CTR Seminars Archive 2011
Date: Monday Nov 28, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Richard J.A.M. Stevens, Department of Science and Technology, Twente University, P.O Box 217, 7500 AE Enschede, The Netherlands
Title: Rayleigh-B´enard turbulence
Abstract: Turbulence is seen as one of the last outstanding unsolved problems in classical physics. In the last century, great minds as Heisenberg, von Weizs¨acker, Kolmogorov, Prandtl, and G.I. Taylor had worked on it, and Einstein put his last postdoc Bob Kraichnan on the subject of turbulence – a task which Kraichnan never finished. The rapid development of experimental and numerical techniques in this area and the growth of computing power creates a lot of activity on turbulence research. In turbulence problems encountered in the real world the influence of walls is very important and one of the classical systems to study concepts in fluid dynamics is the Rayleigh-B´enard (RB) system, i.e. the buoyancy driven flow of a fluid heated from below and cooled from above, see figure 1a and figure 1b. Also from an applied viewpoint, thermally driven flows are of utmost importance. Examples are thermal convection in the atmosphere, in the ocean, or in process technology. Rotating turbulent flow is of utmost importance to optimize industrial applications such as the efficient separation of carbon dioxide (CO2) from nitrogen in the emission gases of conventional carbon-based power plants to enable long term CO2 storage or the separation of CO2 from natural methane gas. In both cases the method of choice is pressurization and cooling down of the gas mixture so that finally CO2 condensates into droplets and can be separated in so-called rotational phase separators. Due to the droplet condensation, considerable heat transfers emerge in this process which are strongly affected by rotation. In this presentation experimental I will present experimental, numerical and theoretical results on RB convection. Simulations and experiments on RB convection are complementary. In accurate experimental measurements of the heat transfer a completely isolated system is needed. Therefore, one cannot visualize the flow while the heat transfer is measured. On the positive side, in experiments one can obtain very high Ra numbers and long time averaging. In direct numerical simulations (DNS), on the other hand, one can simultaneously measure the heat transfer while the complete flow field is available for analysis. Unfortunately, up to recently, there was a major disagreement between experimental and numerical measurements of the heat transfer. We showed that this disagreement was due to insufficient numerical resolution in the simulations. Our new high resolution simulations agree excellently with the experimental result, see figure 1c. In addition, we showed some very unexpected transitions between different turbulent states in RB convection. We find that very small changes in the control parameters can completely change the flow structure or can drastically alter the flow dynamics that are observed.
Date: Friday Nov 4, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: M. Massot, Laboratoire EM2C, UPR CNRS 288 – Ecole Centrale Paris – marc.massot@ecp.fr and Center for Turbulence Research – Stanford University –mmassot@stanford.edu
Title: Adaptive time-space algorithms for the simulation of multi-scale reaction waves
Abstract: Numerical simulations of multi-scale phenomena are commonly used for modeling purposes in many applications such as combustion, chemical vapor deposition, or air pollution modeling. In general, all these models raise several diffculties created by the high number of unknowns, the wide range of temporal scales due to large and detailed chemical kinetic mechanisms, as well as steep spatial gradients associated with very localized fronts of high chemical activity. Furthermore, a natural stumbling block to perform 3D simulations with all scales resolution is either the unreasonably small time step due to stability requirements or the unreasonable memory requirements for implicit methods. In this work, we introduce a new resolution strategy for multi-scale reaction waves based mainly on time operator splitting and space adaptive multiresolution, in the context of very localized and stiff reaction fronts. It considers high order time integration methods for reaction, diffusion and convection problems, in order to build a time operator splitting scheme that exploits efficiently the special features of each problem. Based on theoretical studies of numerical analysis, such a strategy leads to a splitting time step which is not restricted neither by fast scales in the source term nor by restrictive stability limits of diffusive or convective steps, but only by the physics of the phenomenon. Moreover, this splitting time step is dynamically adapted taking into account a posteriori error estimates, carefully computed by a second embedded and economic splitting method. The main goal is then to perform computationally very efficient as well as accurate in time and space simulations of the complete dynamics of multi-scale phenomena under study, considering large simulation domains with conventional computing resources and splitting time steps purely dictated by the physics of the phenomenon and not by any stability constraints associated with mesh size or source time scales. The main results of this Tea Seminar are originally the Ph.D. work of M. Duarte, who will graduate in France in December 2011.
Date: Friday Oct 28, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Damien KAH, Center for Turbulence Research, Stanford University, office 501K
Title: Taking into account polydispersity for the modeling of liquid fuel injection in internal combustion engines
Abstract: Multiphase Flow modeling is a critical aspect involved in the study of fuel injection in internal combustion engines, or more generally any industrial device involving a multiphase flow made of liquid fuel injected in a chamber filled with gaz. In and of itself, it is possible to simulate this flow in the context of a direct numerical simulation. However the small structures created during injection (droplets of diameter until 10 ?m or less) lead to a prohibitive computational cost for any industrial application. Therefore modeling is necessary. In this context, two areas are formally distinguished: the dense liquid core close to the injector called separate-phase flow, and the spray made of a polydisperse droplet population (i.e. droplets with different sizes) generated after the atomization processes downstream of the injector. This talk investigates Eulerian models for the description of polydisperse evaporating sprays. They represent a potential alternative to Lagrangian models, widely used at present, yet suffering from major drawbacks. Thus, the Multi-Fluid model ([1] and references herein) has been assessed. Although it is very promising, two difficulties are highlighted: its cost for a precise description of polydispersity, and its inability to describe particle trajectory crossing (PTC). Solutions to these two limitations are studied in [3], relying both on high order moment methods. This talk addresses the first aspect with the presentation of the Eulerian Multi Size Moment (EMSM) model, offering a more efficient resolution of polydisperse evaporating sprays than the Multi-Fluid model does. Mathematical tools are used to close the model and are combined with original finite volume kinetic-based schemes in order to preserve the moment-set integrity, for evaporation and advection. In order to assess its potential, the EMSM model and numerical tools are implemented in the MUSES3D code, an academic DNS solver that provides a framework devoted to spray method evaluation [1]. The extension of the EMSM model to an industrial context is then considered, with its implementation in the IFP-C3D code, a 3D unstructured reactive flow solver with spray. In order to perform computations within a moving domain (due to the piston movement) the Arbitrary Lagrangian Eulerian (ALE) [2] formalism is used. The robustness of the EMSM model in the IFP-C3D code has been successfully demonstrated. Moreover, very encouraging results demonstrate the feasibility of the EMSM model for spray injection. References [1] S. de Chaisemartin, Eulerian models and numerical simulation of turbulent dispersion for polydisperse evaporation sprays, Ecole Centrale Paris, 2009. [2] J. Donea et al., Arbitrary Lagrangian-Eulerian methods, Encyclopedia of Computational Mechanics pp 413 - 437, 2004 [3] D. Kah, Taking into account polydispersity for the modeling of liquid fuel injection in internal combustion engines, Ecole Centrale Paris, 2010.
Date: Friday Sep 9, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: M. Massot, Laboratoire EM2C, UPR CNRS 288 – Ecole Centrale Paris – marc.massot@ecp.fr and Center for Turbulence Research – Stanford University –mmassot@stanford.edu
Title: Eulerian moment methods for evaporating polydisperse sprays: from mathematical issues to HPC
Abstract: The modeling and simulation of multiphase reacting flows covers a large spectrum of applications ranging from combustion in automobile and aeronautical engines to atmospheric pollution as well as biomedical engineering. In the framework of this talk, we will mainly focus on a disperse liquid phase carried by a gaseous flow field which can be either laminar or turbulent; however, this spray can be polydisperse, that is constituted of droplets with a large size spectrum. Thus, such flows involve a large range of temporal and spatial scales which have to be resolved in order to capture the dynamics of the phenomena and provide reliable and eventually predictive simulation tools. Even if the power of the computer ressources regularly increases, such very stiff problems can lead to serious numerical difficulties and prevent efficient multi-dimensional simulations. The purpose of this talk is to show that all the necessary steps in order to develop a new generation of computational code have to be designed at the same time with a high level of coherence: mathematical modeling through Eulerian moment methods, development of new dedicated stable and accurate numerical methods, implementation of optimized algorithms as well as validations of both model and methods using either experimental measurements or other codes. We will introduce both a new class of models for the direct numerical simulation of spray dynamics [1, 2, 3, 4, 5], as well as a set of dedicated numerical methods [6, 7, 2] and prove that such an approach has the ability, once validated [8, 9], to lead to high perfomance computing on parallel architectures [10], obtained during the Summer Program 2010 of the Center for Turbulence Research at Stanford University.
Date: Thursday Aug 11, 2011
Time: 4:00pm
Location: BLDG 530 ROOM 127
Speaker: Amable Liñán, Escuela de Ingeniería Aeronáutica, Universidad Politécnica de Madrid, Spain
Title: The initiation of self-sustained detonations in reactive gases
Abstract: The presentation will be devoted to the analysis of the initiation, by concentrated external energy sources, of self-sustained cylindrical or spherical detonation waves in gaseous reactive mixtures. The dynamics of the detonation front will be described in the fast reaction limit, when the thickness of the reaction layer that follows the shock front is very small compared with the shock radius. At early times, after the initiation of the external thermal energy deposition, the detonation front, which generates a strongly expanding flow, is overdriven; so that it is reached by expansion waves that decrease its velocity to the Chapman-Jouguet value, for which the expansion waves can no longer reach the front. In cylindrical and spherical detonations, the transition to the constant Chapman-Jouguet velocity occurs at a finite value of the detonation radius, which is scaled by the radius for which the energy released by the external source equals the heat released by the chemical reaction. A brief discussion will be given on how the reaction may be quenched by the expansion waves if the initiating energy is smaller than a critical value, thus failing to generate a self-propagating detonation wave (first described, independently, by Ya.B. Zeldovich and G.I.
Date: Tuesday Aug 2, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Chi-Wang Shu, Division of Applied Mathematics, Brown University
Title: High-order finite difference methods with subcell resolution for hyperbolic conservation laws with stiff reaction terms
Abstract: In this talk we will present the result of the joint research with Wang, Yee and Sjogreen on high order finite difference WENO schemes with subcell resolution for hyperbolic conservation laws with stiff reaction terms. For chemical reaction problems, when the reaction time scale is very small, e.g., orders of magnitude smaller than the fluid dynamics time scales, the problems will become very stiff. Wrong propagation speed of discontinuity may occur due to the under resolved numerical solutions in both the space and time. Our proposed method is a modified fractional step method which solves the convection step and reaction step separately. In the convection step, any high-order shock-capturing method can be used, e.g. a fifth-order finite difference WENO scheme is considered in this work. In the reaction step, a modified ODE solver is applied but with the flow variables in the shock region modified by the subcell resolution idea. This method maintains high order accuracy in space for smooth flows and it can capture the correct wave speeds in very coarse meshes with regular time steps. The properties of the proposed method are demonstrated by a wide range of numerical examples for scalar problems and reactive Euler problems in both one and two dimensions.
Date: Monday Aug 1, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Joseph Powers, University of Notre Dame
Title: Verified Calculation of Multiscale Combustion in Gaseous Mixtures
Abstract: Prediction of combustion of gaseous mixtures described by detailed kinetics and multicomponent transport is notoriously difficult due the multiscale nature of the phenomena. For common mixtures such as hydrogen-air at atmospheric pressure, it will be rigorously shown that the finest continuum reaction length scales are a few microns and are well predicted by a simple estimate from Maxwellian diffusion analysis. The coarsest reaction length scales are typically several orders of magnitude larger. Both fine and coarse reaction scales are usually dwarfed by geometric scales of an associated engineering device. For flows characterized by a finite number of thin fronts, a computational strategy for achieving a fully verified, direct numerical simulation (DNS) solution in a demanding multiscale environment is demonstrated. The strategy is based upon a massively parallel implementation of an adaptive algorithm built with the aid of a wavelet-based representation of all state variables in the flow field. Results are shown for inert and reactive compressible multi-dimensional Navier-Stokes models.
Date: Thursday Jul 28, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Shankar Ghosh, Postdoctoral Associate, Aerospace Engineering and Mechanics University of Minnesota, Minneapolis, MN
Title: Towards numerical simulation of high enthalpy turbulent flows
Abstract: The first part of the presentation describes numerical simulation of a laser-induced plasma and it's interaction with isotropic turbulence. The numerical method developed for this problem uses Fourier spectral spatial derivatives, characteristic based shock capturing and logarithm based formulation of the continuity equation. Local thermodynamic equilibrium conditions are assumed to apply. The post-energy deposition flow field is divided into shock formation, propagation and core roll up stages. Each stage is studied in detail and a mechanism is suggested for core roll up. Vorticity is found to be produced in the flow at short and long times due to different mechanisms. For the turbulent simulations, the effect of turbulence on the flow is studied for compressible and incompressible limits. Compressible turbulence is found to slow down the mean shock wave during formation and propagation stages and suppress core roll up in the mean. Also turbulence levels are found to be amplified in the vicinity of the shock wave due to presence of mean compression there. This effect is spatially inhomogeneous and non-stationary in time. Turbulent kinetic energy budgets are computed to explain the mechanism of transfer of energy between the mean flow and the background turbulence. The second part of the presentation describes a numerical method for simulation of high-enthalpy turbulent flows. A non-dissipative algorithm is used for accurate flux reconstruction at the cell faces. A predictor corrector based shock capturing scheme is used to simulate strong shock waves. The Navier-Stokes equations are suitably modified to represent various thermo-chemical processes occurring in high enthalpy flows. A five species model for air is considered. To account for finite rate chemical reactions, individual mass conservation equations are solved for every species. An equation for conservation of vibrational energy is also solved to account for vibrational excitation. The numerical method is evaluated using test problems.
Date: Monday Jul 18, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: ANTHONY RUIZ, CERFACS/Snecma, Toulouse, France
Title: DNS AND LES OF TURBULENT COMBUSTION IN LIQUID ROCKET ENGINES
Abstract: In a rocket engine at nominal operating point, combustion takes place at high pressure (typically 100 bar), well above the thermodynamic critical point of the reactants. At these supercritical conditions, the dense reactants injected into the combustion chamber do not follow the classical atomization, evaporation, and burning path. Due to vanishing surface tension and latent heat of vaporization, this reactive flow can be modeled as a single variable-density fluid and developing DNS or LES for such flows raises new challenges. In this talk, we first present the thermodynamic equations needed for the use of a real gas equation of state in an explicit variable-density solver. Then the stabilization of a 2D LOX/GH2 jet flame behind a splitter plate is studied using DNS of the reacting flow with realistic H2/O2 chemistry to gain insight into the combustion physics for this type of flame. Finally, a 3D LES of a coaxial LOX/GH2 jet flame, correspondingto the C60 Mascotte experiment performed at ONERA is presented.
Date: Friday Jul 15, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Thierry Poinsot, Université de Toulouse, CNRS and CERFACS
Title: LES AND ACOUSTIC STUDIES OF COMBUSTION INSTABILITIES IN ANNULAR COMBUSTION CHAMBERS
Abstract: Annular chambers found in most gas turbines exhibit strong unstable modes where the acoustics of the combustor interact with the flames and can destroy the whole gas turbine. We will describe the physics of these instabilities and present two approaches to study them: (1) brute force compressible LES on massively parallel machines and (2) acoustic solvers where the flow is frozen and only unstable acoustic modes are searched in frequency domain. The nature of these modes will be studied in two configurations: a small helicopter chamber (Turbomeca) and a large industrial gas turbine (Ansaldo). Validations will be discussed as well as possible path to use simulation results to control unstable modes.
Date: Friday Jul 8, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Haecheon Choi (visiting professor, CTR, Stanford University)
Title: Grid-point requirement for large eddy simulation and wall boundary condition for wall-modeled LES
Abstract: We revisit Chapman's estimate on the required numbers of grid points for wall-modeled and wall-resolving LES. Using more accurate correlations of the skin friction coefficient and boundary layer thickness for high Reynolds number boundary layer flow, we present new Reynolds-number dependence of required numbers of grid points. The importance of providing proper mean wall shear stress is examined for wall-modeled LES and tested for turbulent channel flow using very coarse grids. It is shown that mean velocity profile is predicted reasonably well even without wall modeling once the mean shear stress at the wall is accurately given. An extension of this approach to turbulent boundary layer flow is underway and is briefly discussed.
Date: Friday Jul 1, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: van Oijen - Combustion Technology, Mechanical Engineering Eindhoven University of Technology (TU/e) The Netherlands
Title: Modeling turbulent combustion with flamelet-generated manifolds
Abstract: Numerical simulation of combustion has become an important tool in the design and optimization process of modern gas turbines and engines. The ever more strict emission limits are the main driving force for the development of new combustion concepts based on future sustainable fuels. Numerical modeling of combustion is a challenging task. The many different chemical components and reactions introduce a wide range of length and time scales, which demand special numerical solvers. Due to the enormous computational costs, application of detailed reaction models is usually limited to simple academic flow problems. In order to simulate turbulent combustion in real engine geometries, reduction techniques have to be used which simplify the chemistry model without losing (too much) accuracy. A very efficient method is the Flamelet-Generated Manifold (FGM) method developed at TU/e. Since the FGM method was introduced by Van Oijen, it has been systematically analyzed and its application range has been extended to cover almost all combustion modes in existing engines, furnaces and gas-turbines. FGM and very similar methods have been adopted by many academic and industrial research groups, because of their simplicity, efficiency and accuracy. The concentrations of both major and minor species are accurately predicted while the computation time is reduced by approximately two orders of magnitude. The method is valid for laminar and turbulent combustion up to the thin reaction zone regime, in which the internal structure of the chemical reaction layer is not perturbed by turbulent flow structures. In the presentation, several aspects of FGM will be discussed. The procedure to generate FGMs and an analysis of their performance in several test cases will be presented. Special attention will be given to the modeling of preferential diffusion effects and stratification in premixed turbulent flames.
Date: Friday Jun 10, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr Danielle Moreau, Postdoctoral Research Associate, Aerospace, Acoustics and Autonomous Systems Research Group, School of Mechanical Engineering, The University of Adelaide, Australia
Title: An experimental investigation of flat plate trailing edge flow and noise at low-to-moderate Reynolds number
Abstract: Trailing edge noise is a major sound source in many noise sensitive applications that use airfoil shapes such as fans, rotors and propellers, wind turbines and underwater vehicles. This seminar presents results of a comprehensive experimental study on the flow and noise generated by the sharp trailing edge of a flat plate at low-to-moderate Reynolds number (Rec = 0.7 × 105 to 5.0 × 105, based on chord). Simultaneous measurements of the flow and farfield noise have been taken for three flat plate models with symmetric and asymmetric bevelled trailing edge geometries in an anechoic wind tunnel at the University of Adelaide. Far-field acoustic data are scaled according to existing theory and compared with predictions of trailing edge noise obtained using current semi-empirical methods. Mean and unsteady velocity data measured in the very near wake with hot-wire anemometry are related to the far-field noise measurements. One of the flat plate models with an asymmetric bevelled trailing edge is observed to radiate high amplitude tonal noise at low-to-moderate Reynolds number. Flow and far-field noise data reveal that, in this particular case, the tonal noise appears to be governed by vortex shedding processes at the trailing edge. Also related to the existence of the tonal noise is a region of separated flow slightly upstream of the trailing edge. Hydrodynamic fluctuations at selected vortex shedding frequencies are strongly amplified by the inflectional mean velocity profile in the separated shear layer. The amplified hydrodynamic fluctuations are diffracted by the trailing edge, producing strong tonal noise.
Date: Friday May 27, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Jan Nordstrom, Professor in Scientific Computing Department of Mathematics, Linkoping University, Sweden
Title: Initial Boundary Value Problems, Summation-By-Parts Operators and Weak Boundary Conditions with Multi-Physics Applications
Abstract: During the last decade, stable high order finite difference methods as well as finite volume methods applied to initial-boundary-value-problems have been developed. The stability is due to the use of so-called summation-by-parts operators (SBP), penalty techniques for implementing boundary and interface conditions, and the energy method for proving stability. In this talk we discuss some aspects of this technique including the relation to the initial-boundary-value-problem. By reusing the main ideas behind the recent development, new coupling precodures for multi-physics applications have been developed. We will present the theory by analyzing simple examples and apply to very complex multi-physics problems such as fluid flow problems, elastic and electromagnetic wave propagation, fluid-structure interaction and conjugate heat transfer.
Date: Friday May 20, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Professor Patrick Jenny, Institute of Fluid Dynamics, Zurich, Switzerland
Title: Modeling Premixed Turbulent Combustion Using a PDF Method and Scale-Separation
Abstract: Probability density function (PDF) methods for turbulent reactive flows have the well known advantage that averaging of reaction source terms requires no additional modeling. Moreover, if one solves for joint velocity-scalar PDFs, also turbulent convection appears in closed form, which is in particular attractive in situations where counter gradient diffusion may occur. So far, however, most of the success of PDF methods is attributed to their applications to non-premixed turbulent combustion. Here, a closure for premixed turbulent combustion in the corrugated flamelet regime is presented, which is based on scale separation and where a PDF method is employed to provide the required joint statistics of velocity, turbulence frequency, and a scalar vector. Unlike in most other PDF methods, the scalar source term not only describes reaction rates, but accounts for ignition of reactive unburnt fluid elements due to propagating embedded quasi laminar flames within a turbulent flame brush. If one assumes unperturbed embedded flame structures and a constant laminar flame speed (as more or less expected in the corrugated flamelet regime), then the mean flame surface density, and thus the probability for an individual fluid element (represented by a Lagrangian particle in the PDF method) to ignite during a time step, can rigorously be calculated within the presented modeling framework. This is achieved by introducing the following particle properties: a flag indicating whether a particle represents the unburnt mixture; a flame residence time, which allows to resolve the embedded quasi laminar flame structure; and a flag indicating whether the flame residence time lies within a specified range. Latter, together with precomputed one dimensional laminar flames, allows to accurately estimate the flame surface density and thus the ignition probability. In most flames, in addition to the turbulent flame brush, also molecular mixing of the products with a co-flow has to be modeled by a mixing model. It has to be emphasized however, that in the presented modeling approach latter is not critical for the turbulent premixed flame propagation. To validate the new model, PDF simulation results of three piloted methane-air Bunsen flames (Aachen flames) are compared with experimental data. At the end of the presentation, the necessary modeling steps to extend this approach for also non- premixed and partially premixed turbulent combustion are discussed.
Date: Friday May 13, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Stanley Ling, Department of Mechanical and Aerospace Engineering, University of Florida
Title: Particle dispersal in multiphase explosions
Abstract: Particle dispersal in multiphase explosions is an interesting and complicated phenomenon. Modeling and simulation of this problem are challenging due to the complex interactions between the particles and the compressible flow features. Previous experiments and direct numerical simulations (DNS) have shown that the particle force and heating during the interaction with shock waves can be much larger than those predicted by the standard quasi-steady correlations. However, little work has been done to improve the inter-phase interaction models. Based on recent advances in our understanding of particle force and heating in compressible flows, this work proposes a rigorous inter-phase interaction model for unsteady compressible multiphase flows that includes unsteady force and heating contributions. The model is first validated by comparing with DNS results for particle interaction with planar shock and detonation waves. The importance of unsteady contributions to force and heating are evaluated by measuring their peak values and net effects. The model is applied to investigate the particle dispersal in the classical explosion problem considered by Brode (1955). The simulation results show that ignoring compressibility and unsteady force and heating contributions in the inter-phase interaction model introduces significant errors.
Date: Thursday May 12, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: V. Olshevsky, Postdoctoral Scholar, Stanford University Center for Turbulence Research
Title: Some salient aspects of shock wave/boundary layer interaction
Abstract: Shock wave/boundary layer interaction (SWBLI) and shock/shock interference are two fascinating problems which are closely linked. They are present nearly in all high-speed flows. In particular, shock induced separation is at the origin of low-frequency unsteadiness, that can be highly detrimental for vehicle structure. The present talk will give an overview of some recent advances in LES modeling of SWBLI at M=2.3 over an adiabatic flat plate. Additionally, the study will address the question of the relevance of the remaining sub-grid terms appearing in the energy equation, terms that are often neglected under the assumption of weakly compressible small-scale turbulence. Also, the issue of filtering across the shock will be highlighted.
Date: Friday May 6, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Robert Martin, Staff Research Associate, Center for Energy Research, University of California San Diego
Title: Modeling and Validation of Mixture Separation Effects in Compressible Gasses
Abstract: For complex integrated systems such as hypersonic aircraft, high fidelity direct numerical simulations play a critical role in linking experimentally accessible physical phenomena to the tractable reduced order models necessary for design. However, in compressible turbulence modeling, gas kinetic effects resulting from disparate mass mixture components are often neglected even at the level of direct numerical simulation. In addition to Fickian diffusion driven by concentration gradients, kinetic theory predicts mass fluxes relative to mixture average velocity due to gradients in pressure and temperature as well as external fields. In this talk, the incorporation and validation of thermal- and baro-diffusive effects within the framework of an adaptively refined multi-dimensional shock capturing mixture fluid algorithm will be discussed. Through validation studies with respect to experimental data as well as prior analytical and numerical models, the impact these effects have on the composition and structure of inert shocks and under-expanded jets will be demonstrated. Comparison with Monte Carlo kinetic simulations also enables investigation of continuum breakdown for mixtures involving high Mach numbers, low density, and small spatial scales. Finally, potential regimes for which macroscopic turbulent mixing properties may be impacted by these effects will be briefly considered with particular emphasis on rarefied supersonic hydrogen injection as an area of future study.
Date: Friday Apr 15, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Alessandro Gomez, Prof. of Mechanical Engineering at Yale Center for Combustion Studies Department of Mechanical Engineering, Yale University
Title: Highly Turbulent Strained Flames: a Benchmark for Computational Studies?
Abstract: I will review ongoing experimental research in my group on turbulent nonpremixed and premixed flames in the counterflow configuration. This type of system was originally pioneered in the ‘60s for laminar flows and has been widely used in the combustion community in laminar flame studies. Recently developments in our laboratory have shown that it is also a promising benchmark for highly turbulent flames, at turbulent Reynolds numbers on the order of one thousands. The system may offer significant advantages from a computational viewpoint, including the following ones: a) the flame stabilization is aerodynamic, near the stagnation plane between two opposed jets of either reactants or reactants and products, with ensuing simplifications in the prescription of boundary conditions; b) the domain of interest is dramatically more compact as compared to conventional jet flames at the same Reynolds number; and c) the mean residence time is on the order of one millisecond, which is particularly useful from a DNS/LES standpoint, and to avoid soot complications in the combustion of practical fuels. Case studies will demonstrate the versatility of the system in mimicking real flame effects such as heat loss and flame stratification under well controlled conditions. Speaker’s biographical sketch: Professor Gomez received a Laurea in Ingegneria Aeronautica from University of Naples (Italy) in 1980, and a Masters and Ph.D. in Mechanical and Aerospace Engineering from Princeton University in 1982 and 1986, respectively. After a postdoctoral and lectureship experience in the Department of Chemical Engineering at Yale, he joined the Faculty of the Department of Mechanical Engineering in 1989, where he raised through the ranks to his current position as Professor. His research interests focus on fundamentals of combustion and of electrostatic spray processes with applications. Coauthor of more than 90 articles in the peer-reviewed literature, he has been the recipient of a NSF Young Investigator Award, the Whitby Award from the American Association for Aerosol Research and awards from the Fulbright Foundation, the Von Karman Institute of Fluid Mechanics, ATA Fiat Research Center, and Aeritalia. He is Associate Editor of Combustion Science and Technology and Director of the Yale Center for Combustion Studies. For further details see http://www.eng.yale.edu/gomez-lab/
Date: Friday Apr 8, 2011
Time: 4:00pm
Location: Building 300, Room 300
Speaker: Haecheon Choi, Professor School of Mechanical and Aerospace Engineering Seoul National University, Korea. Also Visiting Professor Department of Mechanical Engineering Stanford University
Title: Flow over a Sphere and Its Controls
Abstract: The flow over a sphere contains complicated flow phenomena, although its geometry is simple. We simulate the flow over a sphere in the subcritical regime (Re = 3700, 10,000 and 100,000) using large eddy simulation with an immersed boundary method. The computational results are compared with the experimental ones and some of the characteristics of flow over a sphere are described. Next, we briefly introduce passive and active controls applied to flow over a sphere. These controls include the dimples, surface trip wires, background turbulence, and surface blowing/suction. These controls provide more than 50% drag reduction, and we show that the drag-reduced flows are very similar to the flow over an uncontrolled sphere under drag crisis, suggesting an important flow control strategy for drag reduction. Finally, we present our on-going efforts on new design of golf ball.
Date: Friday Mar 4, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Julien Bodart, Postdoc, Stanford University Center for Turbulence Research
Title: DNS of shearless turbulent boundary layers
Abstract: The main purpose of studying shearless boundary layers is to understand how energy is transferred between normal and tangential velocity components in the near-wall region. In the context of wall-bounded turbulent flows, this remains one of the main issues to be solved to build appropriate second-order closure turbulence models. We revisit the problem of Perot and Moin (1995) where decaying homogeneous turbulence is interacting with a wall/free-slip surface. In the present case, the shearless turbulent boundary layer is sustained by applying an outer random forcing acting in a central layer of the domain. The produced shearless turbulence self-diffuses and interacts with a solid wall/free-slip surface on both sides, which makes the configuration comparable to oscillating grid experiments. The proposed DNS database of this case brings new conclusions: in addition to the viscosity, we identify the skewness of the velocity field in the outer region as the main responsible of the energy redistribution close to the wall and we propose new length scales associated with this transfer. Rotational events are also shown to play a role.
Date: Friday Feb 11, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Ivo F. Sbalzarini, Assistant Professor of Computational Science ETH Zurich
Title: From stochastic optimization to more efficient simulation algorithms (Ivo F. Sbalzarini, MOSAIC Group, ETH Zurich, Switzerland)
Abstract: Stochastic dynamics provides a rich framework for analyzing and designing optimization and simulation methods. In continuous black-box optimization, bio-inspired concepts from Darwinian evolution have triggered the development of numerous algorithms, including Genetic Algorithms and Evolution Strategies. Despite the success of these algorithms in real-world applications, however, the underlying stochastic dynamics remains poorly understood, and guarantees of solution quality and convergence are not available for any but the simplest toy cases. We analyze stochastic optimization algorithms as probability processes based on Gaussian Adaptation and illustrate several analogies and connections with stochastic processes in chemistry and physics, as well as with adaptive MCMC sampling. Exploiting these connections helps understand stochastic optimization algorithms, but also inspires more efficient algorithms for adaptive sampling and stochastic simulations. The latter is shown for chemical kinetics on the example of the PDM exact Stochastic Simulation Algorithm.
Date: Friday Feb 4, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Claude Cambon, Laboratoire de Mécanique des Fluids et d’Acoustique, École Centrale de Lyon, France
Title: What remains to be done using Rogallo’s techniques. Some proposals for Geophysical and Astrophysical Flows
Abstract: Unbounded turbulent shear flows subjected to large-scale effects of shear, rotation, stratification, MHD, can be studied using pseudo-spectral DNS in a comoving frame (Rogallo 1981). A brief historical survey is given, including connection with linear theory, sometime called "Rapid Distortion Theory", towards a fully nonlinear approach. Different communities use this approach, namely engineering, applied mathematics, geophysics and astrophysics. In the last community, numerous studies, even recent, are performed in the context of thin radially stratified accretion disks, but they use their own terminology: This is clarified here as far as possible. I will present studies in my team, which all take into account the gyroscopic torque induced by the misalignment of system vorticity, in a rotating frame, and vorticity of the rotating mean flow. The case of precessing rotating flow is first addressed. The gyroscopic torque induced by a precession can be exactly balanced by a mean shear, which in turn can trigger an instability very close to the "elliptical flow instability" (Kerswell 1994, Salhi & Cambon 2009). Just recall that turbulence forced by the basic eliptical flow instability was studied at CTR by Blaisdell & Shariff, and Mansour & Lundgren, namely. The second case is the turbulence forced by a baroclinic instability. In this case, the starting point is a mean shear flow rotating in the vertical direction and subjected to a vertical stabilizing mean density gradient. In this case, the gyroscopic torque resulting from vertical system vorticity and spanwise shear-induced vorticity is exactly balanced by an horizontal new component of the density gradient: This yields tilting the mean isopycnal lines vs. their horizontal direction, so that a baroclinic instability is triggered. Typical results from "RDT" and DNS are presented, following Salhi & Cambon (2006) Simon et al., Pieri et al.
Date: Wednesday Jan 26, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Robert Rosner – William E. Wrather Distinguished Service Professor Departments of Astronomy & Astrophysics and Physics, Enrico Fermi Institute, and Computational Institute The University of Chicago
Title: Key Problems of Plasma Astrophysics: A Foray into new Territories
Abstract: N/A
Date: Friday Jan 14, 2011
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Eric Darve
Title: Fast linear algebra using low rank properties of matrices
Abstract: N/A