Two-Dimensional Temperature Measurements in Krypton-Seeded Co-Flow Flames

Poster presented at the Fall 2020 ACS National Meeting.

This work is now included in a peer-reviewed publication available at the link below under "References".

Temperature is a critical variable in combustion devices, since it controls the rates of the chemical reactions that determine efficiency and emissions. Conventional temperature diagnostics used in combustion research include thermocouples and laser-based methods in the UV or visible portions of the spectrum. Thermocouples require radiation corrections and can perturb the flame, while laser-based methods suffer interferences from soot particles and can depend on many uncertain parameters (bulk gas composition for Rayleigh scattering, collisional quenching rates for fluorescence, etc.). This study used a novel x-ray-based technique to measure in-situ temperatures in methane/air co-flow flames and compared the results to numerical simulations. The methane and air streams were seeded with a small concentration of krypton (2.4% by mole), and the krypton number density was measured using x-ray fluorescence (XRF). The Kr atoms were converted to core-hole ions by ionization of 1s electrons with 15.1 keV photons, then the K-alpha fluorescence at 12.6 keV was detected as 2p electrons relaxed into the 1s holes. To obtain adequate signals, the measurements were performed using a synchrotron source: beamline 7-BM at the Advanced Photon Source. The measured number densities were converted to temperature using the ideal gas law and assuming that the Kr mole fraction and pressure were constant. The errors from these assumptions were quantified with the simulations and found to be negligible (pressure) and 3.0% or less (Kr mole fraction). The simulations included a detailed chemical kinetic mechanism and species-dependent transport properties; the adequacy of the transport model was validated by comparing measurements and simulations for a nonreacting flow where krypton was only present in the inner jet stream. The XRF data analysis does not depend on the collisional environment since the lifetime of the core-hole ion is only 0.18 fs. The XRF signal was integrated for 1 s at each flame location, which produced random uncertainties of 4.9% or less in each individual measurement. The flame contained significant amounts of soot (maximum soot volume fraction = 0.3 ppm), but the XRF measurements were unaffected by it due to the weak interaction of hard x-rays with carbon and phase boundaries. Thus, temperatures could be obtained with high accuracy throughout the flame. Excellent agreement was found between the measured and simulated temperatures.