Coherent Anti-Stokes Raman Spectroscopy CARS Measurements in a Laminar Flame Coal Combustor

Page 10

constant over the 23 meter path length. but the diameter of the dye beam was decreased to about 20 mm before it reached the final field lens next to the reactor window. The OG 515 Schott long-pass filter was placed in the beams just before the reactor window in order to fitter out any CARS signal that had been generated outside of the reactor. Montana subb~uminous coal was used to seed the natural gas flame . The coal particles were less than 38 microns in diameter. Other information regarding the two flames investigated are shown below. Natural~ Coal-Seeded Natural~ Nat. Gas Flow Rate (kglhr) 0.489 0.488 Air Flow Rate (kglhr) 10.137 10.126 Coal Feed Rate (kglhr) 0.000 0.039 Overall Stoichiometric Ratio 1.20 1.16 Since the flame was laminar. it was assumed that temperature measurements in the horizontal direction across the flame would be symmetrical and fairly constant about the central axis of the burner. Therefore. in this study the laser beams were focussed to the center of the 11.4 cm diameter laminar flame. and this focal volume was moved axially through the flame by moving the optical breadboards w~h the mechanical crank and chain assembly. Measurements were taken at axial pos~ions from 0.64 to 10.2 cm (0.25 to 4 inches) downstream of the burner face. The average spectra were obtained by collecting 100 shots on the OMA and then dumping these averaged shots to the computer. Background spectra were taken immediately before or after each data set. The single shot data presented in this paper were obtained at the 1.3 cm (0.5 inch) position. A dye profile used to normalize the data was obtained prior to each set of tests conducted. The two flames investigated were fuel-lean. therefore. carbon monoxide was not detected in the flames . However. had the flames been fuel rich. carbon monoxide would have been in sufficient concentration to detect it simultaneously with the nitrogen signal. DATA REDUCTION All of the coal-seeded gas flame spectra taken showed some nonresonant enhancement with the laser energies used (ca 70 mJ per pulse in the pump beam. ca 10 mJ per pulse in the dye beam). The amount of nonresonant enhancement varied from spectrum to spectrum (See Figure 7). The gas flame spectra did not show any nonresonant enhancement and were therefore fit with an unmodffied version of a nonlinear least-squares fitting program from Sandia National Laboratories (Palmer 1987). The coal­seeded gas flame spectra were fit with a modffied version of the fitting code. Temperatures for the flames 10