No description available.

No description available.

In this study we highlight the influence of mean dilatation and mean density gradient on the Reynolds stress modeling of compressible, heat-releasing and supercritical turbulent flows.Firstly, the modeling of the rapid pressure-strain correlation has been extended to self-consistently account for the influence of mean dilatation.Secondly, an algebraic model for the turbulent density flux has been developed and coupled to the tensor equationfor Reynolds stress anisotropy via a 'local mean acceleration',a generalization of the buoyancy force. We applied the resulting differential Reynolds stress model (DRSM) and the corresponding explicit algebraic Reynolds stress model (EARSM) to homogeneously sheared and compressed or expanded two-dimensional mean flows. Both formulations have shown that our model preserves the realizability of the turbulence, meaning that the Reynolds stresses do not attain unphysical values, unlike earlier approaches. Comparison with rapid distortion theory (RDT) demonstrated that the DRSM captures the essentials of the transient behaviour of the diagonal anisotropies and gives good predictions of the turbulence kinetic energy. A general three-dimensional solution to the coupled EARSM  has been formulated. In the case of turbulent flow in de Laval nozzle we investigated the influence of compressibility effects and demonstrated that the different calibrations lead to different turbulence regimes but with retained realizability. We calibrated our EARSM against a DNS of combustion in a wall-jet flow. Correct predictions of turbulent density fluxes have been achieved and essential features of the anisotropy behaviour have been captured.The proposed calibration keeps the model free of singularities for the cases studied. In addition,  we have applied the EARSM to the investigation of supercritical carbon dioxide flow in an annulus. The model correctly captured mean enthalpy, temperature and density as well as the turbulence shear stress. Hence, we consider the model as a useful tool for the analysis of a wide range of compressible flows with large density variation. / <p>QC 20160314</p>

Thermal cracking is an industrial process sensitive to both temperature and pressure operating conditions. The use of internally ribbed reactors is a passive method to enhance the chemical selectivity of the process, thanks to a significant increase of heat transfer. However, this method also induces an increase in pressure loss, which is damageable to the chemical yield and must be quantified. Because of the complexity of turbulence and chemical kinetics, and as detailed experimental measurements are difficult to conduct, the real advantage of such geometries in terms of selectivity is however poorly known and difficult to assess. This work aims both at evaluating the real benefits of internally ribbed reactors in terms of chemical yields and at proposing innovative and optimized reactor designs. This is made possible using the Large Eddy Simulation (LES) approach, which allows to study in detail the reactive flow inside several reactor geometries. The AVBP code, which solves the Navier-Stokes compressible equations for turbulent flows, is used in order to simulate thermal cracking thanks to a dedicated numerical methodology. In particular, the effect of pressure loss and heat transfer on chemical conversion is compared for both a smooth and a ribbed reactor in order to conclude about the impact of wall roughness in industrial operating conditions. An optimization methodology, based on series of LES and Gaussian process, is finally developed and an innovative reactor design for thermal cracking applications, which maximizes the chemical yield, is proposed

Particle-laden flows involve in many energy and industrial processes within a wide scale range. Solid fuel combustion and gasication, drying and catalytic cracking are some of the examples. It is vital to have a better understanding of the phenomena inside the reactors involving in particle-laden flows for process improvements and design. Computational fluid dynamics (CFD) can be a robust tool for these studies with its advantage over experimental methods. The large variation of length scales (101- 10-9 m) and time scales (days-microseconds) is a barrier to execute detailed simulations for large scale reactors. Current state-of-the-art is to use models to bridge the gap between small scales and large scales. Therefore, the accuracy of the models is key to better predictions in large scale simulations.    Particle-laden flows have complexities due to many reasons. One of the main challenge is to describe how the particle-fluid interaction varies when the particles are reacting. Particle and the fluid interact through mass, momentum and heat exchange. Mass, momentum and heat exchange is presented by the Sherwood number (Sh), drag coefficient (CD) and Nusselt number (Nu) in fluid dynamics. Currently available models do not take into account for the effects of net gas flow generated by heterogeneous chemical reactions. Therefore, the aim of this research is to propose new models for CD and Nu based on the flow and temperature fields estimated by particle-resolved direct numerical simulations (PR-DNS). Models have been developed based on physical interpretation with only one fitting parameter, which is related to the relationship between Reynolds number and the boundary layer thickness. The developed models were compared with the simulation results solving intra-particle flow under char gasification. The drawbacks of models were identied and improvements were proposed.    The models developed in this work can be used for the better prediction of flow dynamics in large scale simulations in contrast to the classical models which do not consider the effect of heterogeneous reactions. Better predictions will assist the design of industrial processes involving reactive particle-laden flows and make them highly effcient and low energy-intensive.

In the present work, two phase reacting flow in a single cavity Trapped Vortex Combustor (TVC) is studied at atmospheric conditions. KIVA-3V, numerical program for simulating three dimensional compressible reacting flows with sprays using Lagrangian-Drop Eulerian-fluid procedure is used. The stochastic discrete droplet model is used for simulating the liquid spray. In each computational cell, it is assumed that the volume occupied by the liquid phase is very small. But this assumption of very low liquid volume fraction in a computational cell is violated in the region close to the injection nozzle. This introduces grid dependence in predictions of liquid phase in the region close to the nozzle in droplet collision algorithm, and in momentum coupling between the liquid and the gas phase. Improvements are identified to reduce grid dependence of these algorithms and corresponding changes are made in the standard KIVA-3V models. Pressure swirl injector which produces hollow cone spray is used in the current study along with kerosene as the liquid fuel. Modifications needed for modelling pressure swirl atomiser are implemented. The Taylor Analogy Breakup (TAB) model, the standard model for predicting secondary breakup is improved with modifications required for low pressure injectors. The pressure swirl injector model along with the improvements is validated using experimental data for kerosene spray from the literature. Simulations of two phase reacting flow in a single cavity TVC are performed and the temperature distribution within the combustor is studied. In order to identify an optimum configuration with liquid fuel combustion, the following parameters related to fuel and air such as cavity fuel injection location, cavity air injection location, Sauter Mean Diameter (SMD) of injected fuel droplets, velocity of the fuel injected are studied in detail in order to understand the effect of these parameters on combustion characteristics of a single cavity TVC.

Le procédé de craquage thermique est un procédé industriel sensible aux conditions de température et de pression. L’utilisation de réacteurs aux parois nervurées est une méthode permettant d’améliorer la sélectivité chimique du procédé en augmentant considérablement les transferts de chaleur. Cependant, cette méthode induit une augmentation des pertes de charge dans le réacteur, ce qui est dommageable pour le rendement chimique et doit être quantifié. En raison de la complexité de l’écoulement turbulent et de la cinétique chimique, le gain réel offert par ces géométries en termes de sélectivité chimique est toutefois mal connu et difficile à estimer, d’autant plus que des mesures expérimentales détaillées sont très rares et difficiles à mener. L’objectif de ce travail est double: d’une part évaluer le gain réel des parois nervurées sur le rendement chimique; d’autre part proposer de nouveaux designs de réacteurs offrant une sélectivité chimique optimale. Ceci est rendu possible par l’approche de simulation numérique aux grandes échelles (LES), qui est utilisée pour étudier l’écoulement réactif à l’intérieur de diverses géométries de réacteurs. Le code AVBP, qui résout les équations de Navier Stokes compressibles pour les écoulements turbulents, est utilisé pour simuler le procédé grâce à une méthodologie numérique adaptée. En particulier, les effets des pertes de charge et du transfert thermique sur la conversion chimique sont comparés pour un réacteur lisse et un réacteur nervuré afin de quantifier l’impact de la rugosité de paroi dans des conditions d’utilisation industrielles. Une méthodologie d’optimisation du design des réacteurs, basée sur plusieurs simulations numériques et les processus Gaussiens, est finalement mise au point et utilisée pour aboutir à un design de réacteur de craquage thermique innovant, maximisant le rendement chimique / Thermal cracking is an industrial process sensitive to both temperature and pressure operating conditions. The use of internally ribbed reactors is a passive method to enhance the chemical selectivity of the process, thanks to a significant increase of heat transfer. However, this method also induces an increase in pressure loss, which is damageable to the chemical yield and must be quantified. Because of the complexity of turbulence and chemical kinetics, and as detailed experimental measurements are difficult to conduct, the real advantage of such geometries in terms of selectivity is however poorly known and difficult to assess. This work aims both at evaluating the real benefits of internally ribbed reactors in terms of chemical yields and at proposing innovative and optimized reactor designs. This is made possible using the Large Eddy Simulation (LES) approach, which allows to study in detail the reactive flow inside several reactor geometries. The AVBP code, which solves the Navier-Stokes compressible equations for turbulent flows, is used in order to simulate thermal cracking thanks to a dedicated numerical methodology. In particular, the effect of pressure loss and heat transfer on chemical conversion is compared for both a smooth and a ribbed reactor in order to conclude about the impact of wall roughness in industrial operating conditions. An optimization methodology, based on series of LES and Gaussian process, is finally developed and an innovative reactor design for thermal cracking applications, which maximizes the chemical yield, is proposed

Pour améliorer l'efficacité des procédés thermiques de craquages et réduire les phénomènes de cokage liés à la température de paroi trop élevée, l'utilisation de tubes nervurés est une technique potentiellement car elle permet d'améliorer le mélange et d'augmenter les transferts de chaleur. Cependant, la perte de charge est significativement augmentée. En raison de la complexité de l'écoulement turbulent, du système chimique et du couplage turbulencechimie, il est difficile d'estimer a priori la perte réelle en termes de sélectivité des tubes nervurés. Les expériences représentatives de laboratoire combinant turbulence, transferts de chaleur et chimie sont très rares et trop coûteuses à l'échelle industrielle. Dans ce travail, l'approche simulation aux grandes échelles résolue à la paroi (WRLES) est utilisée pour étudier écoulement non-réactif puis réactif dans des tubes à la fois lisses et nervurés, pour quantifier leur impact sur la turbulence et sur la chimie. Le code AVBP, qui résout les équations de Navier-Stokes compressibles pour les écoulements turbulents, est utilisé avec des schémas chimique réduites du craquage de l'éthane puis du butane. L'écoulement à la paroi est analysé en détail et comparé pour les deux géométries, fournissant des informations utiles pour le développement ultérieur de modèles de parois pour ce type de rugosité. L'impact de la résolution du maillage et du schéma numérique est également discuté, pour trouver le meilleur compromis entre coût et précision de calcul pour une application industrielle. L'impact des structures d'écoulement turbulent ainsi que leurs effets sur le transfert thermique et le mélange sur les réactions chimique sont étudiés à la fois pour les tubes lisses et les tubes nervurés. Perte de pression, transfert de chaleur et conversion chimique sont finalement comparés. / To improve the efficiency of thermal-cracking processes, and to reduce the coking phenomena due to high wall temperature, the use of ribbed tubes is an interesting technique as it allows better mixing and heat transfer. However it also induces significant increase in pressure loss. The complexity of the turbulent flow, the chemical system, and the chemistry-turbulence interaction makes it difficult to estimate a priori the real loss of ribbed tubes in terms of selectivity. Experiments combining turbulence, heat transfer and chemistry are very rare in laboratories and too costly at the industrial scale. In this work, Wall-Resolved Large Eddy Simulation (WRLES) is used to study non-reacting and reacting flows in both smooth and ribbed tubes, to show the impact of the ribs on turbulence and chemistry. Simulations were performed with the code AVBP, which solves the compressible Navier-Stokes equations for turbulent flows, using reduced chemistry scheme of ethane and butane cracking for reacting cases. Special effort was devoted to the wall flow, which is analyzed in detail and compared for both geometries, providing useful information for further development of roughness-type wall models. The impact of grid resolution and numerical scheme is also discussed, to find the best trade-off between computational cost and accuracy for industrial application. Results investigate and analyze the turbulent flow structures, as well as the effect of heat transfer efficiency and mixing on the chemical process in both smooth and ribbed tubes. Pressure loss, heat transfer and chemical conversion are finally compared.