Experimental and modeling study of laminar burning velocities and nitric oxide formation in premixed ethylene/air flames
Forskningsoutput: Tidskriftsbidrag › Artikel i vetenskaplig tidskrift
Adiabatic laminar burning velocities and post-flame NO concentrations for flat, non-stretched, premixed C2 H 4 /air flames were experimentally determined with a heat flux burner of improved design, over equivalence ratios ranging from 0.7 to 2, at atmospheric pressure and initial temperature of 298 K. Recognizing that C2 H 4 is a main intermediate in high-temperature oxidation pathways of heavy hydrocarbons, these data are essential for the development, validation and optimization of kinetic models for any fuel. The present measurements were then compared with the data available in the literature obtained with different techniques under the same experimental conditions. Regarding burning velocity measurements, the comparison showed considerable scatter among existing stretch-corrected data, which corroborate the necessity for the present adiabatic, non-stretched results. Regarding NO concentrations, an excellent agreement was observed between the present in situ, non-intrusive laser-induced fluorescence measurements and the only dataset available in the literature, determined by the phenol disulfonic acid method. A comparison of experimental and computational results using two contemporary comprehensive, detailed chemical kinetic mechanisms, along with one from the authors presented in this work, was also conducted and discussed. Discrepancies between experiments and model predictions and among models themselves were observed under rich conditions. Notwithstanding, the present updated model showed overall good performances in reproducing both laminar burning velocities and nitric oxide concentrations. Further numerical analyses were performed to identify the main causes of the observed differences. The results showed that the description of the relative importance of reactions involving vinyl and hydrogen cyanide consumption pathways, due to remaining uncertainties, lead to the different model behaviors.