Flow Reactor Autoignition Studies of Iso-octane at High Pressures and Low-to-intermediate Temperatures

Open Access
Christensen, Michelle Kathleen
Graduate Program:
Mechanical Engineering
Doctor of Philosophy
Document Type:
Date of Defense:
April 05, 2012
Committee Members:
  • Robert John Santoro, Dissertation Advisor
  • Thomas Litzinger, Committee Member
  • Harold Harris Schobert, Committee Member
  • Stephen R Turns, Committee Member
  • autoignition
  • ignition delay time
  • iso-octane
  • flow reactor
The trend in internal combustion engine and aviation gas turbine design is moving increasingly toward a computational approach, requiring well-validated chemical kinetics models. Autoignition delay measurements are among the key set of validation data used by chemical kinetics modelers. Autoignition is an ignition event that occurs when the mixture reaches conditions of temperature, pressure, and equivalence ratio such that the rate of chain branching exceeds the rate of chain termination. In this case there is no external ignition source, such as a spark or initiating flame. Iso-octane, the compound studied in this work, is one of the fuels used to establish the octane rating for gasoline. Recently it has become of further interest as a component in fuel mixtures referred to as surrogate fuel mixtures such as JP-8. These surrogate fuel mixtures consist of only a few compounds that behave chemically in a manner similar to a practical fuel. Studies of autoignition of iso-octane were conducted in a high-pressure flow reactor. A novel methodology was employed in which the minimum equivalence ratio required for ignition was established for specific conditions of pressure, temperature, and residence time. Liquid iso-octane was spray injected into the products of a hydrogen/oxygen/argon preburner, and the resulting vaporized mixture was then rapidly mixed with air in the high-pressure flow reactor. Autoignition delay times were obtained at pressures of 15, 17.5, 20, and 22.5 atm, for a temperature range of 640-850 K. Residence times of approximately 70, 100, 125, 155, and 175 ms were investigated over equivalence ratios ranging from 0.25 to 0.8. Very few previous iso-octane ignition delay studies have investigated the conditions included in the present study. However, the current experiments have some overlapping pressure and temperature conditions (15 atm and 650-850 K) with rapid compression machine experiments. The rapid compression machine results are all for stoichiometric mixtures, whereas the current results are for equivalence ratios ranging from 0.37-0.8. The results from this work are useful for model validation as they are consistent with trends from previous studies and extend into a range of pressures, temperatures, and equivalence ratios not previously investigated. The onset of autoignition for all pressures occurred at approximately 640K. Trends in the results showed that as temperature increased, the threshold equivalence ratio decreased. As temperature increased further into the negative temperature coefficient (NTC) region, the threshold equivalence ratio increased with increasing temperature. The NTC behavior started at temperatures of approximately 700-725K and ended at approximately 775K. At temperatures above 775K, the threshold equivalence ratio again decreased with increasing temperature. Pressure also had a strong effect on ignition delay. For all conditions, as pressure increased the threshold equivalence ratio decreased. Results were compared with two chemical kinetics models. Good agreement was shown with a model from Lawrence Livermore National Laboratory (LLNL) for the conditions studied in this work. The main discrepancy between the experimental and modeling results was the overprediction of the threshold equivalence ratio in the NTC region at the shorter ignition delay times. For example, at the residence time of 70 ms, a pressure of 15 atm, and temperature of 775 K, the threshold equivalence ratios found experimentally and by the model were 0.75 and 0.95, respectively. At the longest residence time of 175 ms at the same conditions of pressure (15 atm) and temperature (775 K), the threshold equivalence ratios found experimentally and by the model were much closer at 0.44 and 0.47, respectively. The second model from Chemical Reaction Engineering and Chemical Kinetics (CRECK) gave similar results as the experiment for the temperature at the onset of autoignition, but showed no presence of the NTC region that was observed experimentally and predicted by the LLNL model. Neither the LLNL nor CRECK model predicted autoignition below temperatures of 650 K. The LLNL model was used to identify key reactions controlling ignition through reaction pathway and temperature sensitivity analyses for the conditions of this study. The results showed the importance of the location of H-atom abstraction from the iso-octane molecule in determining how reactions proceeded. H-atom abstraction from the tertiary site on iso-octane served to slow overall reactivity, while abstraction from the primary and secondary locations increased overall reactivity.