The behavior of premixed flames has been examined by many authors. In fact the problem of combustion which develops in a turbulent medium depends on two scalings. One makes reference to the scales of the flame the other one is related to the turbulent field. Comparisons between these two scalings allow us to identified what sort of regime is expected.In this paper we first study the development of a material surface which may be identify with a flame front under rather severe conditions. An analytical approach is first used. Hereafter a numerical simulation will be introduced. The role of a fine grained turbulence is more active on the extension of the surface than large structures. To a large extent big eddies convey the surface without distorting it. The risks of extinction are generally predicted by making comparisons between the scales of the flame and the scales of the turbulent field starting from a direct simulation. Poincot et al show that the smallest structures are not responsible for the extinction: intermediate structures are more efficient than the smallest ones. In a previous paper the role of these structures was examined: the distorting mechanism are acting in a cumulative way. The life time of the smallest structures is too short to have them playing a decive role in the extinction process. Intermediate sized structures are less active but they strain the flame during a longer period. This idea requires a detailed description of the turbulent field. That is made possible by using the beta model which accounts for the location of turbulent structures whose ranks in the whole sequence is termed << n >>. The cumulative role of the velocity gradients is given as a function of << n >>. The influence of the intermediate structure on the extinction process is thereby emphasized. Finally the beta model is also used to describe the domain of distributed combustion zones. Flames propagate in limited regions of space. These regions are disconnected from each other. The turbulent diffusion process is very active to create new burning zones. This diffusion process is closely related to both the intensity and the length scale of the turbulent field. That explains the dominant role of turbulence and the adaptation of the burning rate to the rotation speed in combustion engines.

A new configuration of homogeneous turbulent shear flow is investigated experimentally and numerically. The flow is created by superposition of plane strain and solid body rotation of equal rates. This provides uniform shear in planes perpendicular to the main flow direction. Very satisfactory homogeneity conditions are thus achieved. The evolution of the Reynolds stresses and associated integral length scales are determined by hot-wire measurements. The shear gives rise to a definite decrease of the turbulent energy decay and a pronounced anisotropisation of the Reynolds stresses and length scales, as compared to the evolution without shear. The experimental results are compared with numerical predictions based on a two-point closure approach of the EDQNM type. The numerical results are in good qualitative agreement with the measurements, especially for the evolution of the length scales.

Turbulence in solid-body rotation is generated by a flow of air passing through a rotating cylinder containing a dense honeycomb structure and a turbulence-producing grid. The velocity field is probed downstream of this device by hot-wire probes. Using the statistical quantities characterizing the fluctuating field, we show that the rotation affects mainly the components normal to the rotation axis and that these effects are triggered when the Rossby numbers constructed from macroscopic turbulent quantities, are less than unity. These results are discussed in the framework of other available experimental results on the subject. A theoretical interpretation, chiefly based on spectral analysis, is then proposed to explain the trends of the observations.

Scientific and technical approach concerning the behaviour of flames developing in a turbulent medium are related in many recent papers. On the whole the problem is very complex!The chemical reaction develops inside a turbulent flow which requires a double scaling. Characteristic times and characteristic lengths have to be defined for both flame and the turbulent fields. With a view to enlarging these comparisons a spectral analyses of the turbulent field is proposed. It is widely supported by previous experimental data. The flame can be acted upon by an external turbulent fields. That supposes the flame to be thicker that the smallest turbulent structures in connection with the Kolmogorov scale. With increasing Reynolds numbers turbulent structures penetrate the flame front, they can disturb the preheat zone or event the chemical zone. The passage of a flame front regime to the case of a chemical reaction developing in a volume is thereby emphasized. As the reaction rate is decreasing, the domain affected by the reaction is increased chemical reactions generate a segregation process whereas the chemical species are mixed by the turbulent motion. In the premixed combustion engine a large range of operating points can be defined. The diagram usually used is that of Barrere Borghi. Several modeling methods should probably be developed according to the positions of the operating points in the diagram. Modeling methods are not presented herein. However the existence of typical structures in connection with the architecture of the combustion chamber could be examined in subsequent paper. The flame front can be subjected to distorting effects due to isolated rolling or to a sequence of vortices. Previously this last case has been touched upon. Using a spectral approach no discremination has to be made as for the sizes of these rollings: that could lead to new modeling methods if restricted shapes of vortices are accepted. By using a spectral method the behaviour of flames developing in a turbulent field can be easily understood. This approach could also open the way to new modeling methods. Indépendamment de l'intérêt scientifique que présentent les flammes de prémélange, les applications industrielles sont nombreuses. Dans cet article la flamme est supposée se développer dans un milieu turbulent ce qui oblige à une comparaison précise des échelles associées à la flamme et de celles associées à la turbulence. Cette confrontation n'est pas nouvelle. Cet article montre que ces comparaisons peuvent s'enrichir si l'on accepte une description spectrale de la turbulence faisant largement référence à de nombreux travaux antérieurs. La notion de champ extérieur turbulent agissant sur la flamme et celle de champ turbulent envahissant la flamme apparaît clairement lors d'une représentation spectrale. On voit également comment l'on passe continûment d'une situation avec front de réaction à une autre où la réaction s'effectue en volume tandis que se développe une compétition entre une réaction chimique qui tend à générer une ségrégation des espèces et une diffusion turbulente qui provoque le mélange. Dans les moteurs à combustion avec prémélange il existe un grand nombre de situations possibles dépendant de la vitesse de rotation. On conçoit aisément qu'une modélisation unique valable quel que soit le régime ne soit guère envisageable. Il y a là aussi matière à réflexions. Autre sujet au demeurant ouvert est celui de la modélisation. Dans le présent article certaines structures turbulentes sont étudiées. Resterait à savoir si une architecture spécifique de moteur n'est pas capable de générer des structures turbulentes particulières. Jusqu'ici le rôle de tourbillons privilégiés avait été introduit par la voie de séquences de tourbillons. L'approche spectrale permet de considérer un ensemble continu de structures analogues intervenant dans un phénomène de combustion. Si la voie n'est pas explorée du point de vue des modèles du moins peut-on entrevoir certaines possibilités. Le présent article en utilisant une approche spectrale permet une compréhension plus simple de certains phénomènes, il peut aussi suggérer d'autres voies de recherches dans le domaine de la modélisation.

With a view to provide some understanding of the effects of compressibility on turbulence, a two-point closure analysis is developed. The equations for the double correlations are derived for a weakly compressible fluid, and closed according to the Eddy Damped Quasi Normal Markovian theory (E.D.Q.N.M.). So far, only a few authors (Yoshizawa 1986, Hartke et al. 1988), have used two-point closures to study compressible turbulence. Recently, results of Direct Numerical Simulations have shown interesting behaviours (Passot & Pouquet 1986, Delorme 1985).

In order to investigate the behavior of a turbulence associated with a non- equilibrium spectrum, a grid equipped with propellers is used. Reynolds stress profiles clearly show the effects of the mean flow induced by the propellers: a spatially periodic distribution of anisotropy is obtained. The rotation of the propellers introduces a peak in the kinetic energy spectrum. The spectral disturbance, is found to be strongly anisotropic near the grid and to relax towards an axisym metric state. A large eddy simulation is carried out; the results show good qualitative agreement with the experimental spatial periodicity of inhomogeneities.

The evolution of slow, irreversible, isothermal, second-order chemical reactions between two nonpremixed species is experimentally studied in a field of quasiisotropic grid turbulence developing in a water tunnel. The experiments provide some basic data on the interaction between the turbulent mixing and the reaction, which tends to remove the mixed species.

Turbulence in fluid dynamics is one of the most challenging effects in connection with non-linear mechanisms. Starting from a deterministic equation for the instantaneous motion a statistical formalism has been introduced with the moments about the mean value of the velocity. These moments have been treated in a more or less sophisticated manner. The complete properties of the basic equations are considered with a numerical scheme that should not alter them. Even though the exact influence of such alterations is always open to debate, available results are interesting. Anyway, the turbulence problem can be considered in several different ways. The first group is controlled by deterministic equations. However, random initial conditions generate random solutions. The whole turbulent flow is characterized by adequate length and time scales. In fact, a spectral equilibrium is assumed so that turbulent scales are linked to each other.

The properties of the decaying turbulence downstream from a grid depend upon its design (diameter and spacing of the rods). Over a certain range the turbulent energy dissipation rate decays as the inverse square of the distance from the grid. The local rate of micromixing, which is due to the engulfing action of energy-dissipating vortices, then decreases linearly with increasing distance. It is first shown how micromixing in this inhomogeneous field can be calculated. Semi-batch reactor operation with a single, concentric feed tube, situated in grid turbulence, is considered. Modeling encompasses micromixing, self-engulfment and radial turbulent dispersion of the feed. The product distribution of four competitive-consecutive reactions is calculated by integrating three ordinary differential equations. Experimental results (30 runs), using various flow rates and flow rate ratios, feed locations, concentrations and feed pipes, are reported and compared with the model. The principal conclusion is that the feed tube was a source of additional turbulence, which also decays rapidly. A correlation, which includes the grid Reynolds number, the feed point location and the bore of the feed tube, accounting for this additional turbulence was established.