How well do we know thermal-NO?An investigation of NO formation in flames over a wide temperature range

Abstract

This study investigates the large variability and uncertainty in the thermal-initiation reaction rates found in the literature. An experimental study is conducted at atmospheric pressure in lean, premixed, laminar methane-air stagnation flames. Flame temperatures ranging from 1900K to 2500K are achieved by varying the argon concentration, in 21% and 40% oxygen mixtures balanced with nitrogen. The conditions are selected to promote the thermal route in the overall post-flame NO formation. One-dimensional velocity, temperature, and NO concentration profiles are measured with Particle Tracking Velocimetry (PTV), NO multi-line thermometry, and NO Laser-Induced Fluorescence (NO-LIF), respectively. While velocity and temperature measurements are accurately predicted by different thermochemical models, the simulated NO-LIF signal profiles show significant disagreement and large variability, relative to the measurements. Results show that, across all conditions, none of the studied mechanisms are able to capture accurately both the NO concentration, and formation rate in the post-flame region, suggesting that the fundamental chemistry remains inaccurate. The discrepancy in the predictions appears to be linked to the chosen parameters of the Arrhenius rate, specifically the pre-exponential factor, and the activation energy. Sensitivity and Reaction Pathway Analyses suggest that the differences in the Arrhenius parameters could originate from different consideration of the base radical chemistry, as well as different relative contributions of the four NO-formation routes. As a result, some models can predict similar NO concentrations but using significantly different underlying base and NOx chemistry. This implies that the models could diverge significantly in conditions where other non-thermal routes are more important. This study demonstrates the need for spatially-resolved experimental data across a broad range of experimental conditions, promoting and considering a variety of routes, in order to optimize NOx chemistry models with reduced uncertainty.