Effect of turbulence on flame initiation and combustion cyclic variation in spark-ignition engines.

Abstract

The objective of this thesis was to study the effect of turbulent flow on combustion cyclic variation during the flame initiation period. The research consisted of building an apparatus for simultaneously measuring cylinder pressure, ignition energy, flame kernel volume, and mixture velocities, on an individual cycle basis in a firing spark ignition engine. Resulting data was conditionally sampled to control interactions. The main innovation of the present work was the use of direct measures of turbulent flow strain rates in evaluating the effect of turbulence on combustion in a firing spark ignition engine using Particle Image Velocimetry. Testing was performed at the lean limit of Propane and air at typical road load running conditions. It was found that: Ignition energy deposition during breakdown and high voltage discharge did not vary from cycle to cycle for given mixture composition and initial conditions. This led to the conclusion that time scales governing breakdown and high voltage discharge were smaller than those of turbulent flow. Cycle by cycle variation in combustion pressure was correlated with variation in kernel volume at 20 degrees after spark. This proved that near the lean limit, cycle by cycle combustion variation originated in the flame initiation period. With mixture composition and state held constant, cycle by cycle variation in kernel volume was attributed to variation in turbulent flow and ignition energy characteristics. A mild correlation was observed between cycle average ignition voltage and kernel volume. Large variation in kernel volume was, however, present for a given ignition voltage value. It was mainly attributed to variation in surrounding turbulent flow since variation due to heat transfer was considered negligible when ignition voltage was constant. For such conditions, flame growth rate decreased with increasing fluctuating shear strain rates of the turbulent flow. It was then experimentally proven, by directly measuring turbulent flow shear strain rates, that they resulted in reducing burn rates, and that their random variation caused cycle by cycle variation in combustion.