Comparison of numerical methods and combustion models for LES of a ramjet

Ramjets are very sensitive to instabilities and their numerical predictions can only be addressed adequately by Large Eddy Simulation (LES). With this technique, solvers can be implicit or explicit and handle structured, unstructured or hybrid meshes, etc. Turbulence and combustion models are other sources of differences. The impact of these options is here investigated for the ONERA ramjet burner. To do so, two LES codes developed by ONERA and CERFACS compute one stable operating condition. Preliminary LES results of the two codes underline the overall robustness of LES. Mean flow features at the various critical sections are reasonably well predicted by both codes. Disagreement mainly appear in the chamber where combustion positions differ pointing to the importance of the combustion and subgrid mixing models. The two LES produce different energy containing motions. With CEDRE, a low frequency dominates while AVBP produces different ranges of low frequencies that can be linked with acoustic modes of the configuration. To cite this article: A. Roux et al., C. R. Mecanique 337 (2009).     

Using LES to Study Reacting Flows and Instabilities in Annular Combustion Chambers

Great prominence is put on the design of aeronautical gas turbines due to increasingly stringent regulations and the need to tackle rising fuel prices. This drive towards innovation has resulted sometimes in new concepts being prone to combustion instabilities. In the particular field of annular combustion chambers, these instabilities often take the form of azimuthal modes. To predict these modes, one must compute the full combustion chamber, which remained out of reach until very recently and the development of massively parallel computers. Since one of the most limiting factors in performing Large Eddy Simulation (LES) of real combustors is estimating the adequate grid, the effects of mesh resolution are investigated by computing full annular LES of a realistic helicopter combustion chamber on three grids, respectively made of 38, 93 and 336 million elements. Results are compared in terms of mean and fluctuating fields. LES captures self-established azimuthal modes. The presence and structure of the modes is discussed. This study therefore highlights the potential of LES for studying combustion instabilities in annular gas turbine combustors.     

Coupling tabulated chemistry with Large Eddy Simulation of turbulent reactive flows

A new modeling strategy is developed to introduce tabulated chemistry methods in the LES of turbulent premixed combustion. The objective is to recover the correct laminar flame propagation speed of the filtered flame front when the subgrid scale turbulence vanishes. The filtered flame structure is mapped by 1D filtered laminar premixed flames. Closure of the filtered progress variable and the energy balance equations are carefully addressed. The methodology is applied to 1D and 2D filtered laminar flames. These computations show the capability of the model to recover the laminar flame speed and the correct chemical structure when the flame wrinkling is completely resolved. The model is then extended to turbulent combustion regimes by introducing subgrid scale wrinkling effects on the flame front propagation. Finally, the LES of a 3D turbulent premixed flame is performed. To cite this article: R. Vicquelin et al., C. R. Mecanique 337 (2009).     

Injection coupling with high amplitude transverse modes: Experimentation and simulation

High frequency combustion instabilities have technical importance in the design of liquid rocket engines. These phenomena involve a strong coupling between transverse acoustic modes and combustion. They are currently being investigated by combining experimentation and numerical simulations. On the experimental level, the coupling is examined in a model scale system featuring a multiple injector combustor (MIC) comprising five coaxial injectors fed with liquid oxygen and gaseous methane. This system is equipped with a novel VHAM actuator (Very High Amplitude Modulator) which comprises two nozzles and a rotating toothed wheel blocking the nozzles in an alternate fashion. This device was designed to obtain the highest possible levels of transverse oscillation in the MIC. After a brief review of the VHAM, this article reports cold flow experiments using this modulator. Velocity maps obtained under resonant conditions using the VHAM are examined at different instants during a cycle of oscillation. Experimental data are compared with numerical pressure and velocity fields obtained from an acoustic solver. The good agreement observed in the nozzle vicinity indicates that numerical simulations can be used to analyze the complex flow field generated by the VHAM. To cite this article: Y. Mery et al., C. R. Mecanique 337 (2009).     

Dynamics of premixed confined swirling flames

Considerable effort is currently being extended to examine the fundamental mechanisms of combustion instabilities and develop methods allowing predictions of these phenomena. One central aspect of this problem is the dynamical response of the flame to incoming perturbations. This question is examined in the present article, which specifically considers the response of premixed swirling flames to perturbations imposed on the upstream side of the flame in the feeding manifold. The flame response is characterized by measuring the unsteady heat release induced by imposed velocity perturbations. A flame describing function is defined by taking the ratio of the relative heat release rate fluctuation to the relative velocity fluctuation. This quantity is determined for a range of frequencies and for different levels of incoming velocity perturbations. The flame dynamics is also documented by calculating conditional phase averages of the light emission from the flame and taking the Abel transform of these average images to obtain the flame geometry at various instants during the cycle of oscillation. These data can be useful to the determination of possible regimes of instability. To cite this article: P. Palies et al., C. R. Mecanique 337 (2009).     

The structure of multidimensional strained flames under transcritical conditions

Strained flames are commonly used to study the structure of reactive layers and describe the local properties of turbulent combustion. This model is attractive because constant strain rate flames only depend on a transverse coordinate and can be treated as a one-dimensional problem. This configuration is considered in a multidimensional context in which the strained flow is obtained by two counterflowing streams of reactants. It is used to examine the structure of transcritical strained flames in which one or two reactants are injected at a high pressure exceeding the critical value while their temperature is below the critical value. Calculations are carried out in a two-dimensional domain to test numerical models developed for multidimensional simulations and test thermodynamic and transport models devised to deal with high pressure real gas effects. Multidimensional strained flame calculations carried out in this study serve to check the validity of a new version of a Navier–Stokes flow solver (AVBP) conceived to deal with transcritical combustion of interest to liquid propellant rocket applications. This article describes the basic elements of such simulations and discusses results of calculations. It is shown that the calculated multidimensional strained flames have the expected features in terms of structure and response to the imposed strain rate. To cite this article: L. Pons et al., C. R. Mecanique 337 (2009).     

A DES method applied to a Backward Facing Step reactive flow

Hybrid Reynolds Averaged Navier Stokes–Large Eddy Simulation is a trend which is becoming of common use in aerodynamics but has seldom been employed to simulate reactive flows. Such methods, like the Delayed Detached Eddy Simulation (DDES) presented in this article, have been created to treat near wall flows with a RANS approach while switching to LES in the separated flow region. It is indeed an affordable solution to simulate complex and unsteady compressible flows and to have access to accurate skin friction and wall thermal fluxes. In order to validate this technique in combustion, we chose a simple and well documented Backward Facing Step combustor. To account for turbulent combustion a Dynamic Thickened Flame was used. The results obtained on this case show a good agreement with the experimental database and are of the same quality as LES in the separated region for both inert and reactive flows. To cite this article: B. Sainte-Rose et al., C. R. Mecanique 337 (2009).     

Chemical kinetics modeling and LES combustion model effects on a perfectly premixed burner

Chemical kinetics modeling and coupling with turbulent combustion models for compressible Large Eddy Simulations (LES) is a critical issue. Accurate flow predictions can only be guaranteed if the coupling is well mastered. In a first attempt to qualify the effect of each model, the case of a lean premixed swirled combustor with comprehensive measures is targeted (species mass fractions and temperature fields). For the investigation, two turbulent combustion models are considered. The first model relies on a presumed PDF approach coupled to a look-up chemistry table obtained with a reduced chemical scheme. The second model makes use of the thickened flame approach using the same reduced chemical scheme but with reaction rates computed explicitly as the computation advances. Then, to estimate kinetic schemes reduction effects, the first model is compared to a third one, with the same PDF approach, but coupled to a look-up chemistry table obtained with a complete chemical scheme. All LES are very close to each other. The main difference between the different predictions relies on CO mass fractions. Although they are all able to return good outlet mass fractions, CO values inside the flame are different depending on the model used. To cite this article: G. Albouze et al., C. R. Mecanique 337 (2009).     

Direct numerical simulation and global stability analysis of three-dimensional instabilities in a lid-driven cavity

The first bifurcation in a lid-driven cavity characterized by three-dimensional Taylor–Görtler-Like instabilities is investigated for a cubical cavity with spanwise periodic boundary conditions at Re=1000. The modes predicted by a global linear stability analysis are compared to the results of a direct numerical simulation. The amplification rate, and the shape of the three-dimensional perturbation fields from the direct numerical simulation are in very good agreement with the characteristics of the steady S1 mode from the stability analysis, showing that this mode dominates the other unstable unsteady modes. To cite this article: J. Chicheportiche et al., C. R. Mecanique 336 (2008).     

Use of Faraday instabilities to enhance fuel pulverisation in air-blast atomisers

The atomization of liquids into a spray is an important process in many industrial applications and particularly in the aero-engine sector. Conventional air-blast injectors in aircraft engines today use aerodynamic shearing effects to atomize the liquid fuel. However, at operating conditions where the air velocity is below 30 m/s (such as ground start and high altitude restart) the atomization quality is poor. Consequently combustion is less efficient with high pollutant emissions. The objective of this study is to validate a new concept of injector which couples the shearing effects with the principle of ultrasonic atomization. The latter consists of using piezoelectric actuators to generate the oscillations of a wall in contact with the liquid film. This excitation perpendicular to the liquid film surface creates Faraday instabilities at the liquid/air interface. Amplitudes higher than a defined threshold value induce the break-up of ligaments and the formation of droplets. To cite this article: M. Boukra et al., C. R. Mecanique 337 (2009).     

Stabilization of non-premixed flames in supersonic reactive flows

A Lagrangian framework is set out to describe turbulent non-premixed combustion in high speed coflowing jet flows. The final aim is to provide a robust computational methodology to simulate, in various conditions, the underexpanded GH2/GO2 torch jet that is used to initiate combustion in an expander cycle engine. The proposed approach relies on an early modelling proposal of Borghi and his coworkers. The model is well suited to describe finite rate chemistry effects and its recent extension to high speed flows allows one to take the influence of viscous dissipation phenomena into account. Indeed, since the chemical source terms are highly temperature sensitive, the influence of viscous phenomena on the thermal runaway is likely to be all the more pronounced since the Mach number values are high. The validation of the extended model has been recently performed through the numerical simulation of two distinct well-documented experimental databases. Only a brief summary of this preliminary validation step is provided here. The main purpose of the present work is to proceed with the numerical simulation of geometries that bring together the essential peculiarities of the underexpanded GH2/GO2 torch. The behavior of the corresponding supersonic coflowing jet flames for various conditions is discussed in the light of computational results. To cite this article: J.-F. Izard, A. Mura, C. R. Mecanique 337 (2009).     

Large-Eddy Simulation of transcritical flows

The present study uses the LES code AVBP, developed at CERFACS, to simulate transcritical flows. Real gas effects are accounted for by the use of a cubic equation of state, in conjunction with appropriate viscosity and thermal conductivity coefficients. First a single nitrogen round jet at supercritical pressure injected in a gaseous reservoir is simulated. Two cases are considered, one demonstrating a transcritical injection (high density injection), the other being directly injected at supercritical temperature (lower density injection). Comparison with available measurements shows good agreement. Finally, the simulation of a reacting case from the Mascotte bench (ONERA) is performed, consisting in a single coaxial injector injecting transcritical oxygen and supercritical hydrogen in a 60 bar chamber. Mean flow characteristics are in good agreement with the experimental observations of OH^* emission, whereas temperature comparisons are more difficult to interpret. To cite this article: T. Schmitt et al., C. R. Mecanique 337 (2009).     

Eulerian and Lagrangian Large-Eddy Simulations of an evaporating two-phase flow

Large-Eddy Simulations (LES) of an evaporating two-phase flow in an experimental burner are performed using two different solvers, CDP from CTR-Stanford and AVBP from CERFACS, on the same grid and for the same operating conditions. Results are evaluated by comparison with experimental data. The CDP code uses a Lagrangian particle tracking method (EL) while the code AVBP can be coupled either with a mesoscopic Eulerian approach (EE) or with a Lagrangian method (EL). After a validation of the purely gaseous flow in the burner, liquid-phase dynamics, droplet dispersion and fuel evaporation are qualitatively and quantitatively evaluated for three two-phase flow simulations. They are respectively referred as: CDP-EL, AVBP-EE and AVBP-EL. The results of the three simulations show reasonable agreement with experiments for the two-phase flow case. To cite this article: J.M. Senoner et al., C. R. Mecanique 337 (2009).     

Extended three-dimensional digital image correlation (X3D-DIC)

A correlation algorithm is proposed to measure full three-dimensional displacement fields in a three-dimensional domain. The chosen kinematic basis for this measurement is based on continuous finite-element shape functions. It is furthermore proposed to account for the presence of strong discontinuities, similarly to extended finite element schemes, with a suited enrichment of the kinematics with discontinuities supported by a (crack) surface. An optimization of the surface geometry is proposed based on correlation residuals. The procedure is applied to analyze one loading cycle of a fatigue-cracked nodular graphite cast iron sample by using computed tomography pictures. Subvoxel crack openings are revealed and measured. To cite this article: J. Réthoré et al., C. R. Mecanique 336 (2008).     

Effect of multiperforated plates on the acoustic modes in combustors

The analytical model derived by Howe assessing the acoustic effect of perforated plates has been implemented in a 3D Helmholtz solver. This solver allows one to compute the acoustic modes of industrial chambers taking into account the multiperforated plates present for the cooling of the walls. An academic test case consisting of two coaxial cylinders, with the inner one being perforated is used to validate the implementation in the general purpose AVSP code. This case is also used to show the effects of the presence of the plates. In particular, the sensitivity of the acoustic damping to the bias flow speed will be studied. A maximum absorption speed is shown, and the behaviour towards an infinite speed will be illustrated by the academic case. Computations are also conducted in the case of an industrial helicopter chamber. The value of the maximum absorption speed is discussed to explain why the modes are in fact not much absorbed by the perforated plates, and that the frequencies are the same as for walls. To cite this article: E. Gullaud et al., C. R. Mecanique 337 (2009).     

Nonlinear models of vehicular traffic flow – new frameworks of the mathematical kinetic theory

This paper deals with the design of mathematical frameworks for the modeling of traffic flow phenomena by suitable developments of classical models of the kinetic theory. Various types of evolution equations are deduced, and different mathematical structures are proposed toward conceivable applications. To cite this article: M. Delitala, C. R. Mecanique 331 (2003).     

Resolution of linear viscoelastic equations in the frequency domain using real Helmholtz boundary integral equations

Boundary integral equations are well suitable for the analysis of seismic waves propagation in unbounded domains. Formulations in elastodynamics are well developed. In contrast, for the dynamic analysis of viscoelastic media, there are very seldom formulations by boundary integral equations. In this Note, we propose a new and simple formulation of time harmonic viscoelasticity with the Zener model, which reduces to classical elastodynamics if a compatibility condition is satisfied by boundary conditions. Intermediate variables which satisfy the classical elastodynamic equations are introduced. It makes it possible to utilize existing numerical tools of time harmonic elastodynamics. To cite this article: S. Chaillat, H.D. Bui, C. R. Mecanique 335 (2007).     

Espaces d'écoulements dits « universels », 3

Universal motions with uniform steady vorticity form a corolla of linear spaces derived from rigid body motions. Closely related to potential flows, they satisfy two extensions of Lagrange theorem, are investigated with the help of complex functions, as stand celebrated when be plane. They take place in hydrodynamics, aerodynamics, geophysics, astrophysics, turbulence, physics of plasmas and superfluid helium. In all the cases, arbitrary unsteady span-wise translations permit to generalise as well as to exhibit helical or 3D universal motions. Three misunderstood periodic flows illustrate our purpose, as they approach shear instabilities in numerous fluids. To cite this article: M. Bouthier, C. R. Mecanique 332 (2004).     

Nonlinear behavior of matrix-inclusion composites under high confining pressure: application to concrete and mortar

This paper is devoted to a micromechanics-based simulation of the response of concrete to hydrostatic and oedometric compressions. Concrete is described as a composite made up of a cement matrix in which rigid inclusions are embedded. The focus is put on the role of the interface between matrix and inclusion which represent the interfacial transition zone (ITZ). A plastic behavior is considered for both the matrix and the interfaces. The effective response of the composite is derived from the modified secant method adapted to the situation of imperfect interfaces. To cite this article: T.H. Le et al., C. R. Mecanique 336 (2008).     

Weighted Korn inequalities for thin-walled elastic structures

Appropriate weighted norms in H¹ are presented such that the Korn type inequality is asymptotically sharp with respect to relative thickness and stiffness of the elastic plates. The weights depend crucially on the geometric structure of the plates' junction. To cite this article: O.V. Izotova et al., C. R. Mecanique 334 (2006).