this content.