Simultaneous, single-shot imaging of all major species (N2, O2, H2, and H2O), OH, temperature, and mixture fraction is demonstrated for the first time in H2N2 non-premixed jet flames at 12 bar. The spatial distribution of mole fraction is obtained for the four major species by recording images of Raman scattering on four separate back-illuminated CCD cameras. The available field-of-view is 25(H) × 8(V) mm2. Temperature and mixture fraction are derived from Raman scattering images. Images of OH fluorescence intensity are recorded using OH-PLIF and are converted into OH mole fraction by calculating the Boltzmann population distribution and the quenching rate accurately for each pixel. Precision and accuracy are assessed by comparing 2-D Raman/OH-PLIF measurements in laminar flames to 1-D flame computations accounting for differential diffusion. Single-shot precision on N2, O2, H2, and temperature is better than 5%. It is better than 6% and 10% for H2O and OH, respectively. Accuracy lies within these values, except for H2O with 10%. Such good performance of Raman imaging is attributed to (a) the use of non-intensified CCD cameras, (b) wavelet adaptive thresholding (WATR) image denoising, (c) rejection of flame luminosity by a Pockels cell electro-optical shutter with 500 ns gating, and (d) elevated pressure that boosts the Raman signal intensity. Measurements in a turbulent flame (3.6N2:H2 by vol. and Re = 29,000) show that most of the flame's thermochemical structure is accurately captured by unity Lewis number computations, suggesting that effects of differential diffusion are less important above some Reynolds number. This is consistent with expectations and it engenders confidence in the Raman imaging technique. Because images of both mixture fraction and OH mole fraction are available, it is also possible to reconstruct a more accurate scalar dissipation rate by projection onto the 2-D flame front normal.
Bibliographical noteKAUST Repository Item: Exported on 2020-10-01
Acknowledgements: The research reported in this publication was supported by funding from King Abdullah University of Science and Technology (KAUST). The help of Dr. M.J. Dunn with the WATR denoising method was greatly appreciated.