Stéphane Jamme

Direct Numerical Simulation of the Interaction between a Shock Wave and Various Types of Isotropic Turbulence

Direct Numerical Simulation (DNS) is used to study the interaction between normal shock waves of moderate strength (M1= 1.2 and M1 = 1.5) and isotropic turbulence. A complete description of the turbulence behaviour across the shock is provided and the influence of the nature of the incoming turbulence on the interaction is investigated. The presence of upstream entropy fluctuations satisfying the Strong Reynolds Analogy enhances the amplification of the turbulent kinetic energy and transverse vorticity variances across the shock compared to the solenoidal (pure vorticity) case. Budgets for the fluctuating-vorticity variances are computed, showing that the baroclinic torque is responsible for this additional production of transverse vorticity. More reduction of the transverse Taylor microscale and integral scale is also observed in the vorticity-entropy case while no influence can beseen on the longitudinal Taylor microscale. When the upstream turbulence is dominated by acoustic and vortical fluctuations, less amplification of the kinetic energy (for Mach numbers between 1.25 and 1.8), less reduction of the transverse microscale and more amplification of the transverse vorticity variance are observed through the shock compared to the solenoidal case. In all cases, the classic estimation of Batchelor relating the dissipation rate and the integral scale of the flow proves to be invalid. These results are obtained with the same numerical tool and similar flow parameters, and they are in good agreement with Linear Interaction Analysis (LIA).

Evaluation of the Thrust Recovery of an Aircraft Flapped Outflow Valve

This paper presents a detailed study of a flapped outflow valve. Well-known as part of the pressurization system of aircraft, this type of valve is also designed for thrust recovery even if the efficiency of the system has never been demonstrated. Aground experimental test rig is first designed to provide global and local measurements to be used as validation data for numerical simulations. Once the validation of the numerical approach is achieved on a ground configuration, additional three-dimensional computations are then conducted for cruise conditions. They lead to a reliable estimation of thrust recovery as well as interesting insight in the aerodynamic behavior of the flow across the valve and its associated three-dimensional effects.

Direct Numerical Simulation of the Interaction Between a Shock Wave and Anisotropic Turbulence

Direct Numerical Simulations of the interaction between a shock wave at Mach 1.5 and several anisotropic turbulent fields are described and compared to the isotropic reference case. The various anisotropic turbulent fields considered upstream of the shock are of axisymetric type. The amount of the longitudinal normal Reynolds stress component is found to drive the evolution of the kinetic energy in the near field behind the shock wave as well as the modification of the anisotropic state across the shock. No new fundamental mechanism is found compared to the isotropic situation. Further insight is however provided concerning the near field evolution of the turbulent kinetic energy where two different zones have to be distinguished. Length scales evolutions are also considered and the dissipation length scale is found to behave in a radically different way than the integral scale.

Laser Doppler Velocimetry Measurements in Turbulent Gaseous Mixing Induced by the Richtmyer–Meshkov Instability: Statistical Convergence Issues and Turbulence Quantification

A statistical characterization of the turbulent flow produced in a vertical shock tube dedicated to the study of the Richtmyer-Meshkov instability (RMI) is carried out using Laser Doppler Velocimetry (LDV), time-resolved Schlieren images and pressure histories. The time evolution of the phase-averaged velocity field and the fluctuating velocity levels produced behind the shock wave are first investigated for different configurations of a pure air, homogeneous medium. This allows us to determine the background turbulence of the experimental apparatus. Second, the RMI-induced turbulent Air/SF6 mixing zone (TMZ) is studied both in its early stage of development and after its interaction with a reflected shock wave (reshock phenomenon). Here the gaseous interface is initially produced by a thin nitrocellulosic membrane trapped between two grids. One of the most consistent issue regarding such a process is the generation of a large number of fragments when the incident shock wave crosses the interface. These fragments are likely to corrupt the optical measurements and to interact with the flow. This work seeks to clarify the influence of these fragments on the statistical determination of the velocity field. In particular it is shown that statistical convergence cannot be achieved when the fragments are crossing the LDV measurement volume, even if a significant number of identical experiments are superimposed. Some specific locations for the LDV measurements are however identified to be more favourable than others in the Air/SF6 mixing configuration. This finally allows us to quantify the surplus of turbulence induced by the reshock phenomenon.

Experimental and numerical investigation of the growth of an air/SF6 turbulent mixing zone in a shock tube

Shock-induced mixing experiments have been conducted in a vertical shock tube of 130mm square cross section located at ISAE. A shock wave traveling at Mach 1.2 in air hits a geometrically disturbed interface separating air and SF6, a gas five times heavier than air, filling a chamber of length L up to the end of the shock tube. Both gases are initially separated by a 0.5 lm thick nitrocellulose membrane maintained parallel to the shock front by two wire grids: an upper one with mesh spacing equal to either ms=1.8mm or 12.1 mm, and a lower one with a mesh spacing equal to ml=1 mm. Weak dependence of the mixing zone growth after reshock (interaction of the mixing zone with the shock wave reflected from the top end of the test chamber) with respect to L and ms is observed despite a clear imprint of the mesh spacing ms in the schlieren images. Numerical simulations representative of these configurations are conducted: the simulations successfully replicate the experimentally observed weak dependence on L, but are unable to show the experimentally observed independence with respect to ms while matching the morphological features of the schlieren pictures.

Experimental Characterization of Turbulence Produced in a Shock Tube: A Preliminary Work for the Study of the Turbulent Gaseous Mixing Induced by the Richtmyer-Meshkov Instability

The Richtmyer-Meshkov Instability (RMI) occurs in several physical and technological processes such as supernova explosion, supersonic combustion, detonics or inertial confinement fusion. This instability develops when interfacial perturbations, between two fluids of different densities, grow because of a shock wave induced impulsive acceleration. The basic mechanism for the initial growth of perturbations on the interface is the baroclinic generation of vorticity which results from the misalignment of the pressure and density gradientswhen the shock crosses the interface. Early time linear and following nonlinear growth of the RMI have been, and are still widely investigated, either theoretically, numerically and experimentally [1]. Nevertheless, experimental investigation of the Turbulent Mixing Zone (TMZ) induced by a rapidly growing RMI is still nowadays poorly documented, even if we can mention for instance the work of Leinov et al. [2] who characterized the growth of the MZ with time following the passage of the re-shock (with an emphasis on the influence of the initial amplitude of the MZ and the reshock strength), and the study of Poggi et al. [3] in which the production of turbulence by the TMZ has been investigated in a vertical shock tube using two-components Laser Doppler Velocimetry (LDV).

On the Return to Isotropy of Compressible Anisotropic Turbulence

In the present paper, the behaviour of compressible anisotropic turbulence is investigated using Direct Numerical Simulation (DNS). The computational code is based on a finite volume formulation of the explicit predictor-corrector MacCormack scheme. Decay simulations of homogeneous turbulence are carried out through specifying a constant mean flow and adding an initial anisotropic fluctuating velocity field. A specific parameterization of anisotropy allows us to explore several anisotropic decay situations. The impact of the nature of the initial anisotropic turbulence on turbulent kinetic energy decay is underlined. A specific decomposition of the turbulent kinetic energy budget permits to study the return to-isotropy phenomenon.

A Study of Sheared Turbulence/Shock Interaction: Velocity Fluctuations and Enstrophy Behaviour

Direct Numerical Simulations of the idealized interaction of a normal shock wave with several turbulent shear flows are conducted. We analyse the behaviours of velocity and vorticity fluctuations and compare them to what happens in the isotropic situation. Investigation of the budgets of these quantities allows to isolate the mechanisms underlying the physics of the interaction, and reveals the importance of enthalpic production and baroclinic torque in such flows.

Experimental Determination of the Growth Rate of Richtmyer-Meshkov Induced Turbulent Mixing after Reshock

The Richtmyer-Meshkov Instability (RMI) develops when a shock wave impulsively accelerates an initially perturbed interface between two gases of different density, promoting their mixing inside a delimited zone, hereafter denoted the mixing zone (MZ). This mixing is a key issue for inertial confinement fusion process. It also finds applications in many different scientific and engineering issues, e.g. in supernova explosion or supersonic combustion [1]. In the context of inertial confinement fusion, the characterization of the RMI-induced mixing zone initiated by a shock and further amplified by a reshock is largely based on the temporal evolution of integral parameters such as the width of the MZ. This is classically achieved through shock tube experiments involving the time-resolved acquisition of Schlieren images.

Compressibility effects on the return to isotropy of homogeneous anisotropic turbulence

In this study, we investigate the behaviour of decaying compressible anisotropic turbulence using Direct Numerical Simulations (DNS). High-order schemes are employed to accurately solve the full three-dimensional Navier-Stokes equations. The study of compressibility effects is based on decay simulations of anisotropic turbulence with two parametric studies, one concerning the ratio between the compressible part and the total kinetic energy, and the other one concerning the turbulent Mach number. We focuse our attention on the return to isotropy phenomenon. We analyse our results in respect with previous studies.