Bulletin of the American Physical Society
70th Annual Meeting of the APS Division of Fluid Dynamics
Volume 62, Number 14
Sunday–Tuesday, November 19–21, 2017; Denver, Colorado
Session Q37: Reacting Flows: Modeling and SimulationsCFD Reacting
|
Hide Abstracts |
Chair: Sadaf Sobhani, Stanford University Room: 303 |
Tuesday, November 21, 2017 12:50PM - 1:03PM |
Q37.00001: Flame-conditioned turbulence modeling for reacting flows Jonathan F. MacArt, Michael E. Mueller Conventional approaches to turbulence modeling in reacting flows rely on unconditional averaging or filtering, that is, consideration of the momentum equations only in physical space, implicitly assuming that the flame only weakly affects the turbulence, aside from a variation in density. Conversely, for scalars, which are strongly coupled to the flame structure, their evolution equations are often projected onto a reduced-order manifold, that is, conditionally averaged or filtered, on a flame variable such as a mixture fraction or progress variable. Such approaches include Conditional Moment Closure (CMC) and related variants. However, recent observations from Direct Numerical Simulation (DNS) have indicated that the flame can strongly affect turbulence in premixed combustion at low Karlovitz number. In this work, a new approach to turbulence modeling for reacting flows is investigated in which conditionally averaged or filtered equations are evolved for the momentum. The conditionally-averaged equations for the velocity and its covariances are derived, and budgets are evaluated from DNS databases of turbulent premixed planar jet flames. The most important terms in these equations are identified, and preliminary closure models are proposed. [Preview Abstract] |
Tuesday, November 21, 2017 1:03PM - 1:16PM |
Q37.00002: Application of Pareto-efficient combustion modeling framework to large eddy simulations of turbulent reacting flows Hao Wu, Matthias Ihme The modeling of turbulent combustion requires the consideration of different physico-chemical processes, involving a vast range of time and length scales as well as a large number of scalar quantities. To reduce the computational complexity, various combustion models are developed. Many of them can be abstracted using a lower-dimensional manifold representation. A key issue in using such lower-dimensional combustion models is the assessment as to whether a particular combustion model is adequate in representing a certain flame configuration. The Pareto-efficient combustion (PEC) modeling framework was developed to perform dynamic combustion model adaptation based on various existing manifold models. In this work, the PEC model is applied to a turbulent flame simulation, in which a computationally efficient flamelet-based combustion model is used in together with a high-fidelity finite-rate chemistry model. The combination of these two models achieves high accuracy in predicting pollutant species at a relatively low computational cost. The relevant numerical methods and parallelization techniques are also discussed in this work. [Preview Abstract] |
Tuesday, November 21, 2017 1:16PM - 1:29PM |
Q37.00003: Evaluation of model constant sensitivities for subfilter mixture fraction variance using adjoint and sensitivity derivative approaches Kevin Griffin, Michael Mueller The subfilter mixture fraction variance is a critical quantity in Large Eddy Simulation (LES) models for turbulent nonpremixed combustion. In the transport equation for the subfilter mixture fraction variance, two terms require modeling: the subfilter mixture fraction dissipation rate and the subfilter scalar flux. Conventional models for both of these terms require specification of model constants: the subfilter mixture fraction dissipation rate model constant and the subfilter turbulent Schmidt number. In this work, two approaches are compared for computing the sensitivity of the subfilter mixture fraction variance to these two model constants. In the first approach, explicit transport equations are derived and solved for the sensitivity derivatives. In the second approach, the sensitivity is obtained from the continuous adjoint equation of the subfilter mixture fraction variance. To stabilize the forward solution of the adjoint equation in LES, an efficient bootstrapping approach is proposed. The two methods are applied to a non-reacting nonpremixed bluff body flow, and the relative magnitudes of the two model constant sensitivities are discussed. The two methods are compared in terms of computational cost and apparent accuracy. [Preview Abstract] |
Tuesday, November 21, 2017 1:29PM - 1:42PM |
Q37.00004: Sensitivity Analysis to Turbulent Combustion Models for Combustor-Turbine Interactions Kenji Miki, Jeff Moder, Meng-Sing Liou The recently-updated Open National CombustionCode (Open NCC) equipped with alarge-eddy simulation (LES) is applied to model the flow field inside the Energy Efficient Engine (EEE) in conjunction with sensitivity analysis to turbulent combustion models. In this study, we consider three different turbulence-combustion interaction models, the Eddy-Breakup model (EBU), the Linear-Eddy Model (LEM) and the Probability Density Function (PDF)model as well as the laminar chemistry model. Acomprehensive comparison of the flow field and the flame structure will be provided. One of our main interests isto understand how a different model predicts thermal variation on the surface of the first stage vane. Considering that these models are often used in combustor/turbine communities, this study should provide some guidelines on numerical modeling of combustor-turbine interactions. [Preview Abstract] |
Tuesday, November 21, 2017 1:42PM - 1:55PM |
Q37.00005: Multiscalar Subfilter PDF Modeling for Large Eddy Simulation of Turbulent Piloted Flames with Inhomogeneous Inlets Bruce A. Perry, Michael E. Mueller Reduced-order manifold approaches for Large Eddy Simulation (LES) of turbulent combustion are usually combined with presumed subfilter Probability Density Function (PDF) models to close filtered quantities dependent on the thermochemical state. To further reduce computational cost, convolution of the manifold with the presumed PDF is conducted\textit{ a priori}. This work addresses two challenges associated with increasing the dimensionality of reduced-order manifolds to incorporate multiple mixture fractions for systems with multiple or inhomogeneous inlets. First, many models for the required multiscalar PDF have been proposed (Dirichlet, Conner-Mosimann, and five-parameter bivariate beta distributions), but the impact of these models has not been assessed in \textit{a posteriori }LES. Second, \textit{a priori }convolution of these presumed PDF models with the reduced-order manifolds becomes intractable. In this work, an \textit{a posteriori }analysis comparing results using various presumed PDF models in LES is conducted for a turbulent piloted jet flame with inhomogeneous inlets to assess the sensitivity to the subfilter PDF model form, leveraging a new `on-the-fly' strategy for convolution of the presumed PDF with the manifold. [Preview Abstract] |
Tuesday, November 21, 2017 1:55PM - 2:08PM |
Q37.00006: A Lagrangian mixing frequency model for transported PDF modeling. Hasret Turkeri, Xinyu Zhao In this study, a Lagrangian mixing frequency model is proposed for molecular mixing models within the framework of transported probability density function (PDF) methods. The model is based on the dissipations of mixture fraction and progress variables obtained from Lagrangian particles in PDF methods. The new model is proposed as a remedy to the difficulty in choosing the optimal model constant parameters when using conventional mixing frequency models. The model is implemented in combination with the Interaction by exchange with the mean (IEM) mixing model. The performance of the new model is examined by performing simulations of Sandia Flame D and the turbulent premixed flame from the Cambridge stratified flame series. The simulations are performed using the pdfFOAM solver which is a LES/PDF solver developed entirely in OpenFOAM. A 16-species reduced mechanism is used to represent methane/air combustion, and in situ adaptive tabulation is employed to accelerate the finite-rate chemistry calculations. The results are compared with experimental measurements as well as with the results obtained using conventional mixing frequency models. Dynamic mixing frequencies are predicted using the new model without solving additional transport equations, and good agreement with experimental data is observed. [Preview Abstract] |
Tuesday, November 21, 2017 2:08PM - 2:21PM |
Q37.00007: Can we predict differential diffusion effects on the turbulent flame structure in non-premixed flames? Nicholas Burali, Guillaume Blanquart Differential diffusion effects in turbulent non-premixed flames have been the subject of a vast body of work spanning over four decades. These effects have been shown to have a strong impact on the flame structure close to the burner exit plane, even at high turbulence intensities, and are observed to diminish with increasing downstream distance and increasing Reynolds numbers. Yet, the transition from molecular diffusion controlled mixing, to turbulence dominated transport in non-premixed flames remains poorly understood. The correct representation of these effects is important for phenomena which are sensitive to accurate scalar transport, such as soot formation. In recent work, we proposed a quantitative approach to extract ``effective'' Lewis numbers from conditional mean species profiles. This methodology is based on the popular flamelet assumption, and was applied to the ``Sandia flames'' experimental data set. In this work, the analysis is extended to Direct Numerical Simulation data. Statistics of the mixture fraction and its scalar dissipation are shown, differential diffusion effects are presented, and a budget analysis of the flamelet equations is discussed. [Preview Abstract] |
Tuesday, November 21, 2017 2:21PM - 2:34PM |
Q37.00008: Differential diffusion in tabulated chemistry: From model development to practical applications Jason Schlup, Guillaume Blanquart Tabulated chemistry has been used to reduce the computational cost of chemistry and help close terms in the transport equations of Large Eddy Simulations. Various tabulated chemistry models utilize a single Lewis number in the progress variable transport equation to account for differential diffusion effects. In this work, a tabulated chemistry model is extended to include thermal diffusion coefficients and non-unity Lewis numbers for more than just the progress variable or fuel. A derivation of the model is given, and a wide range of lean hydrogen/air flame configurations are examined, including one-, two-, and three-dimensional flames under laminar and turbulent conditions. Comparisons of flame speeds, surface areas, source terms, and flame curvatures are done between the previous and new tabulated chemistry models, and a practical application of the tabulated chemistry method is considered in a low-swirl burner. [Preview Abstract] |
Tuesday, November 21, 2017 2:34PM - 2:47PM |
Q37.00009: Assessment of two progress variable implementation procedures for supersonic combustion Wenhai Li, Foluso Ladeinde In the traditional non-premixed flamelet model, the reactive scalars are expressed in terms of the mixture fraction Z and its scalar dissipation rate $\chi $. However, the parameter set (Z, $\chi )$ cannot behave as a unique identifier to parameterize all flame thermochemical states on the S-shaped curve. Therefore, the progress variable C has been introduced to replace $\chi $ so that unique flamelet solutions can be determined by the values of (Z, C). However, $\chi $ can be identified as one of the most important parameters in non-premixed combustion since its mean value represents the rate of molecular scalar mixing and its fluctuation can directly influence the variance of Z. Therefore, $\chi $ should be kept as a control parameter in the flamelet table in order to correctly account for the compressive strain effects in high speed combustion. In this study, an interpolation-based progress variable implementation procedure is introduced so that (Z, $\chi )$ can still be used to obtain multiple flamelet solutions on each branch of the S-curve. This way, unique flamelet solutions can be obtained by an interpolation procedure based on C. [Preview Abstract] |
Tuesday, November 21, 2017 2:47PM - 3:00PM |
Q37.00010: Results from flamelet and non-flamelet models for supersonic combustion Foluso Ladeinde, Wenhai Li Air-breathing propulsion systems (scramjets) have been identified as a viable alternative to rocket engines for improved efficiency. A scramjet engine, which operates at flight Mach numbers around 7 or above, is characterized by the existence of supersonic flow conditions in the combustor. In a dual-mode scramjet, this phenomenon is possible because of the relatively low value of the equivalence ratio and high stagnation temperature, which, together, inhibits thermal choking downstream of transverse injectors. The flamelet method has been our choice for turbulence-combustion interaction modeling and we have extended the basic approach in several dimensions, with a focus on the way the pressure and progress variable are modeled. Improved results have been obtained. We have also examined non-flamelet models, including laminar chemistry (QL), eddy dissipation concept (EDC), and partially-stirred reactor (PaSR). The pressure/progress variable-corrected simulations give better results compared with the original model, with reaction rates that are lower than those from EDC and PaSR. In general, QL tends to over-predict the reaction rate for the supersonic combustion problems investigated in our work. [Preview Abstract] |
Tuesday, November 21, 2017 3:00PM - 3:13PM |
Q37.00011: The Geometry and Velocity of Propagating Fronts in Complex Flow Fields Saikat Mukherjee, Mark Paul We numerically investigate the velocity and geometry of propagating fronts in a range of complex flow fields generated by Rayleigh-B\'enard convection. The fronts are computed using a reaction-advection-diffusion equation with a Fischer-Kolmogorov-Petrovskii-Piskunov (FKPP) non-linearity. We explore the fronts in rectangular and cylindrical convection domains for a range of flow fields including straight-parallel rolls, concentric rolls, and patterns where the rolls exhibit spatiotemporal chaos. We are interested in the low Damk\"ohler number regime where the fluid dynamics plays an important role. We study the geometry of the front and compute its box counting dimension. The front is found to be fractal for the chaotic flow fields we explored. We also compute the variation of the front speed with the magnitude of the underlying fluid velocity. We connect with analytical results where possible and build a description of the propagating front using local geometric features of the convective pattern that include the local wavenumber, angle, and curvature of the convection rolls. [Preview Abstract] |
Tuesday, November 21, 2017 3:13PM - 3:26PM |
Q37.00012: A Quantum Algorithm for Modeling of Reactant Conversion Rate in Homogeneous Turbulence Guanglei Xu, Andrew Daley, Peyman Givi, Rolando Somma Developments in quantum computing techniques have the potential to revolutionise a range of computational subjects. With significant progress in the construction of necessary quantum hardware, it is important to identify possible applications in a wide range of fields. Turbulent reactive flows have been the subject of significant computational investigations. In particular, probability density function (PDF) methods simulated via Monte Carlo (MC) methods have been widely used for modeling and simulation of these flows. However, the cost of such computation in high precision parameter estimations can be enormous and problematic. We have developed a quantum algorithm for reacting flow simulations with a quadratic speed-up over classical MC methods in terms of the number of repetitions required to reach a certain accuracy. We analyze our algorithm as it would apply to simulate the limiting rate of reactant conversion rate in homogeneous turbulence via a transported PDF model. By computing statistical error scaling, we identify regimes in which our quantum algorithm would outperform MC methods. This is a starting point for identifying further applications of quantum algorithms in turbulent combustion, and to analyze the hardware requirements for applications in this area of science. [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