Jeffrey A. Sutton, PhD , Ohio State University
The approach taken in this research project is to formulate a series of canonical turbulent DME-fueled flames that are “equivalent” to well-characterized CH4-fueled flames and use the CH4-fueled flames as a “baseline” of which to compare to the DME flame structure and flame behavior. In this manner we use the so-called DLR flames (.221 CH4/.332 H2/.447 N2) as a basis for forming similar DME flames. Serendipitously, using a fuel mixture of .221 DME/.332 H2/.447 N2 results in a flame with a stoichiometric mixture fraction (ξs = 0.17), which is identical to the DLR flames, thus the two flame systems can be compared directly with laser-based measurements to examine mixing, reaction zone structure, and reaction chemistry.
The initial portion of this program focused on selecting the operating conditions of the DME-fueled flames. We have recently finalized a test matrix and are currently mapping out our test space which consists of laminar flame conditions through high-Reynolds number-flames near blowout. In terms of laser-based imaging, we have performed preliminary measurements of the CH2O distribution in both the DLR and DME turbulent flames using planar laser-induced fluorescence (PLIF). Initial results indicate a vast difference in both the magnitude (CH2O number density) and spatial distribution of the PLIF signal for the DME flames as compared to the DLR flames. For the DLR flames, the CH2O layers occur in the “cool” flame region between centerline and the high-temperature reaction zone as expected and demonstrated repeatedly in hydrocarbon-fueled flames. However, for the DME flames, the CH2O occurs closer to centerline and radially inward in comparison to the CH4 flames. This is most likely due to the fact that for conventional CxHy fuels, CH2O is an intermediate which involves the decomposition of the hydrocarbon fuel (through many routes) until the methyl radical, CH3 is formed. CH3 then reacts to form CH3O and ultimately CH2O. In DME (an oxygenated fuel), there is a more direct pathway to CH2O, including CH3OCH3→CH3OCH2→ CH2O or the direct production of CH3 through severing the C-O bond of the CH3-O-CH3 fuel molecule. In this manner, the CH2O distribution more closely resembles a fuel decomposition product in the DME flames vs. a higher-temperature intermediate in the CH4 flames.
In addition to the laser-based planar imaging occurring at Ohio State, a student, Katie Gabet, spent 6 weeks at Sandia National Laboratories this past summer working with Dr. Robert Barlow on some preliminary 1D Raman/Rayleigh scattering imaging in DME flames. The DME system presents an increased challenge for the Raman scattering measurements as compared to simple methane-based flames and thus, the measurements are in their infancy.
Two students (Hind Alkandry and Katie Gabet) worked on this project in various capacities during the past year. As mentioned above, this project allowed for collaboration with Sandia National Laboratories and a significant research experience for Katie Gabet. It is anticipated that during the next year, extensive laser-based measurements will be performed within the turbulent DME flames including OH, CO, CH, CH2O, and NO PLIF imaging, which will not only detail the structure of the turbulent DME flames, but provide new target flames/data for the International Workshop on Measurement and Computation of Turbulent Nonpremixed Flames and provide preliminary data which will “fuel” future proposal submissions to agencies such as the National Science Foundation and the Department of Energy.