Soot formation in turbulent flames and fires in an important processes. Soot is a pollutant with adverse health effects. Its emission reduces combustion efficiency. High soot concentrations strongly impact radiative heat transfer through emission, but soot that breaks through flames results in radiative shielding. Accurate soot formation and transport models are important for correctly predicting and simulating flames and fires. Modeling these processes is challenging due to the complex chemical formation processes, and soot transport with differential diffusion relative to a flame, which dictates soot concentrations through the local temperature and composition fields the soot experiences as it is formed, grows, coagulates, and oxidizes. Detailed simulations of soot formation have highlighted the importance of differential transport on soot concentrations. However, detailed simulations (e.g., DNS) have high computational costs and are prohibitive for practical configurations such as fires. The one-dimensional turbulence model (ODT) is able resolve a full range of length and timescales and solves the evolution of diffusive and reactive scalars in the natural physical coordinate, essentially resolving flame structures in one dimension. The model is computationally affordable and has been successfully applied to a wide range of reacting flows. We present results of soot formation in ODT and compare the model directly to simulation results from DNS. The configuration is a temporally-evolving planar ethylene jet flame, which is an ideal configuration for comparison with the ODT model, Detailed combustion models for ethylene are applied along with a four step soot model using the method of moments with three transported moments. The same models are used in the DNS and ODT. Comparisons between the models are made for the mean jet evolution along with mean and fluctuating profiles in the physical and flame coordinates. The limitations of ODT in capturing only flame-normal transport (no flame curvature) are explored. Validation of the ODT model in this configuration will lend confidence in extending the model to configurations for which detailed temperature-velocity-composition data are not available, as well as extensions to three-dimensional versions consisting of coupled ODT lines that allow simulation of more complex configurations. Successful validation would also allow for the use of ODT as a DNS surrogate under parameter spaces (i.e., high Reynolds number) not available to DNS, or to provide statistical information at lower computational cost. These data could be used to great effect for subgrid model development and validation.
展开▼
