par Ferrarotti, Marco 
;Fürst, Magnus ;Cresci, Enrico;De Paepe, Ward ;Parente, Alessandro
Référence Energy & fuels, 32, 10, page (10228-10241)
Publication Publié, 2018-10-01
;Fürst, Magnus ;Cresci, Enrico;De Paepe, Ward ;Parente, Alessandro Référence Energy & fuels, 32, 10, page (10228-10241)
Publication Publié, 2018-10-01
Article révisé par les pairs
| Résumé : | A numerical and experimental investigation of a quasi-industrial furnace operating in moderate or intense low-oxygen dilution combustion regime, and fed with natural gas, is presented. The study analyzes the effect of various parameters, including the combustion model [eddy dissipation concept (EDC) and partially stirred reactor (PaSR)], the definition of the chemical and mixing time scale, the turbulence model, and the choice of the kinetic mechanism. The numerical results are validated against in-flame temperature profiles, pollutant emission, and OH∗ chemiluminescence images. It was found that EDC fails in providing a reasonable estimation of the ignition region, while improved predictions can be obtained using the PaSR model. A sensitivity analysis was carried out to determine the optimal mixing time scale formulation for the PaSR model. Indeed, a static time scale approach, based on defining a prescribed fraction of the integral time scale, was compared to a dynamic mixing time scale formulation, based on the ratio between the variance of the mixture fraction and its dissipation rate. Results indicate the need to modify the coefficients appearing in the scalar dissipation rate transport equation, as the latter was originally derived for homogeneous turbulence and two-dimensional configurations. Results obtained with an optimized set of transport equation coefficients are in good agreement with the experimental data and in line with those obtained calibrating the mixing constant, Cmix, in the static approach. |
