• Energy Research

Abstract A sub-grid scale closure for Large Eddy Simulation (LES) of turbulent combustion based on physical-space filtering of laminar flames is discussed. Applied to an unstructured grid, the combustion LES filter size is not fixed in this novel approach devoted to LES with refined meshes, but calibrated depending on the local level of unresolved scalar fluctuations. The context is premixed or stratified flames, the derived model relies on four balance equations for mixture fraction and its variance, and a progress variable and its variance. The proposed formalism is based on a presumed probability density function (PDF) derived from the filtered flames. Closures for the terms of the equations that are unresolved over LES grids are achieved through the PDF. The method uses flamelet tabulated detailed chemistry and is first applied to the simulation of laminar flames (1D and 2D) over various grids for validation, before simulating a turbulent burner studied experimentally by Sweeney et al. (2012). Since this burner also features differential diffusion effects, the numerical model is modified to account for accumulation of carbon in the recirculation zone behind the bluff-body. A differential diffusion number based on the gradient of residence times is proposed, in an attempt to globally quantify differential diffusion effects in burners.

International audience Dynamic procedures to automatically determine flame wrinkling factors from known resolved fields in large eddy simulations of turbulent premixed combustion are investigated from a priori tests processing a DNS database of a turbulent swirled flame. These flame wrinkling factors measure the ratio of total to resolved flame surfaces in the filtering volume and enter directly or indirectly into various flamelet combustion models through the sub-grid scale turbulent flame speed. They are usually modeled by algebraic expressions derived assuming equilibrium between turbulence motions and flame dynamics, a situation generally not reached during early stages of flame developments. Dynamic models then appear as a promising alternative to flame wrinkling factor or flame surface density balance equations to handle out-of-equilibrium situations. Attention is paid to three key requirements: (i) the correct prediction of propagating laminar flame fronts; (ii) the replacement of the averaging volume introduced to determine resolved and test-filtered flame wrinkling factors by a Gaussian operator easier to implement on unstructured meshes and/or massively parallel machines; (iii) the use of a local model parameter, evolving both in space and time. The two first requirements suggest basing the procedure on flame surface conservation instead of on chemical reaction rates. The saturated form of the Charlette et al. efficiency function [1], ΞΔ=β(Δ/δl),ΞΔ=(Δ/δl)β, where Δ is the filter width and δl the flame thickness, is found to be very well suited to dynamic determination of the model parameter β, easy to implement and very robust in practice, as confirmed by preliminary a posteriori tests.

Abstract A quasi–cubic meso-scale air/methane whirl flow combustion chamber, with no moving parts, is analyzed by means of high-fidelity Large-Eddy Simulations (LES). In order to consider partially premixed flames and differential diffusion process into the combustor, 3D LES computations include semi-detailed chemical kinetics mechanism and complex transport. They show that the reacting flow structure, flame topology and global efficiency are in line with previous experimental and less detailed numerical modeling. Combustion is stabilized by a Central Recirculation Zone (CRZ) and mainly takes place in premixed lean regime. Wall heat losses are also estimated and incomplete combustion zones are identified. Performances improvement was tested with hydrogen addition to fuel mixture. A small amount of hydrogen shows a global efficiency rise while keeping the whirl flow topology. The addition of a large amount of hydrogen implies a change in flame topology leading to a less complete combustion and pollutant formation. These results are in agreement with experimental data.

Abstract Swirl burners are widely used in aeronautical engines and industrial gas turbines. These burners enhance the flame stabilization by bringing back hot products to the reactive zone and their topology ensures a good compactness. However, the outer recirculation zones created in these burners induce wall heat loss that affects the flame structure. This paper proposes a numerical strategy based on Large-Eddy Simulation (LES) and non-adiabatic boundary conditions with a skeletal chemistry approach coupled to the Dynamic Thickened Flame model (TFLES). Simulations were performed on the PRECCINSTA burner with meshes up to 877 millions elements. Results demonstrated the impact of the addition of wall heat loss on the lift-off of the external flame front, leading to a flame topology change from a M-shape to a V-shape. This lift-off height increases with the grid resolution showing the strong influence of the mesh resolution on the flame topology. Comparisons of the non-adiabatic LES results on the finest mesh with the experimental data show an unprecedented agreement. The flame quenching process is analyzed and exhibits the role of strain rate which controls the level of penetration of cooled products within the inner reaction zone.

Accurate simulation of wind turbine wakes is critical for the optimization of turbine efficiency and prediction of fatigue loads. These wakes are three-dimensional, complex, unsteady and can evolve in geometrically complex environments. Modeling these flows calls thus for high-quality numerical methods that are able to capture and transport thin vortical structures on an unstructured grid. It is proposed here to assess the performances of a fourth-order finite-volume LES solver to perform massively parallel scale-resolving simulations of wind turbines wakes. In this framework, the actuator line method that takes the effect of the wind turbine blades on the flow into account is implemented. It is demonstrated that both near and far parts of the turbine wakes are accurately modeled as well as geometrical details. The methodology is assessed on two different test cases and validated with experimental results. It is demonstrated that the flow predictions are of equivalent quality on both structured and unstructured grids. The influence of the geometrical details (e.g. nacelle and tower) on the wake development as well as the influence of the discretization scheme are also investigated.

Abstract Large-Eddy Simulations (LES) and Direct Numerical Simulation (DNS) are applied to the analysis of a swirl burner operated with a lean methane–air mixture and experimentally studied by Meier et al. [ 19 ]. LES is performed for various mesh refinements, to study unsteady and coherent large-scale behavior and to validate the simulation tool from measurements, while DNS enables to gain insight into the flame structure and dynamics. The DNS features a 2.6 billion cells unstructured-mesh and a resolution of less than 100 microns, which is sufficient to capture all the turbulent scales and the major species of the flame brush; the unresolved species are taken into account thanks to a tabulated chemistry approach. In a second part of the paper, the DNS is filtered at several filter widths to estimate the prediction capabilities of modeling based on premixed flamelet and presumed probability density functions. The similarities and differences between spatially-filtered laminar and turbulent flames are discussed and a new sub-grid scale closure for premixed turbulent combustion is proposed, which preserves spectral properties of sub-filter flame length scales. All these simulations are performed with a solver specifically tailored for large-scale computations on massively parallel machines.

Abstract Large eddy simulation of the two stratified nonswirling configurations of the Cambridge burner studied by Sweeney et al. (2012) is presented. The sub-grid-scale combustion closure relies on a physical space filtering operation with a filter size determined locally depending on the resolved and sub-grid-scale flame properties, which is discussed in a companion paper. Similarly to the premixed configuration of the same burner, the modeling reproduces the differential diffusion effects leading to accumulation of carbon and an enhancement of mixture fraction in the recirculation zone, an effect that is less pronounced than in the fully lean premixed case, because of the modification of the topology of the reaction zone that is induced by the mixture stratification. The study of the LES combustion regimes shows that the reaction zones develop under a quite large range of flame topologies, from wrinkled flamelets up to thin reaction zones. Instantaneous and time-averaged LES data were analyzed to extract information concerning the degree of stratification and the orientation of flame and mixing vectors. A decomposition of the flame response into premixed, diffusion, and partially premixed flamelets is performed, to conclude that the premixed mode dominates close to the burner, with a partially premixed burning regime further downstream. Overall, the length scales associated with stratification were found to be much larger than that of the reaction zone and flame, resulting in a quasi-homogeneous propagation, predominantly in a back supported stratified combustion regime. Overall good agreement between simulation and measurements was obtained for either configurations.