Search

Turbulent combustion models are substantially important in CFD simulation of combustion. Some simple models frequently cannot well simulate the finite-rate chemistry. Widely recognized models are good only for certain flame types and flame structures. More general is the PDF equation model, but its computation requirement is very large, in particular when using large-eddy simulation (LES). Alternatively, the statistical moment models, such as unconditional moment models were developed based on the idea of turbulence modeling. However, the difficulty in treating the highly nonlinear exponential function of temperature in the reaction term was encountered. Different investigators adopted various approximations by neglecting different correlation terms, leading to under-predicting the time-averaged reaction rate. The present author entirely abandoned different approximations, proposed a final version of the second-order moment (SOM) model. This final version was used in both RANS modeling and LES of different single-phase and two-phase flames. It was validated by both experimental results and direct numerical simulation (DNS). [ABSTRACT FROM AUTHOR]

Direct numerical simulation (DNS) of three-dimensional turbulent reacting channel flows with buoyancy is carried out using a spectral method. Statistical results from the DNS database are used to validate an algebraic second-order-moment sub-grid-scale (ASOM-SGS) combustion model and show that the ASOM-SGS model is reasonable. Furthermore, a methane-air jet flame is simulated by large-eddy simulation (LES) using the ASOM-SGS model and indicates that the Reynolds-averaged Navier-Stokes ASOM combustion model is a reasonable model. [ABSTRACT FROM AUTHOR]

A methane-air swirling diffusion flame is studied by large-eddy simulation (LES) using second-order moment (SOM) and simplified PDF subgrid-scale (SGS) combustion models, Smagorinsky-Lilly and dynamic kinetic energy (DKE) subgrid-scale (SGS) stress models. The predictions are validated by the experimental data. For LES statistics, comparison of the predicted axial and tangential time-averaged and root-mean-square (RMS) fluctuation velocities, temperature, RMS fluctuation temperature, CO2 and H2O mass fractions with the experimental results shows that all of the SGS models adopted here give results with only slight differences, but the DKE + SOM models are somewhat better than other SGS models. For instantaneous results, the predicted flame shape with a neck region is in agreement with that observed in experiments. The predicted flame shape and length obtained using the DKE + SOM models are better than those obtained using other SGS models. There are both large-scale and small-scale structures in the swirling flame; combustion reduces the large-scale structures and enhances the small-scale structures. [ABSTRACT FROM AUTHOR]

Swirling flows and methane–air swirling diffusion combustion are studied by large-eddy simulation (LES) using a Smagorinsky-Lilly subgrid-scale (SGS) turbulence model, a second-order moment (SOM) SGS combustion model, proposed by the present authors, and an eddy-break-up (EBU) combustion model used by some investigators. The LES statistical results are compared with the experimental results and the Reynolds-averaged Navier-Stokes (RANS) modeling results using the Reynolds stress equation model and the SOM combustion model. For swirling flows, the LES statistical results give better agreement with the experimental results than the RANS modeling, indicating that the adopted SGS turbulence model is suitable for swirling flows. For swirling combustion, both the SOM SGS combustion model and the RANS-SOM model give results in good agreement with the experimental results, but the LES-EBU modeling results are not in agreement with the experimental results. The LES instantaneous results show large vortices formed in the shear layer and a complex vortex shedding pattern in the downstream region. It is shown that combustion weakens the vortex structures, combustion is intensified by the coherent structures, and swirl thickens the flame front. [ABSTRACT FROM AUTHOR]