화학공학소재연구정보센터
Fuel, Vol.212, 1-11, 2018
A calibrated soot production model for ethylene inverse diffusion flames at different Oxygen Indexes
A numerical investigation was carried out in order to gain insights into the soot production mechanisms by studying the effect of the oxygen concentration of the oxidant stream, known as Oxygen Index (OI), on soot production and flame radiation in laminar Inverse Diffusion Flames (IDF). Several laminar axisymmetric IDF were simulated, varying the OI from 17% to 35%. All flames were fueled with pure ethylene. Comparisons to experimental data were intended to assess and improve the capabilities of a two-equation acetylene/benzenebased semi-empirical soot production model and a Full-Spectrum correlated-k (FSCK) radiative property model. Higher OI were found to generate shorter flames, presenting higher temperatures and an increase in the soot production and energy irradiated. Results show that simulations predict correctly the experimental behavior observed by changing the OI, but reliable predictions are limited to values under 25%. In order to improve the generality of the results, the soot production model was modified, incorporating the soot surface aging effect in an approximated way on the surface growth mechanism, and then it was calibrated. A square root dependence was considered in the specific soot surface area rather than a linear dependence. A sensitivity analysis shows that the surface growth process is the most sensitive in terms of predicting the soot content produced. Simulations after calibration were in satisfactory agreement with the experimental measurements, presenting accurate predictions from both local and overall points of view. Results demonstrate that IDF is a relevant configuration to obtain insights concerning soot modeling, providing a decoupled oxidation process, similar residence times at different OI, and thermal age controlled by temperature alone. Results also demonstrate that considering an aging effect on the soot surface reactivity is necessary in order to properly model the variations induced by changing the OI.