Abstract
Stochastic pre-ignition remains one of the major barriers limiting further engine downsizing and down-speeding; two widely used strategies for improving the efficiency of spark-ignited engines. One of the most cited mechanisms thought to be responsible for pre-ignition is the ignition of a rogue droplet composed of lubricant oil and fuel. This originates during mixture formation from interactions between the fuel spray and oil on the cylinder liner. In the present study, this hypothesis is further examined using a single cylinder supercharged engine which employs a range of air-fuel mixture formation strategies. These strategies include port-fuel injection (PFI) along with side and central direct injection (DI) of an E5 gasoline (RON 97.5) using single and multiple injection events. Computational fluid dynamic (CFD) calculations are then used to explain the observed trends. Overall, this study reinforces that interactions between the fuel spray and oil on the cylinder liner can be an important contributor towards stochastic pre-ignition. The occurrence of pre-ignition, as shown by CFD calculations, is successful after completion of two stages. The first stage involves the formation of precursors from interactions between the fuel spray and oil on the cylinder liner. This is shown to be dependent upon the mass of the fuel impinging on the cylinder liner. The second stage involves the ignition of the precursor, which is shown to be dependent upon the temperature of the air-fuel mixture near top dead center.
| Original language | English (US) |
|---|---|
| Title of host publication | SAE Technical Paper Series |
| Publisher | SAE International |
| DOIs | |
| State | Published - Sep 10 2018 |
Bibliographical note
KAUST Repository Item: Exported on 2020-10-01Acknowledgements: The authors wish to acknowledge funding from the Clean Combustion Research Center at King Abdullah University of Science and Technology and Saudi Aramco under the FUELCOM II program. The authors also thank Nimal Naser for discussions on the project. The simulations utilized the KAUST supercomputing facility provided by KASUT Information Technology Service. We thank Convergent Science for providing the CONVERGE licenses.