Effect of Aromatic Components in Surrogate Fuels on Soot in Co-flow Flames and a Model Gas Turbine Combustor
Open Access
- Author:
- Iyer, Venkatesh Ramasubramanian
- Graduate Program:
- Mechanical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 09, 2012
- Committee Members:
- Thomas Litzinger, Dissertation Advisor/Co-Advisor
Robert John Santoro, Committee Member
Daniel Connell Haworth, Committee Member
Andre Louis Boehman, Committee Member - Keywords:
- Soot
diffusion flames
surrogate fuels
Threshold Soot Index
smoke point
aromatic compounds. - Abstract:
- This study is part of a larger effort to formulate surrogate fuels for the military jet fuel, JP-8. In order for the surrogate fuel to match the combustion properties of JP-8, the composition of the surrogate fuel was selected so that its H/C ratio, molecular weight, derived cetane number, and the threshold soot index (TSI) match the corresponding values for the JP-8. The first part of this study focuses on establishing the utility of TSI as the parameter to match the sooting tendency of the surrogate fuels and JP-8. Measurements were conducted to compare the soot volume fractions for JP-8 and its TSI-matched surrogates in laboratory flames and in a model gas turbine combustor. Because the sooting tendency of the aromatic component has a critical influence on the TSI of the surrogate fuel, it is necessary to understand the relationship of TSI and soot formation for aromatic compounds in order to pick the most suitable aromatic component for the surrogate fuel. The second part of the thesis focuses on understanding the relationship between TSI and soot characteristics for aromatic-fueled laboratory flames. A laminar co-flow wick burner was developed for use with liquid hydrocarbon fuels. The burner was calibrated to correlate the sooting tendencies of all the fuels studied using TSI. Peak soot volume fractions for binary mixtures of iso-octane and toluene were correlated with the smoke heights using Kent’s model which showed that the wick burner was consistent with the burners of Kent and Olson et al. In order to establish the extent to which the soot fields of surrogate fuels match that of JP-8, laser extinction measurements were used to obtain soot fields for JP-8 and surrogate fuels on the wick burner. In addition, net soot production for JP-8 and solvent-based surrogate fuels were measured on a model gas turbine combustor for JP-8 and its TSI-matched surrogate fuels. The soot fields in the flames and the soot concentrations in the combustor showed excellent agreement within ± 15% which confirmed that TSI works well as an indicator of sooting tendency of surrogate fuels matched to the TSI of JP-8. The study of the relationship of TSI and the soot characteristics of aromatic compounds involved four fuels. 1,3,5-trimethylbenzene and n-propylbenzene were studied because they were potential aromatic components for the surrogate fuels in the MURI program. iso-propylbenzene and m-xylene were selected because they had similar TSI to 1,3,5-trimethylbenzene and n-propylbenzene respectively, but substantially different soot volume fractions (20-30%). Axial soot measurements and radial soot profiles for the aromatic fuels were used to characterize the formation and oxidation dominated regions in the flames. 1,3,5-trimethylbenzene and m-xylene showed about 20% higher soot volume fractions than the propylbenzenes. Differences were also observed in the overall soot formation and oxidation rates among the fuels. Estimates of formation and oxidation based on temperature measurements from the literature showed that differences in temperatures cannot fully account for differences in sooting. Consequently, studies were done to determine whether morphology of soot could account for the differences. Morphology was investigated using multi-angle scattering (MAS) and thermophoretic sampling/TEM analysis. MAS showed that primary particle sizes were the same within the experimental uncertainty of ± 15% for all the fuels. 1,3,5-trimethylbenzene and m-xylene had 2.5 times and 2 times higher particle number densities, respectively, than n-propylbenzene. TEM data for n-propylbenzene and m-xylene also showed that the fuels had similar particle sizes. Thermogravimetric measurements of oxidation rates of soot from all four aromatics showed that the mass specific rates of oxidation were identical for all four fuels. Analysis of these results provided additional confirmation that primary particle sizes were the same. These results pointed towards different particle number densities for the fuels as the reason that they form different amounts of soot. Since number densities are related to nucleation, modeling analysis were performed to examine nucleation rates for m-xylene and n-propylbenzene using state-of-the-art chemical kinetic mechanisms. Higher concentrations of PAH and acetylene from the simulations were consistent with the hypothesis that m-xylene had a greater number of primary particles than n-propylbenzene.