A computational study of pre-ignition to detonation transition in a one-dimensional chamber

Sow, Aliou; Jaasim, Mohammed; Hernandez Perez, Francisco; Im, Hong G.

A computational study of pre-ignition to detonation transition in a one-dimensional chamber

Type

Conference Paper

Authors

Sow, Aliou
Jaasim, Mohammed
Hernandez Perez, Francisco
Im, Hong G.

KAUST Department

Computer, Electrical and Mathematical Sciences and Engineering (CEMSE) Division
Mechanical Engineering Program
Physical Science and Engineering (PSE) Division
Clean Combustion Research Center

Date

2017-01-01

Abstract

Regulations in CO2 emissions have led to the development of downsized, highly boosted and direct injection technologies that enhance power density and fuel consumption in Spark Ignition (SI) engines. Despite these enhancements, SI engines are prone to exhibit uncontrolled pre-ignition, whose occurrence, specially at low speed and high load, can trigger detonation development which is associated to violent peak pressures and pressure oscillations that can damage the engine. The uncontrolled combustion is a major challenge to overcome for the new generation of SI engines to be widely used. Here, we studied detonation development using high resolution direct numerical simulations of a closed one-dimensional combustion chamber. A highly reactive hydrogen-oxygen mixture with a detailed chemical reaction mechanism is used in this study. Unlike previous studies where bulk mixture properties were arbitrarily defined, we used results from engine simulations obtained with the CONVERGE software to set the initial bulk mixture properties. Depending on the initial conditions two different paths to detonation development were identified: 1) auto-ignition near the wall followed by detonation combustion and 2) transition of the initial propagating flame to detonation combustion mode through flame acceleration.

Acknowledgements

The work presented in this paper was sponsored by KAUST and Saudi Aramco under the FUELCOM II Project. The computational work utilized resources from the KAUST Supercomputing Laboratory.