Comparison of Fuel Rich and Stoichiometric Premixed Toluene Flames

Wednesday, November 10, 2010: 9:54 AM
150 A/B Room (Salt Palace Convention Center)
Wenjun LI1, Bin Yang2, Phillip Westmoreland1, Tina Kasper3 and Nils Hansen3, (1)Chemical and Biomolecular Engineering, North Carolina State University, Raleigh, NC, (2)Cornell University, Ithaca, NY, (3)Sandia National Laboratories, Livermore, CA

Aromatic fuels are important and significant components in the commercial fuels, having several common advantages, including higher energy density and octane number, and better knock rating. Mean while, they are also more potential for high formation of benzene, poly-cyclic aromatics hydrocarbon (PAH) and soot, one important pollution issue. As the simplest aromatic fuel, toluene chemistry captures the key features common to other substituted aromatics, and it is also chosen as the aromatics representative in most surrogate fuel models.

In the current work, the detailed flame structure for a fuel-lean toluene premixed flame (15 torr, phi=0.9, toluene/O2/50.0% Ar, 120.0 cm/s feed velocity at 298 K) was mapped, with mole fraction profiles for more than 40 stable and radical species measured by photo-ionization molecular beam mass spectrometry (PIMBMS) from Advanced Light Source at Lawrence Berkeley National Laboratory[1], and temperature profiles measured by NO-doped Laser Induced Fluorescence (LIF). High-resolution energy scans (ion signal along the photon energy) were used for species identification by their mass charge ratios and ionization thresholds, while burner scans (signal along the flame axis) were converted to species mole fraction profiles. This measured flame structure is compared with a literature fuel-rich toluene flame (30 torr, phi=1.9, toluene/O2/42.7% Ar, 34.4 cm/s feed velocity)[2].

A recently updated mechanism model was incorporated for both flames simulation, with generally good agreement with the experimental result. This mechanism was found to be able to simulate the fuel-rich flame better, due to some species over-prediction in the simulation of the fuel-lean flame. Overall reaction flux and detailed reaction pathway analysis were further carried out to illustrate the combustion chemistry differences at the two stoichiometric flame conditions.

The comparison showed that: (1) the fuel-rich flame had higher magtitudes for most smaller alkyls C1-C6, especially for C2H2, C3H3, C4H4, C5H5 and C5H6; (2) only a couple of higher mass species were identified in this fuel-lean flame, including C9H8, C9H14 and C10H8 different isomers with very small magnitude (1E-5), while in the fuel-rich flame, their magnitudes were all more than 20 times higher, and especially there were many more other higher aromatics species; (3) toluene oxidation into C6H4CH3 and OC6H4CH3 contributed much more in the fuel-lean flame; (4) C6H5CH2 dominantely broke the aromatic ring into C5H5+C2H2 in the fuel-rich flame, with no significance from other side chain oxidation; (5) similarly C5H5 decompositon into C3H3+C2H2 and its H-addition to form C5H6 was much more important in fuel-rich flame.

The main flame chemistry difference was found to be from the different abundance of the most active radicals H, O and OH. In the fuel-lean flame, all H, O and OH had similar high abundance (mole fraction peaking 2E-2), this made those oxidation chemistry via O and OH attack very important for the key species destruction, such as C6H5CH3, C6H5CH2 and C5H5. While in the fuel-rich flame, H radical (mole fraction peaking 1E-2) dominated much more than O and OH radicals (mole fraction peaking 1E-3), which made much higher dominance of H radical attack and direct decomposition chemistry, due to the lack of oxidation via O and OH attack.

References

1. N. Hansen, T. A. Cool, K. Kohse-Höinghaus, P. R. Westmoreland, “Recent Contributions of Flame-Sampling Molecular-Beam Mass Spectrometry to a Fundamental Understanding of Combustion Chemistry,” Progress in Energy and Combustion Science 35(2) (2009) 168-191.

2. Y. Li, L. Zhang, Z. Tian, T. Yuan, J. Wang, B. Yang, and F. Qi. Energy & Fue 23(3) (2009) 1473–1485.


Extended Abstract: File Not Uploaded