Bulletin of the American Physical Society
2005 58th Annual Meeting of the Division of Fluid Dynamics
Sunday–Tuesday, November 20–22, 2005; Chicago, IL
Session BH: Reacting Flow II: Modeling |
Hide Abstracts |
Chair: Heinz Pitsch, Stanford University Room: Hilton Chicago Williford B |
Sunday, November 20, 2005 10:56AM - 11:09AM |
BH.00001: Effects of the subgrid-scale mixture fraction structure on scalar and temperature dissipation in turbulent partially premixed flames Chenning Tong, Danhong Wang, Robert Barlow, Adonios Karpetis In LES of turbulent nonpremixed combustion subgrid-scale scalar mixing require modeling. Our previous studies have shown that when the SGS scalar variance is small and large the SGS mixture fraction has Gaussian and bimodal distributions, respectively. Here we study the effects of these SGS structure on the SGS mixing of mixture fraction and temperature. Our experimental results show that or fully burning SGS samples, the conditionally filtered temperature is similar to the conditionally filtered scalar dissipation for small SGS variance but differs from the latter for large SGS variance. For extinguished SGS samples the scalar dissipation generally has large values whereas the temperature dissipation is generally small. For large SGS variance the temperature dissipation is the largest for intermediate temperatures. Both the conditional SGS mixing time scales of mixture fraction and temperature increases as the SGS variance is increased, although the latter's increase is slower. The results show that the two mixture fracture fraction distributions affect the SGS mixing of in non-reactive and reactive scalars different ways, which have implications for modeling their SGS mixing. [Preview Abstract] |
Sunday, November 20, 2005 11:09AM - 11:22AM |
BH.00002: A new solver for Large-Eddy Simulations of turbulent premixed combustion in complex geometries. Vincent Moureau, Heinz Pitsch A new low Mach number numerical scheme for Large Eddy Simulation of premixed turbulent combustion in complex flow geometries has been developed. A Ghost-Fluid like method is introduced that allows the solver to accurately handle large density jumps and small flame brush thickness while using non-dissipative centered schemes for spatial integration. By coupling this solver with the G-equation combustion model, the flame dynamics are well described without resolving the flame front structure. The new numerical method is implemented in the variable-density unstructured flow solver CDP developed at the Center for Turbulence Research. Verification of the numerical methods will be presented in various simple test cases. The formalism has been used to compute a swirling turbulent premixed flame in an industrial premixed combustor. The solver proves to be very robust and to predict velocity statistics with good accuracy. It also captures the main hydrodynamic instabilities caused by the strong swirling motion. [Preview Abstract] |
Sunday, November 20, 2005 11:22AM - 11:35AM |
BH.00003: A numerical method for DNS of turbulent reacting flows Jeffrey Doom, Yucheng Hou, Krishnan Mahesh A non-dissipative, implicit numerical method is described to simulate turbulent reacting flows with simple chemistry over a range of Mach numbers. The compressible Navier-Stokes equations are rescaled so that the zero Mach number equations are discretely recovered in the limit of zero Mach number. The dependent variables are colocated in space, and thermodynamic variables are staggered from velocity in time. A novelty of the algorithm is that it discretely conserves kinetic energy in the incompressible limit. This makes it robust without compromising accuracy. Details and numerical examples of both premixed and diffusion flames will be presented. [Preview Abstract] |
Sunday, November 20, 2005 11:35AM - 11:48AM |
BH.00004: Resolution Matters: Issues in Computational Simulation of Detailed Kinetics Gas Phase Combustion Joseph M. Powers, Samuel Paolucci A robust method is employed to provide rational estimates of fine scale reaction zone thicknesses in one-dimensional steady gas phase detonations in mixtures of inviscid ideal reacting gases described by detailed kinetics models. The equations for the evolution of $N$ species mass fractions $\bf Y$ are formulated as a standard dynamical system of the form $d{\bf Y}/ dx = {\bf f}({\bf Y})$. These equations are integrated from a shock to an equilibrium end state. The eigenvalues of the Jacobian of $\bf f$ are calculated at every point in space, and their reciprocals give robust estimates of all length scales. Our main conclusion is that most computational results using detailed kinetics are not properly resolved. The finest physical length scale intrinsic to such models, $\sim1\times 10^{-5}~cm$, is much smaller than the discretization scale typically employed, $\sim 1\times 10^{-3}~cm$. The finest physical scale is consistent with that provided by the molecular collision theory. Predictions of under-resolved numerical simulations are being artificially stabilized by numerical viscosity. While our conclusion has been drawn in the context of a detonation problem, we speculate that similar conclusions can be drawn for a wide range of reactive flow simulations which employ detailed kinetics. [Preview Abstract] |
Sunday, November 20, 2005 11:48AM - 12:01PM |
BH.00005: The influence of chemical mechanisms on PDF calculations of non-premixed turbulent flames Stephen B. Pope, Renfeng Richard Cao A series of calculations is reported of the Barlow \& Frank non-premixed piloted jet flames D, E and F, with the aim of determining the level of description of the chemistry necessary to account accurately for the turbulence-chemistry interactions observed in these flames. The calculations are based on the modeled transport equation for the joint probability density function of velocity, turbulence frequency and composition (enthalpy and species mass fractions). Seven chemical mechanisms for methane are investigated, ranging from a five-step reduced mechanism to the 53-species GRI 3.0 mechanism. The results show that, for C-H-O species, accurate results are obtained with the GRI 2.11 and GRI 3.0 mechanisms, as well as with 12 and 15-step reduced mechanisms based on GRI 2.11. But significantly inaccurate calculations result from use of the 5-step reduced mechanism (based on GRI 2.11), and from two different 16-species skeletal mechanisms. As has previously been observed, GRI 3.0 over-predicts NO by up to a factor of two; whereas NO is calculated reasonably accurately by GRI 2.11 and the 15-step reduced mechanism. [Preview Abstract] |
Sunday, November 20, 2005 12:01PM - 12:14PM |
BH.00006: Large-Eddy Simulation/ Filtered-Density Function Modeling of Turbulent Reactive Flows Venkatramanan Raman, Heinz Pitsch Hybrid large-eddy simulation/filtered-density function (LES-FDF) approach for turbulent combustion simulations has the unique feature that the chemical source terms appear closed and do not require modeling. The FDF evolution equation is a high-dimensional transport equation that can be practically solved using a notional particle-based Monte-Carlo scheme. This approach has been widely used in RANS-based modeling. However, the inherent unsteady nature of the LES method combined with large computational cost require efficient, robust and numerically consistent algorithms. In this work, we provide verification and validation of a consistent numerical algorithm for LES-FDF simulation of turbulent combustion. We illustrate the accuracy of the algorithm using several test cases. In addition, we show comparisons with experimental flame data for several standard configurations including the Sandia D/E flames and the bluff-body stabilized flame. [Preview Abstract] |
Sunday, November 20, 2005 12:14PM - 12:27PM |
BH.00007: Numerical Prediction of Nitrogen Oxide Emission Using Flamelet/Progress Variable Model Matthias Ihme, Heinz Pitsch Unsteady flamelet models applied in large-eddy simulation of turbulent non- premixed combustion have been proven capable of predicting pollutants, such as CO and NO, with good accuracy. However, their application so far has been limited to simple geometries. The flamelet/progress variable (FPV) model developed for LES of non-premixed turbulent combustion in complex geometry is based on the steady flamelet approach. As a result of this assumption, slow physical and chemical processes such as radiation or the formation of NO$_x$ cannot be described by this model. An extended FPV model has been developed, where, in addition to the solution of transport equations for the mixture fraction and progress variable, a scalar equation for the NO mass fraction is solved. The chemical source term appearing in this equation is modeled using the production term from a steady flamelet library. Since the NO consumption depends on the local NO mass fraction, additional closure assumptions are introduced for this term. This extended FPV model is applied in LES of Sandia flame D. Since the unsteady flamelet model yields good predictions for NO in this flame, the results are used to assess the importance of the individual parts of the filtered chemical source term for NO and to evaluate the accuracy of the modeling assumptions. [Preview Abstract] |
Sunday, November 20, 2005 12:27PM - 12:40PM |
BH.00008: Effect of Mixing Model on Flame Extinction and Reignition using Large Eddy Simulation/Filtered Mass Density Function Approach Abhilash Chandy, David Glaze, Steven Frankel The effect of using three different subgrid mixing models on the dynamics of flame extinction and reignition is studying using a hybrid Large Eddy Simulation (LES) and Filtered Mass Density Function (FMDF) approach. An idealized, piloted non-premixed jet flame with a one-step exothermic reaction $A + B \rightarrow P$ is studied with chemical kinetics parameters chosen to result in local extinction and reignition for the given Reynolds number. The mixing models to be studied include the Interaction by Exchange with the Mean (IEM), the Modified Curl (MC) and the Euclidean Minimum Spanning Tree (EMST). The relative performance of the mixing models is considered regarding extinction/reignition dynamics through examination of instantaneous flame images and scatter plots, as well as relevant statistical measures. Preliminary results towards efficient simulations of Sandia Flame D obtained using the above approach with realistic chemistry and the in situ adaptive tabulation or ISAT approach will be presented. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700