Analysis of Fuel Behavior in the Spark-Ignition Engine Start-Up Process 950678

An analysis method for characterizing fuel behavior during spark-ignition engine starting has been developed and applied to several sets of start-up data. The data sets were acquired from modern production vehicles during room temperature engine start-up. Two different engines, two control schemes, and two engine temperatures (cold and hot) were investigated. A cycle-by-cycle mass balance for the fuel was used to compare the amount of fuel injected with the amount burned or exhausted as unburned hydrocarbons. The difference was measured as “fuel unaccounted for”. The calculation for the amount of fuel burned used an energy release analysis of the cylinder pressure data. The results include an overview of starting behavior and a fuel accounting for each data set

Overall, starting occurred quickly with combustion quality, manifold pressure, and engine speed beginning to stabilize by the seventh cycle, on average. To facilitate rapid starting at cold engine conditions, approximately five times the amount of fuel required for a stoichiometric mixture is injected during the first one or two cycles. A large portion of this fuel, equivalent to nearly ten injections at stoichiometric idle conditions, remains “unaccounted for” after ten cycles of this analysis. About 10% of the fuel injected during the initial overfueling that is “unaccounted for” at first, shows up later in underfueled cycles as burned fuel or as hydrocarbon emissions. Similar trends occurred with both engines, and start-up strategies, although, during warm engine start-up conditions the overfueling is only 130% of stoichiometric and the mass “unaccounted for” after ten cycles represents only one injection at idle. The most successful start-up strategies that were analyzed injected close to the stoichiometric requirement for each cycle after the initial overfueling. The stoichiometric requirement for a particular cycle is directly proportional to the manifold pressure at a given temperature. It is recommended that methods for using manifold pressure in start-up strategies be investigated.