On using ab initio calibration to fit temperature from AlO B-X emission

Aluminum Oxide (AlO) emission can be an important indicator of reaction and temperature in combustion and explosion systems. The Δv=-1 and particularly Δv=-2 sequences are attractive for temperature measurement, but have presented some challenges in modelling, especially for high vibrational energy transitions. Recent studies and ab initio simulations have offered parameters for the simulation of high resolution AlO spectra (Patrascu et.al. Mon. Not. R. Astron. Soc., 2015). In this paper we compare simulated spectra with high resolution emissions of AlO, both from laser induced breakdown spectroscopy and carefully calibrated shock tube experiments (3000<T<5000 K). In this manner, we assess the spectral model and its capacity for temperature assignment, rather than assume model accuracy and a best-fit-temperature. The experimental and modeled spectra agree below 0.1 Å resolution, and line features agree within our experimental resolution, in some regions down to 3 pm. Least squares fitting was performed on the experimental spectra, resulting in temperatures in agreement with calibration temperatures. The Δv=-2 sequence fits with residuals less than 4%. Our analysis suggests that experimental measurements with a resolution of 0.2 nm in this sequence can reliably fit the temperature within 100 K. Because of the apparently accurate fits to temperature, apparent accuracy of line position list, and reasonableness of line width model, we assess that the line strengths are likely also accurate and that ab initio simulated calibrations are useful for measuring temperature in AlO.