Thermometry by vibrational Raman spectroscopy of nitrogen: Identification and impact of spatial averaging effects

Thermometry utilizing the vibrational spectrum of nitrogen is a very common tool in combustion diagnostics via spontaneous and coherent Raman scattering spectroscopy. In the presence of strong temperature gradients, however, it is likely that hot and cold gas portions contribute to the signal. This is known as spatial averaging, and it can result in severe measurement errors. The present work proposes and demonstrates two straightforward approaches that allow the identification of spatial averaging as well as estimating the resulting error. For this purpose, simulated Raman spectra are considered in order to avoid any impact from experimental artifacts. The first approach utilizes difference spectra, whereas the second one is based upon principal component analysis (PCA). Two scenarios were tested: (1) spectra of 50:50 mixtures exhibiting a common mean temperature and (2) spectra of mixtures with systematically varied fractions of hot (2000 K) and cold (300 K) gas. It is shown that the resulting measurement error can be as high as 500 K.

Automated summary

This study proposes and demonstrates two straightforward approaches to identify spatial averaging and estimate the resulting error. By considering simulated Raman spectra, it is shown that the resulting measurement error can be as high as 500 K.