Large Eddy Simulation(LES)

Turbulent flow structures forming around isolated bridge abutments with curved toes are investigated in this study. Detached Eddy Simulation (DES) at a channel Reynolds number of 45,000 is used. Channel bottom is taken to be horizontal which corresponds to the initiation of the scouring process. Incoming flow in the simulations were fully turbulent containing unsteady velocity fluctuations. Changes in the structure of the horseshoe vortex system and the bed shear stress and pressure r.m.s. fluctuations on the bed is investigated for two different abutment lengths. These quantities are very important in sediment entrainment and can be used to identify the possible erosion regions on the bed. Although the pattern of the bed shear stress is similar in both cases there are some quantitative differences. This is mainly because of the changes in the strength of the main necklace vortex forming around the abutment and due to the different blockage ratio at the cross-section where the abutm...

This work investigates flow structures in Couette-Poiseuille (C-P) flows with the adverse pressure gradient (APG) adjusted to create zero mean skin friction on the stationary wall. The vortical structures are identified as regions with a positive value of the discriminant of the velocity gradient tensor (Chong et al., 1990; Soria et al., 1994). The data used in current study are from direct numerical simulation (DNS). The enstrophy density and the dissipation of kinetic energy carried by flow structures are instigated by examining the second invariants of the symmetric strain rate tensor S i j and the skew-symmetric rate of rotation tensor W i j (Chong et al. (1998)). Some characterizations of the topological flow structures in the near-wall region are similar to a viscous sub-layer in channel flows (Blackburn et al., 1996).

An explicit time filter is applied to the Navier-Stokes equation prior to a space filter. The time filter is supposed to be smooth, and an exact expansion depending on the time derivatives of the velocity is derived for the associated stress tensor. On the contrary, the effect of the space filter is treated as usual and an eddy viscosity model is introduced in the LES equation. The total stress is thus represented using a new class of mixed models combining time and space derivatives of the LES field.

Various alternative formulations of the LES equations have been explored in which additional evolution equations for variables such as the acceleration, the subgrid-scale stress tensor, or the subgrid-scale force are explicitly carried. Statistics of the velocity field obtained from the equation for the acceleration are shown to depend strongly on the initial conditions. This feature, which is independent of LES modeling issues, seems to prove that the velocity-acceleration formulation of the Navier-Stokes is not useful for numerical simulation. Equations for the subgrid-scale quantities appear to be much more stable. However, models required by this formulation of the LES problem still require additional study.

The Rover Environmental Monitoring Station (REMS) on the Mars Science Laboratory (MSL) offers the opportunity to explore the near surface atmospheric boundary layer (ABL) over an extended region of the Martian surface. The atmospheric boundary ...

A Large Eddy Simulation code based on a non-orthogonal, multiblock, finite volume approach with co-located variable storage, was ported to three different parallel architectures: a Cray T3E/1200E, an Alpha cluster and a PC Beowulf cluster. Scalability and parallelisation issues have been investigated, and merits as well as limitations of the three implementations are reported. Representative LES results for three flows are also presented.

In order to gain insight into the one-dimensional turbulence (ODT) model of Kerstein [1] as it pertains to residual stress closure in large-eddy simulation (LES), we develop ensemble mean closure (EMC), an algebraic stress closure based on the mappings and time scale physics employed in ODT. To allow analytic determination of the stress the ODT model is simplified, conceptually, such that eddy events act upon a velocity field linearized by the local resolved scale strain. EMC can account for viscous effects, addressing the laminar flow finite eddy viscosity problem without implementation of the dynamic procedure . The algebraic form of the model lends itself to analysis and we are able to derive a theoretical value for the eddy rate constant. This value is a bound on the rate constant for full ODT subgrid closure and yields good results in LES of decaying isotropic turbulence with EMC.

A new method to close the large eddy simulation (LES) equations using the one-dimensional turbulence (ODT) model of Kerstein et al. [1] is presented. In contrast to traditional eddy viscosity closures, LES/ODT explicitly models subgrid inertial cascading (a totally inviscid process), which is responsible for practically all of the energy transfer off the LES grid in high Reynolds number flows.

High delity n umerical simulation of wall-bounded turbulence requires physically sound representation of the small scale unsteady processes governing near-wall momentum, heat, and mass transfer. Conventional wall treatments do not capture the diverse multiphysics ow regimes relevant to engineering applications. To obtain a robust yet computationally a ordable near-wall submodel for turbulent o w computations, the ne-grained spatial structure and time evolution of the near-wall ow is simulated using a model formulated on a 1D domain corresponding to the wall-normal direction. This approach captures the strong variation of ow properties in the wall-normal direction and the transient i n teractions between this highly inhomogeneous region and the more nearly homogeneous (at ne scales) ow farther from the wall. The 1D simulation utilizes the One Dimensional Turbulence (ODT) methodology, whose formulation for the present application is described in detail. Demonstrations of ODT performance with regard to aspects of ow p h ysics relevant to near-wall ow modeling are presented. The coupling of ODT to a large eddy simulation (LES) of con ned turbulent o w i s d escribed, and the performance of the coupled formulation is demonstrated. It is concluded that this formulation has the potential to provide the delity needed for engineering applications at an a ordable computational cost.

In order to gain insight into the one-dimensional turbulence (ODT) model of Kerstein [1] as it pertains to residual stress closure in large-eddy simulation (LES), we develop ensemble mean closure (EMC), an algebraic stress closure based on the mappings and time scale physics employed in ODT. To allow analytic determination of the stress the ODT model is simplified, conceptually, such that eddy events act upon a velocity field linearized by the local resolved scale strain. EMC can account for viscous effects, addressing the laminar flow finite eddy viscosity problem without implementation of the dynamic procedure . The algebraic form of the model lends itself to analysis and we are able to derive a theoretical value for the eddy rate constant. This value is a bound on the rate constant for full ODT subgrid closure and yields good results in LES of decaying isotropic turbulence with EMC.

Large-eddy simulation (LES) was used to study the influence and the resulting flow mechanisms of active flow control applied to a two-dimensional vehicle geometry. The LES results were validated against existing Particle Image Velocimetry (PIV) and force measurement data. This was followed by an exploration of the influence of flow actuation on the near-wake flow and resulting aerodynamic forces. Not only was good agreement found with the previous experimental study, but new knowledge was gained in the form of a complex interaction of the actuation with the coherent flow structures. The resulting time-averaged flow shows a strong influence of the extension of the actuation slots and the lateral solid walls on the near-wake flow structures and thereby on the resulting drag.

for reading this study and providing many constructive, valuable comments and discussions. Their suggestions and recommendations on different aspects of my research helped me improve my dissertation. I am very grateful to my friends in Atlanta who made my life and study enjoyable. Many thanks are due to Ercan Cacan, Hatun Cacan and "tombis" Hazal and Hilal for their endless support, understanding and encouragement. Special thanks to "Busracan" (Onur Celik) and "Abla" (Dr. Gence Genc) for providing a pleasant and friendly atmosphere during their stay in Atlanta. v I would like to acknowledge the help of my lab mates, Sandeep Kumar

Energy consumption in the world is increasing rapidly due to population growth and technological developments. Renewable energy
sources have gained importance thanks to fossil fuel reserves being limited and their negative impacts on the environment. The design
of blade geometry is crucial to enhancing the efficiency of wind turbines that convert kinetic energy in the wind, one of the renewable
energy sources, into electrical energy. In this study, the effect of the change in the geometry of the lower surface of the NACA 63-415
airfoil used in wind turbines on aerodynamic performance was investigated numerically with the ANSYS Fluent CFD software. The
experimental lift coefficient (CL) valuesfor a range of angles of attack 0° to 20° at a Reynolds number of 1.6x106 number, were compared
with the numerical CL values obtained using Spalarat Almaras and SST k-ω turbulence models. Since CL values obtained by the SST
k-ω turbulence model fitted better with the experimental data, the calculations were performed using this model. In this study, NCAY10, NCAY20 and NCAY30 blade profiles were produced by reducing the dimensions of the lower surface of the NACA 63-415 airfoil by
10%, 20% and 30%, respectively. Numerical results showed that the new airfoils improved the aerodynamic performance compared to
the NACA 63-415 airfoil and the maximum CL/CD values were obtained using the NCAY30 airfoil. It is thought that this study will
contribute to the literature on enhancing the performance of the existing airfoils.

Sounding mechanism is numerically analyzed to elucidate physical processes in air-reed instruments. As an example, compressible large-eddy simulations (LES) are performed on both two and three dimensional ocarina. Since, among various acoustic instruments, ocarina is known as a combined system consisting of an edge-tone and a Helmholtz resonator, our analysis is mainly devoted to the resonant dynamics in the cavity.

Acoustic mechanics of air-reed instruments is investigated numerically with compressible Large-eddy simulation (LES). Taking a two dimensional air-reed instrument model, we have succeeded in reproducing sound oscillations excited in the resonator and have studied the characteristic feature of air-reed instruments, i.e., the relation of the sound frequency with the jet velocity described by the semi-empirical theory developed by Cremer & Ising, Coltman and other authors based on experimental results.

Natural hazards such as tsunamis, impulse waves and dam-break waves are rare, but extremely destructive. In recent times, more importance was given to structures that could withstand such events, however, uncertainties still exist in the estimation of wave velocities. This project focuses on the estimation of wave front celerity in a laboratory environment for both smooth and rough beds and the results are successfully compared to previous studies and design codes. Based on the experimental data an expression for the wave celerity is presented and discussed. The influence of bed roughness on the propagation velocity is also investigated and a dependence on the bed friction is observed, pointing out the need for further studies.

Since the first meeting in Lyon in 1986, the biannual European Turbulence Conferences have provided an informative survey of the international efforts in understanding turbulence in its fundamental and applied aspects. Now integrated into the conference cycles coordinated by the European Mechanics Society the meetings provide a regular forum for the exchange of ideas and the discussion of the latest developments. The more recent conferences in Barcelona (2000), Southampton (2002), Trondheim (2004), and Porto (2007) have attracted several hundred participants from 30 and more countries. The 12th meeting in this series, ETC 12, which was held in Marburg September 7-10, 2009, continues this tradition. Researchers from 34 countries submitted 336 abstracts to the conference. The number of submissions is somewhat lower than for the Porto or Barcelona meetings, but in line with the previous ones in Northern Europe. The contributions that were presented in Marburg were selected by the international advisory board for the European Turbulence Conferences. The committee was chaired by Professor Arne V. Johannson (Stockholm) and consisted of Professors Helge I. Andersson (Trondheim), Konrad Bajer (Warsaw), Luca Biferale (Rome), Claude Cambon (Lyon), Hans-Hermann Fernholz (Berlin), Peter Davidson (Cambridge, UK), Yuri Kachanov (Novosibirsk), Detlef Lohse (Twente), Jose L. Palma (Porto), Jean-Francois Pinton (Lyon) and the local organizer. for oral presentation, corresponding to a record number of almost 75% of the submissions. In addition, 70 papers were selected for presentation in a poster and seminar session. These numbers attest to the healthy state of the field and the reputation the conferences have achieved. in the sections given in the table of contents. Naturally, the level of activity in the subareas varies from conference to conference. For ETC12, the largest numbers of submissions were recorded for the areas of instability and transition, and wall bounded flows. These were closely followed by intermittency Impressed by the high quality of the abstracts the committee selected 250 As in previous years, the papers are grouped by the subfields as reflected v

The Rover Environmental Monitoring Station (REMS) on the Mars Science Laboratory (MSL) offers the opportunity to explore the near surface atmospheric boundary layer (ABL) over an extended region of the Martian surface. The atmospheric boundary ...

An ap rioristudy of the turbulence radiation interaction (TRI) is performed on numerical data from Direct Numerical Simulation (DNS) of a turbulent flame. The influence of the various correlations that appear in the radiative emission is investigated and their impact is evaluated in the context of Large Eddy Simulation (LES). In LES, only filtered quantities are computed, where the filter

The authors investigate the relationships between coherent structures and turbulence anisotropy in the neutral planetary boundary layer by means of empirical orthogonal function (EOF) analysis of large-eddy simulation (LES) data. The simulated flow contains near-surface transient streaks. The EOF analysis extracts the most energetic patterns from the velocity fluctuations based on their second-order spatial correlations. The scale and direction of streaks obtained from a level-by-level analysis of the LES flow field do correspond to that of the EOFs. It is found that two characteristics of the turbulence anisotropy depend on whether or not the velocity fluctuations with a given horizontal wave vector present distinct patterns: (i) the vertical extent up to which the turbulent kinetic energy (TKE) is concentrated and (ii) the ratio of the vertical TKE EV to the horizontal TKE EH. Although still present in the complete signal, this anisotropy is strongly emphasized when the signal is ...

In this study, the two-dimensional, incompressible numerical analysis is carried out to understand the turbulent fluid flow physics and thermal characteristics of a dual jet consisting of a wall jet and a parallel offset jet using standard k-𝝴 RNG turbulence model. The study is carried out to compare the flow and thermal features of the dual jet for an offset ratio of 3, 7 and 9 at different Reynolds numbers in the range of 5,000 to 20,000. The results of the flow field and turbulent behavior of a dual jet are illustrated in the form of temperature contours, velocity vectors, streamlines, and turbulent kinetic energy at different Reynolds number and offset ratio. The effect of offset ratio and Reynolds number on recirculation region and intensity of upper and lower vortices are discussed in detail. The velocity variation along the axial and vertical direction are also discussed.

Fluid flow of water in a model of a slab caster has been simulated using the large eddy simulation (LES) computational approach, and the simulated results are compared with experimental measurements performed using digital particle image velocimetry (DPIV) techniques. Simulation results agree acceptably well with the experimental measurements of instantaneous velocity fields. Flow patterns change with time as a consequence of the vertical oscillation of the jet core. These oscillations are originated by the residual Reynolds stresses that characterize turbulent flows. The asymmetry of fluid flows caused by these stresses provides biased flows. Thus, turbulence originates natural biasing effects without the influence of other operating factors such as the slide gate opening, gas bubbling, or inclusions clogging of the submerged entry nozzle (SEN). Instantaneous velocities follow periodical behaviors with time whose frequencies increase with increases of flow rate of liquid. Periodical flow changes originate velocity spikes, at some given casting speed, which are physically and mathematically identified. These sudden changes of fluid velocities are responsible of unsteady phenomena associated with fluid dynamics during steady operations of the mold.

A direct numerical simulation of the buoyancy driven turbulent flow inside a horizontal annular cavity at higher Rayleigh number, Ra = 1.18x10 9 , and the cylinders ratio of 4.87 has been carried out using the commercial code STAR-CCM+. Kinetic energy budgets have been calculated to verify the accuracy of the unstructured finite volume code on polyhedral cells in Direct Numerical Simulation (DNS) mode. Comparison of DNS results with wall resolved Unsteady RANS (URANS) models shows that the later models are able to capture the general flow features but fail to predict the large unsteadiness and high turbulence levels in the plume. However local heat transfer rates along the inner and outer cylinder walls are on average of acceptable accuracy for engineering purposes.

Large-Eddy Simulations (LES) with the first order Conditional Moment Closure (CMC) approach of a nitrogen-diluted hydrogen jet, igniting in a turbulent co-flowing hot air stream, are discussed. A detailed mechanism (9 species, 19 reactions) is used to represent the chemistry. Our study covers the following aspects: CFD mesh resolution; CMC mesh resolution; inlet boundary conditions and conditional scalar dissipation rate modelling. The Amplitude Mapping Closure for the conditional scalar dissipation rate produces acceptable results. We also compare different options to calculate conditional quantities in CMC resolution. The trends in the experimental observations are in general well reproduced. The auto-ignition length decreases with an increase in co-flow temperature and increases with increase in co-flow velocity. The phenomena are not purely chemically controlled: the turbulence and mixing play also affect the location of auto-ignition. In order to explore the effect of turbulence, two options were applied: random noise and turbulence generator based on digital filter. It was found that stronger turbulence promotes ignition.

We analyze the incompressible flow past a square cylinder immersed in the wake of an upstream splitter plate, which separates two streams of different velocities, UT (top) and UB (bottom). The Reynolds number associated with the flow below the plate is kept constant at ReB=DUB/ν=56, based on the square cylinder side D as characteristic length. The top-to-bottom flow dissymmetry is measured by the ratio R≡ReT/ReB∈[1,5.3] between the Reynolds numbers above and below the plate. The equivalent bulk Reynolds taken as the mean between top and bottom changes with R in the range Re≡(ReT+ReB)/2∈[56,178]. A Hopf bifurcation occurs at R=2.1±0.1 (Re=86.8±2.8), which results in an asymmetric Kármán vortex street with vortices only showing on the high-velocity side of the wake. A spanwise modulational instability is responsible for the three-dimensionalization of the flow at R≃3.1 (Re≃115) with the associated wavelength of λz≃2.4. For velocity ratios R≥4, the flow becomes spatiotemporally chaotic...

This study investigates how the inclusion of porous block influences the flow development in a serpentine passage, which consists of two straight square-sectioned ducts connected with a square-ended bend. Aluminium porous foam blocks have been attached to two opposite walls of the straight sections normal to the bend in a staggered manner, used as turbulence promoters. The investigation considers flows in the passage in stationary conditions, and in orthogonal rotation. Particle image velocimetry (PIV) is used to map out the flow development. The wall static pressure distribution along two sides of the passage (the leading and trailing under rotating conditions) is also provided. Tests are carried out at Reynolds numbers of 16,000, 26,000 and 36,000 and rotation numbers of 0.32 and 0.64. Preliminary tests, in a smooth (without porous blocks) passage of identical geometry, establish the reference conditions and enable the identification and quantification of the effects of the porous blocks on the flow development. PIV results show the meandering manner of the flow both upstream and downstream of the bend region, due to the presence of the porous blocks. Within the bend, in contrast to the case with a smooth upstream section, a single vortex dominates the flow. While rotation does not change the overall flow character, it does force more fluid through the blocks along the trailing (pressure) side of the duct. For both stationary and rotating conditions, the upstream section is long enough for the flow to become periodic over successive rib intervals, which produces very attractive data for CFD validation. These data include mean flow as well as second-moment profiles. The pressure coefficient is found to decrease linearly along the channel, due to the high blockage by the porous blocks. Rotation does not change the shape of the pressure coefficient profile but causes a greater reduction along the passage.

This study investigates how the inclusion of porous block influences the flow development in a serpentine passage, which consists of two straight square-sectioned ducts connected with a square-ended bend. Aluminium porous foam blocks have been attached to two opposite walls of the straight sections normal to the bend in a staggered manner, used as turbulence promoters. The investigation considers flows in the passage in stationary conditions, and in orthogonal rotation. Particle image velocimetry (PIV) is used to map out the flow development. The wall static pressure distribution along two sides of the passage (the leading and trailing under rotating conditions) is also provided. Tests are carried out at Reynolds numbers of 16,000, 26,000 and 36,000 and rotation numbers of 0.32 and 0.64. Preliminary tests, in a smooth (without porous blocks) passage of identical geometry, establish the reference conditions and enable the identification and quantification of the effects of the porous blocks on the flow development. PIV results show the meandering manner of the flow both upstream and downstream of the bend region, due to the presence of the porous blocks. Within the bend, in contrast to the case with a smooth upstream section, a single vortex dominates the flow. While rotation does not change the overall flow character, it does force more fluid through the blocks along the trailing (pressure) side of the duct. For both stationary and rotating conditions, the upstream section is long enough for the flow to become periodic over successive rib intervals, which produces very attractive data for CFD validation. These data include mean flow as well as second-moment profiles. The pressure coefficient is found to decrease linearly along the channel, due to the high blockage by the porous blocks. Rotation does not change the shape of the pressure coefficient profile but causes a greater reduction along the passage.

An experimental and computational study is presented on highly turbulent non-premixed counterflows under both isothermal and reactive conditions. Experimentally, Hot Wire Anemometry (HWA), twodimensional Particle Image Velocimetry (PIV) and OH Planar Laser Induced Fluorescence (PLIF) were applied. Computationally, Large-Eddy Simulations (LES) with a steady flamelet model were used to simulate the flow inside the nozzles and in the opposed flow region, using three different grid resolutions between 1.0 and 0.2 mm (0.5-70 million cells). The combined experimental and computational approach enabled the cross-validation of the simulation, and provided additional insight into the flow field. Both isothermal and burning conditions were examined with turbulent Reynolds numbers reaching a value of 900, demonstrating the system capability of reaching conditions of relevance to practical systems. Importantly, the simplicity of a compact, bench-top experiment is retained. The extension of the computational domain to a region within the nozzles with no optical access reveals the mechanism by which a specially designed turbulence generating plate (TGP) and burner housing yield turbulence intensities well exceeding 20%. The simulated and measured data were found to be in good agreement for first and second velocity moments, for the axial velocity autocorrelation function and for the normalised mean OH fluorescence. Similarity of OH-based flame morphology between experiments and computations also confirms that the LES successfully captures key features of the flow.

Numerical modelling of turbulence plays a crucial role in providing accurate microscale wind fields which are necessary to predict transport and dispersion of pollutants in the vicinity of buildings reliably. Although more elaborate turbulence models have been developed, two-equation turbulence models are widely applied in atmospheric flow and dispersion models. The large CPU time required for direct numerical simulation (DNS) and large eddy simulation (LES) still prohibits their application to microscale wind flow simulations of practical interest. In this study, the results obtained by the microscale model MIMO for the three-dimensional flow around a surface mounted cubical obstacle, using different two-equation turbulence models, are discussed. The accuracy of the investigated turbulence models is assessed by comparing the numerical results with wind tunnel measurements.

Numerical modelling of turbulence plays a crucial role in providing accurate wind fields, which are necessary to predict transport and dispersion of pollutants in the vicinity of buildings reliably. Although the standard k-e has some well-known deficiencies and more elaborate turbulence models have been developed it is still by far the most widely applied in atmospheric flow and dispersion models. The large CPU time requirement of direct numerical simulation (DNS) and large eddy simulation (LES) still prohibits their application to wind engineering problems of practical interest. Many linear two-equation turbulence models have been proposed in recent years to overcome the deficiencies of the standard k-e turbulence model. However, comparison shows that none achieves much greater accuracy in all regions of typical wind engineering flow fields. In this paper, a new nonlinear turbulence model is proposed which leads to improved results compared to a conventional eddy-viscosity scheme. The flows considered for the validation of the new model range from simple shear and boundary layer flows, to flows around arrays of cubic obstacles. The accuracy of the proposed turbulence model is assessed by comparing the numerical results with wind-tunnel measurements.

A Spectral Difference (SD) algorithm on tensor-product elements which solves the reacting compressible Navier-Stokes equations is presented. The classical SD algorithm is shown to be unstable when a multi-species gas where thermodynamic properties depend on temperature and species mass fractions is considered. It is demonstrated that computing primitive variables from conservative variables at solution points and extrapolating them at flux points, rather than conservative variables, make the SD method stable for such case. In addition, a new methodology, more stable, for computing the diffusive fluxes at an interface is detailed. To allow the simulation of combustion, characteristic and wall boundary conditions for multi-species flows in the SD framework are introduced and the thickened flame turbulent combustion model is adapted to the SD context. Validation test cases from one-dimensional and two-dimensional Flow, Turbulence and Combustion laminar flames to a three-dimensional turbulent flame have been performed showing excellent agreement with each of the reference solutions. To the authors' knowledge, this is the first time that a 3D turbulent flame is simulated using the SD method. The interest of employing a high-order method, such as the SD, in combustion is demonstrated: high values of the polynomial degree improve the quality of the results which advocates for the use of p-refinement when simulating reacting flows.

The current work presents the Large Eddy Simulation (LES) of a kerosene spray ignition phase in a simplified aeronautical combustor for which detailed experimental data are available. The carrier phase is simulated using an unstructured multi-species compressible Navier-Stokes solver while the dispersed liquid phase is modeled with a Lagrangian approach. An energy deposition model neglecting the presence of a plasma phase in the very first instants of the energy deposition process, a reduced kinetic scheme and a simplified spray injection model are combined to achieve both a reasonable computational expense and a satisfactory overall accuracy. Following a brief description of the validation of these models, non reactive gaseous and twophase flow LES's of the target combustor are performed. Excellent agreement with experiments is observed for the non reactive gaseous simulations. The dispersed phase velocity fields are also well reproduced while discrepancies appear for the spatial size distribution of the particles. Finally, numerical snapshots of a successful ignition phase are shown and discussed.

A Spectral Difference (SD) algorithm on tensor-product elements which solves the reacting compressible Navier-Stokes equations (NSE) is presented. The classical SD algorithm is shown to be unstable when a multispecies gas where thermodynamic properties depend on temperature and species mass fractions is considered. In that case, a modification of the classical algorithm was successfully employed making it stable. It uses the fact that it is better for the multispecies case to compute primitive variables from conservative variables at solution points and then extrapolate them at flux points rather than extrapolating conservative variables at flux points and reconstruct primitive variables on these points. Characteristic, wall and symmetry boundary conditions for reactive flows in the SD framework are also introduced. They all use the polynomial form of the variables and of the fluxes to impose the correct boundary condition at a boundary flux point. Validation test cases on one-dimensional and two-dimensional laminar flames have been performed using both global chemistry and Analytically Reduced Chemistry (ARC). Results show excellent agreement with the reference combustion code AVBP validating the implementation of this SD method on laminar combustion.

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

The design challenge of reliable lean combustors needed to decrease pollutant emissions has clearly progressed with the common use of experiments as well as Large Eddy Simulation (LES) because of its ability to predict the interactions between turbulent flows, sprays, acoustics and flames. However, the accuracy of such numerical predictions depends very often on the user's experience to choose the most appropriate flow modeling and, more importantly, the proper spatial discretization for a given computational domain. The present work focuses on the last issue and proposes a static mesh refinement strategy based on flow physical quantities. To do so, a combination of sensors based on the dissipation and production of kinetic energy coupled to the flame-position probability is proposed to detect the regions of interest where flow physics happens and grid adaptation is recommended for good LES predictions. Thanks to such measures a local mesh resolution can be achieved in these zones improving the LES overall accuracy while, eventually, coarsening everywhere else in the domain to reduce the computational cost. The proposed mesh refinement strategy is detailed and validated on two reacting-flow problems: a fully premixed bluff-body stabilized flame, i.e. the VOLVO test case, and a partially premixed swirled flame, i.e. the PRECCINSTA burner, which is closer to industrial configurations. For both cases, comparisons of the results with experimental data underline the fact that the predictions of the flame stabilization, and hence the computed velocity and temperature fields, are strongly influenced by the mesh quality and significant improvement can be obtained by applying the proposed strategy.

A Spectral Difference (SD) algorithm on tensor-product elements which solves the reacting compressible Navier-Stokes equations is presented. The classical SD algorithm is shown to be unstable when a multi-species gas where thermodynamic properties depend on temperature and species mass fractions is considered. It is demonstrated that computing primitive variables from conservative variables at solution points and extrapolating them at flux points, rather than conservative variables, make the SD method stable for such case. In addition, a new methodology, more stable, for computing the diffusive fluxes at an interface is detailed. To allow the simulation of combustion, characteristic and wall boundary conditions for multi-species flows in the SD framework are introduced and the thickened flame turbulent combustion model is adapted to the SD context. Validation test cases from one-dimensional and two-dimensional Flow, Turbulence and Combustion laminar flames to a three-dimensional turbulent flame have been performed showing excellent agreement with each of the reference solutions. To the authors' knowledge, this is the first time that a 3D turbulent flame is simulated using the SD method. The interest of employing a high-order method, such as the SD, in combustion is demonstrated: high values of the polynomial degree improve the quality of the results which advocates for the use of p-refinement when simulating reacting flows.

A Spectral Difference (SD) algorithm on tensor-product elements which solves the reacting compressible Navier-Stokes equations (NSE) is presented. The classical SD algorithm is shown to be unstable when a multispecies gas where thermodynamic properties depend on temperature and species mass fractions is considered. In that case, a modification of the classical algorithm was successfully employed making it stable. It uses the fact that it is better for the multispecies case to compute primitive variables from conservative variables at solution points and then extrapolate them at flux points rather than extrapolating conservative variables at flux points and reconstruct primitive variables on these points. Characteristic, wall and symmetry boundary conditions for reactive flows in the SD framework are also introduced. They all use the polynomial form of the variables and of the fluxes to impose the correct boundary condition at a boundary flux point. Validation test cases on one-dimensional and two-dimensional laminar flames have been performed using both global chemistry and Analytically Reduced Chemistry (ARC). Results show excellent agreement with the reference combustion code AVBP validating the implementation of this SD method on laminar combustion.

An ap rioristudy of the turbulence radiation interaction (TRI) is performed on numerical data from Direct Numerical Simulation (DNS) of a turbulent flame. The influence of the various correlations that appear in the radiative emission is investigated and their impact is evaluated in the context of Large Eddy Simulation (LES). In LES, only filtered quantities are computed, where the filter

Simulation of turbulent combustion has gained high potential with the Large Eddy Simulation (LES) approach, allowing to predict unsteady turbulent reactive flows. In this approach only the largest scales of the turbulence are solved while the smallest scales are modelled. This approach permits to simulate complex industrial geometries on a wide range of Reynolds numbers. Previous works have shown the ability of LES to predict unsteady combustion behaviors such as : instabilites, ignitions and extinctions in industrial systems .

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

In recent years, the Engine Combustion Network (ECN) has developed as a worldwide reference for understanding and describing engine combustion processes, successfully bringing together experimental and numerical efforts. Since experiments and numerical simulations both target the same boundary conditions, an accurate characterization of the stratified environment that is inevitably present in experimental facilities is required. The difference between the core-, and pressure-derived bulk-temperature of pre-burn combustion vessels has been addressed in various previous publications. Additionally, thermocouple measurements have provided initial data on the boundary layer close to the injector nozzle, showing a transition to reduced ambient temperatures. The conditions at the start of fuel injection influence physicochemical properties of a fuel spray, including near nozzle mixing, heat release computations, and combustion parameters. To address the temperature stratification in more detail, thermocouple measurements at larger distances from the spray axis have been conducted. Both the temperature field prior to the pre-combustion event that preconditions the high-temperature, high-pressure ambient, as well as the stratification at the moment of fuel injection were studied. To reveal the cold boundary layer near the injector with a better spatial resolution, Rayleigh scattering experiments and thermocouple measurements at various distances close to the nozzle have been carried out. The impact of the boundary layers and temperature stratification are illustrated and quantified using numerical simulations at Spray A conditions. Next to a reference simulation with a uniform temperature field, six different stratified temperature distributions have been generated. These distributions were based on the mean experimental temperature superimposed by a randomized variance, again derived from the experiments. The results showed that an asymmetric flame structure arises in the computed results when the temperature stratification input is used. In these predictions, first-stage ignition is advanced by 24 μs, while second-stage ignition is delayed by 11 μs. At the same time a lift-off length difference between the top and the bottom of up to 1.1 mm is observed. Furthermore, the lift-off length is less stable over time. Given the shown dependency, the temperature data is made available along with the vessel geometry data as a recommended basis for future numerical simulations.

In the literature, anisotropy-invariant maps are being proposed to represent a domain within which all realizable Reynolds stress invariants must lie. It is shown that the representation proposed by Lumley and Newman has disadvantages owing to the fact that the anisotropy invariants (II, III) are nonlinear functions of stresses. In the current work, it is proposed to use an equivalent linear representation of the anisotropy invariants in terms of eigenvalues. A novel barycentric map, based on the convex combination of scalar metrics dependent on eigenvalues, is proposed to provide a non-distorted visual representation of anisotropy in turbulent quantities. This barycentric map provides the possibility of viewing the Reynolds stress and any anisotropic stress tensor. Additionally the barycentric map provides the possibility of quantifying the weighting for any point inside it in terms of the limiting states (one component, two component, three component). The mathematical basis for the barycentric map is derived using the spectral decomposition theorem for second-order tensors. In this way, an analytical proof is provided that All turbulence lies along the boundaries or inside the barycentric map. It is proved that the barycentric map and the anisotropy-invariant maps are one to one uniquely interdependent and as a result satisfy the requirement of realizability.

This thesis develops and tests various transient and steady-state computational models such as direct numerical simulation (DNS), large eddy simulation (LES), filtered unsteady Reynolds-averaged Navier-Stokes (URANS) and steady Reynolds-averaged Navier-Stokes (RANS) with and without magnetic field to investigate turbulent flows in canonical as well as in the nozzle and mold geometries of the continuous casting process. The direct numerical simulations are first performed in channel, square and 2:1 aspect rectangular ducts to investigate the effect of magnetic field on turbulent flows. The rectangular duct is a more practical geometry for continuous casting nozzle and mold and has the option of applying magnetic field either perpendicular to broader side or shorter side. This work forms the part of a graphic processing unit (GPU) based CFD code (CU-FLOW) development for magnetohydrodynamic (MHD) turbulent flows. The DNS results revealed interesting effects of the magnetic field and its orientation on primary, secondary flows (instantaneous and mean), Reynolds stresses, turbulent kinetic energy (TKE) budgets, momentum budgets and frictional losses, besides providing DNS database for two-wall bounded square and rectangular duct MHD turbulent flows. Further, the low-and high-Reynolds number RANS models (k-ε and Reynolds stress models) are developed and tested with DNS databases for channel and square duct flows with and without magnetic field. The MHD sink terms in k-and ε-equations are implemented as proposed by Kenjereš and Hanjalić using a user defined function (UDF) in FLUENT. This work revealed varying accuracies of different RANS models at different levels. This work is useful for industry to understand the accuracies of these models, including continuous casting. iii After realizing the accuracy and computational cost of RANS models, the steady-state k-ε model is then combined with the particle image velocimetry (PIV) and impeller probe velocity measurements in a 1/3 rd scale water model to study the flow quality coming out of the well-and mountain-bottom nozzles and the effect of stopper-rod misalignment on fluid flow. The mountain-bottom nozzle was found more prone to the longtime asymmetries and higher surface velocities. The left misalignment of stopper gave higher surface velocity on the right leading to significantly large number of vortices forming behind the nozzle on the left. Later, the transient and steady-state models such as LES, filtered URANS and steady RANS models are combined with ultrasonic Doppler velocimetry (UDV) measurements in a GaInSn model of typical continuous casting process. LES-CU-LOW is the fastest and the most accurate model owing to much finer mesh and a smaller timestep. This work provided a good understanding on the performance of these models. The behavior of instantaneous flows, Reynolds stresses and proper orthogonal decomposition (POD) analysis quantified the nozzle bottom swirl and its importance on the turbulent flow in the mold. Afterwards, the aforementioned work in GaInSn model is extended with electromagnetic braking (EMBr) to help optimize a ruler-type brake and its location for the continuous casting process. The magnetic field suppressed turbulence and promoted vortical structures with their axis aligned with the magnetic field suggesting tendency towards 2-d turbulence. The stronger magnetic field at the nozzle well and around the jet region created large scale and lower frequency flow behavior by suppressing nozzle bottom swirl and its front-back alternation.

This work studies the momentum and energy transport mechanisms in the corner between a free surface and a solid wall. We perform large-eddy simulations of the incompressible fully developed turbulent flow in a square duct bounded above by a free-slip wall, for Reynolds numbers based on the mean friction velocity and the duct width equal to 360, 600 and 1000. The flow in the corner is strongly affected by the advection due to two counter-rotating secondary-flow regions present immediately below the free surface. Because of the convection of the inner eddy, as the free surface is approached, the friction velocity on the sidewall first decreases, then increases again. A similar behaviour is observed for the surface-parallel Reynolds-stress components, which first decrease and then increase again very close to the surface. The budgets of the Reynolds stresses show a strong reduction of all terms of the dissipation tensor in both the inner and outer near-corner regions. They exhibit a reduction in both production and dissipation towards the free surface. Very close to the solid boundary, within 15-20 viscous lengths of the sidewall, the turbulent kinetic energy production and the surface-parallel fluctuations rebound in the thin layer adjacent to the free surface. The Reynolds-stress anisotropy appears to be the main factor in the generation of the mean secondary flow. The multi-layer structure of the boundary layer near the free surface is also discussed.

Our long-term goal is to understand the dynamic and thermodynamic processes causing changes in the velocity and density structure of the upper Arctic Ocean. For example we seek to understand the heat and mass balance of the mixed layer. In light of recent changes in the upper ocean structure, our longterm goals are shifting toward processes important to larger-scale changes.