Abstract
Quantitative species measurements are essential for optimizing hydrogen-fueled micro-mixing combustion, yet inherently weak spontaneous Raman scattering signals pose significant challenges for flame diagnostics. This study develops a systematic quantitative Raman spectroscopic methodology featuring simplified multi-pass excitation and automated spectral processing. Three-pass excitation using flat mirrors achieved three-fold signal enhancement with less than 7% energy loss, while avoiding air breakdown through afocal beam compression. An automated data processing workflow integrating Savitzky-Golay filtering, wavelet denoising, and Voigt profile fitting with a priori peak constraints was established to eliminate operator dependence and ensure reproducibility. Validation against standard McKenna flat-flame burner data demonstrates that the method achieves measurement deviations within 3% for major species (H2, O2, N2, H2O), with an expanded uncertainty of 2.8% (k = 1.96). These results reveal a nonlinear coupling between Raman signal intensity and equivalence ratio. As temperature rises, the reduction in molecular number density due to thermal expansion overwhelms the increase in cross-section, which leads to signal attenuation. The methodology was subsequently applied to measurements at 21 spatial locations in a newly designed rectangular micro-mixing hydrogen burner, capturing axial reactant gradients and flame front positions. This study establishes a reproducible quantitative Raman diagnostic methodology, providing a technical reference for species measurements in complex combustion environments.

This publication has 29 references indexed in Scilit: