Ignition Quality Tester: an alternative for characterizing the combustion kinetics of low volatility fuels, The

Abstract

The objective of this thesis is to demonstrate that the Ignition Quality Tester (IQT) can be used to validate the kinetic mechanisms of both high and low volatility fuels. Such validated mechanisms are an essential component for engine models used to improve efficiency and determine the impact of alternative fuels. There are other approaches to measure the ignition kinetics of high volatility fuels, but only very limited data are available for low volatility fuels. The IQT was modified by increasing the range of temperatures it could access and by implementing a purge program so that the accuracy and repeatability of experiments at low pressures could be increased. Experiments were performed to characterize the effect of varying parameters (temperature, pressure, oxygen concentration, equivalence ratio, mass of fuel injected, choice of diluent, fuel physical properties, and fuel structure) on the ignition delay, and whether these effects were due to the chemical kinetics or spray physics. CFD modeling, run without chemistry, was used to show that at long times (>20ms) the IQT becomes pseudo-homogeneous in both temperature and equivalence ratio. This suggested that a 0-D homogeneous batch reactor model could be used to predict the ignition delay at the longer times. Experiments were performed for five heptane isomers where accurate mechanisms are available, and the 0-D model ignition time predictions were consistent with the measurements. Similar favorable comparisons were found for iso-octane, another well studied high volatility fuel. Attention then shifted to validate chemical mechanisms for low volatility fuels. Model predictions for n-hexadecane were a factor of sim 2.5 longer then the observed ignition delays at long times (> 20 ms). This difference could be due to the older rate rules used in the mechanism. Experiments were done with 2,2,4,4,6,8,8-heptamethylnonane (HMN) since the inherently lower reactivity of this fuel allows NTC behavior to be observed without needing to go to the lower pressures (thus allowing experiments more relevant to diesel combustion). The 0-D model significantly underpredicted the ignition delay. This provided an opportunity to develop an improved HMN mechanism. It was discovered that the highly branched structure of HMN meant that additional terms needed to be considered when computing the thermodynamic properties. This updated thermo, in combination with updated estimates for various reaction types, greatly improved the HMN mechanism.