A comparison of multivariate LIBS and chemiluminescence-based local equivalence ratio measurements in premixed atmospheric methane–air flames

In the present work, a comparative study of fuel–air equivalence ratio measurements in premixed atmospheric methane–air flames with ungated laser-induced breakdown spectroscopy (LIBS) and natural chemiluminescence emission spectroscopy is presented. With ungated LIBS, the strong elastically scattered laser light greatly reduces the measurement dynamic range and sensitivity. For effectively suppressing the elastically scattered laser light in ungated LIBS detection, a polarizer and a notch filter in the signal collection system were tested. It was observed that the polarizer performed better compared to the notch filter in suppressing the elastically scattered laser light. The LIBS spectra collected at ten different equivalence ratios ranging from 0.74 to 1.42 were used to develop the equivalence ratio calibration model using atomic line peak intensity ratios (for H656nm, N 500 nm + , and O777nm) as well as broad spectral features with a multivariate, partial least squares regression (PLS-R) approach. The developed calibration models were tested with an unknown data set comprising LIBS spectra taken at similar experimental conditions used in developing the calibration model. The LIBS-based prediction results were also compared with chemiluminescence-based predictions from the same experimental setup. It was found that the LIBS-based calibration provides a more robust measurement approach with better predictions of equivalence ratios compared to the chemiluminescence-based calibration.