effects of fuel molecular structures on pollutants in co-flow laminar flames
Open Access
- Author:
- Wang, Yefu
- Graduate Program:
- Mechanical Engineering
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 01, 2015
- Committee Members:
- Thomas Litzinger, Dissertation Advisor/Co-Advisor
Robert John Santoro, Committee Member
Daniel Connell Haworth, Committee Member
Sarma V Pisupati, Committee Member
Donald Richards, Committee Member - Keywords:
- fuel
molecular structure
co-flow laminar flame
soot
aromatic species
LII
LIF
laser diagnostics - Abstract:
- This study is part of a larger effort to establish a science-based model to predict the emissions from gas turbine engine combustors using alternative fuels. In order to validate and improve the chemical mechanisms in the model, four binary fuel mixtures comprised of the hydrocarbon compounds representative of the classes compounds that are expected in alternative aviation fuels. In each fuel mixture, n-dodecane was the base component. The second component was m-xylene, methylcyclohexane, iso-octane, or n-heptane that were selected to represent the molecular structures of aromatic, cyclo-paraffin, iso-paraffin, and n-paraffin. The binary fuel mixture (25% m-xylene and 75% n-dodecane in liquid volume fraction) was also evaluated as a surrogate for JP-8. A burner system was developed and built to produce co-flow laminar jet flames with liquid fuel mixtures. The experimental conditions for flames were set at three equivalence ratios (Φ) of the fuel jet—Φ=∞, Φ=6, and Φ=2—to simulate the soot-rich zones in gas turbine engine combustors. The combination of laser extinction and laser-induced incandescence (LII) was applied to obtain the spatial distributions of soot volume fraction quantitatively. “Small aromatics” and “large aromatics,” containing 1-2 aromatic rings and 3-4 aromatic rings respectively, were detected by laser-induced fluorescence (LIF). A special configuration of thermocouple probe was developed to obtain the temperature distributions in the soot-free regions of the flames. Experimental results indicated that the PAH and soot from all paraffin fuels are similar, but PAH and soot of the aromatic fuel were much larger than for the paraffin fuels. The amount of soot was found to be higher in aromatic flames than in paraffin flames by a factor of between 2-4. The maximum LIF signals from both small and large aromatics along centerline were found to be approximately ten times higher in the aromatic fuel than in paraffin fuels. Similar results, especially soot volume fraction distributions, was found between JP-8 and the m-xylene/n-dodecane fuel. The experimental results were compared in detail to simulation results provided by Dr. Katta of Innovative Scientific Solutions, Inc. Basic consistent distribution trends for each fuel mixture were established with the simulation results. Similar qualitative distributions of soot volume fraction and semi-quantitative LIF signals from aromatic species as well as temperature were found for flames burnt with all fuel mixtures, even though the simulation always displayed large areas of soot and aromatics existing regions. The maximum soot volume fraction along centerline in flames was estimated with values similar to experimental data for paraffin fuels. Several potential explanations were produced for the significant discrepancy of soot distributions in aromatic flames between the simulation and experimental data. Other simulation results, including the distributions of OH and rates of soot nucleation, soot surface growth, and soot oxidation were presented to gain insight into the reasons for the discrepancies between the simulations and the experiment.