Soot contributes to many of the health and safety hazard associated with air pollution from combustion systems. Scientists lack a fundamental understanding of soot formation and transport in flames, which is crucial to developing accurate combustion models and predictive simulation tools. Modeling soot is difficult because its reactions span many orders of magnitude in both length and time scales. Coupling soot chemistry with turbulent combustion, which comprises most practical combustion systems, further complicates modeling. Direct numerical simulation (DNS) is the only simulation tool that can fully resolve a flame’s flow structures, but is computationally prohibitive for turbulent flows. Reynolds-averaged Navier-Stokes (RANS) methods and large eddy simulation (LES) capture large-scale flame structures, but do not resolve flames at fine dissipation scales. The one-dimensional turbulence (ODT) model, however, resolves fine scales and models large-scale turbulence with stochastic eddy events. ODT is useful for modeling reacting, turbulent flows because it is much less computationally expensive than both DNS and LES, but still represents detailed flame structures.
We present an extension of the ODT model to include soot chemistry, and we perform simulations of turbulent, nonpremixed ethylene jet flames. The soot chemistry consists of reaction mechanisms for soot formation, coagulation, surface growth, and aggregation, such as HACA surface growth and fractal aggregation models. Soot particles are modeled using the quadrature method of moments (QMOM), and particle moment transport equations are introduced. The literature indicates that such extensions will improve the ODT model’s accuracy and validity. We will qualitatively compare results with existing simulations and experimental data to ascertain how well this extended ODT model predicts soot production and transport in flames.
See more of this Group/Topical: Catalysis and Reaction Engineering Division