CTR Seminars Archive 2008
Date: Friday Dec 12, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Riccardo Rossi, Laboratorio di Termofluidodinamica Computazionale, Università di Bologna
Title: Progress in the numerical simulation of scalar dispersion in complex flows
Abstract: The seminar addresses problems arising in the numerical simulation of passive scalar dispersion in complex geometries using RANS and DNS techniques. In the first part of the talk, a review of the theoretical background for gradient-transport modeling of turbulent transport is presented. In spite of the prevalent use of the standard gradient-diffusion hypothesis (SGDH) in the framework of RANS simulations, that is, a simple similarity with random molecular motion where the eddy diffusivity and the turbulent Schmidt number are introduced, the analysis shows that the SGDH model is inadequate for modeling turbulent scalar fluxes even in the case of simple shear flows, leading to the failure of predicting scalar dispersion under strongly inhomogeneous and spatially developing flow conditions. A significant improvement in the modeling of turbulent transport can be obtained through the generalized gradient-diffusion hypothesis (GGDH) and its high-order extension (HOGGDH), where an algebraic closure for turbulent scalar fluxes is adopted. If the basic constraints for the applicability of gradient-transport type models are satisfied and the Reynolds stresses anisotropy is reasonably captured, numerical experiments on scalar dispersion downstream of a square obstacle show that the use of algebraic models leads to reliable predictions of turbulent scalar fluxes even in the presence of the counter-diffusion process. The second part of the talk is concerned to the use of random-forcing techniques for the generation of incoming fully developed turbulence for the scalar dispersion problem downstream of a two-dimensional obstacle. A preliminary analysis of results will be presented, where turbulence statistics obtained from the forcing technique will be compared to the synthetic turbulent-inflow specification method previously adopted in the computations.
Date: Friday Nov 14, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Olaf Marxen, Stanford University, Postdoctoral Fellow
Title: Progress towards a better understanding of disturbance evolution in a laminar hypersonic boundary-layer with roughness elements
Abstract: Prediction of heat load on the surface of vehicles (re-)entering a planetary atmosphere is important for heat-shield design. As turbulent flow induces a much higher heating than laminar flow, the prediction of laminar-turbulent transition is a key factor in defining the dimensions and materials used for the thermal protection system. Yet, fundamental physical processes related to laminar-turbulent transition in high-speed boundary layers are not well understood. Our understanding is even less comprehensive if two- or three-dimensional roughness elements are present inside the boundary layer. Examples of localized roughness elements are fences in 2-d and bolts in 3-d that may be present on modern heat shields. High-speed, compressible boundary layers often exhibit qualitatively different phenomena than low-speed, incompressible ones, such as shocks and multiple instability modes. Appropriate simulation tools are necessary to accurately capture these physical phenomena associated with compressible boundary layers. The presence of roughness-induced shocks, boundary-layer separation, and vortical structures may lead to strong growth of instability modes and the generation of additional disturbances. Recent results obtained from numerical simulations of a boundary layer with two- and three-dimensional roughness elements will be discussed.
Date: Friday Oct 31, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Mohammed Zamir Afsar, University of Cambridge, United Kingdom
Title: Jet noise modeling
Abstract: In this presentation we show very accurate jet noise predictions can be made using an acoustic analogy. The analogy is based on a form of the linearized Navier Stokes equations derived by Goldstein (2002), and we use it to analyze the sound pressure of a non-heated jet flow. We develop a unified approach to jet noise modeling and start by showing how the jet noise spectrum can be thought of as being composed of two terms, one that accounts for the high frequency noise, and another term that represents the peak sound pressure. In this case, the sound predictions we show are based upon a Reynolds averaged Navier Stokes (RANS) calculation of the Stromberg jet, which has a Reynolds number (Re) of 3600 and Mach number (M) of 0.9. Although the jet noise predictions we obtain are reasonable, they require some empirical tuning of the turbulence properties. We therefore extend the jet noise model and show that very accurate noise predictions can be made without having any empirical tuning involved. The turbulence properties are now found by directly post processing a Large Eddy Simulation (LES) of the jet flow and in this particular case we analyze a high Reynolds number jet, where Re = 106 and M = 0.75. We show the LES-based turbulence properties are in good agreement with the data from experiment, for the forth order longitudinal correlation function. The final optimized jet noise model gives very accurate predictions across the spectrum for various observation locations, at 90°, and closer to the jet axis where the peak noise occurs.
Date: Friday Oct 24, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Eric Johnsen, Stanford University, Postdoctoral Fellow
Title: Accurate simulations of slowly moving shocks
Abstract: Many commonly used shock-capturing schemes exhibit unsteady errors when they are applied to problems in which shock waves move slowly compared to the grid. Though this drawback has been known for several decades, the underlying causes are not well understood; current fixes introduce extra dissipation and perform well for only for specific flow conditions. In this talk, we analyze and characterize the causes for such errors carefully. By specifying appropriate bounds on the wave speeds used in the HLL approximate Riemann solver, we find that no spurious oscillations are generated for first-order accurate methods. We further discuss the extension of this fix to higher-order accurate schemes. The effect of these errors on the flow field is illustrated by considering the interaction between a shock and an acoustic wave.
Date: Wednesday Sep 24, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. James Glimm, Department of Applied Mathematics and Statistics, University at Stony Brook
Title: The Mathematics and Numerics of Chaotic Mixing Flows
Abstract: This talk will be divided into two parts. In the first part we will explain the operation and use of Front Tracking from the point of view of a potential user, who might be interested in adding a front tracking capability to an existing CFD code. Locations for additional documentation will be presented. The second part of this talk will illustrate the use of Front Tracking for the study of turbulent mixing. There is current interest in combining the separate capabilities of capturing codes, which are efficient for shock waves (and, as is important, steep gradients of concentration, or near contact discontinuities) with accurate capabilities to model turbulent transport. For this part of the talk, we start by identifying the goals: to establish convergence for locations of primary waves and mixing zone edges (macro variables) and the joint PDFs for concentration and temperature (micro variables). Front tracking can be viewed as an enhanced version of the capturing codes, in which effective numerical control is obtained over contact discontinuities and steep solution gradients. To model turbulent transport in this framework, we have added the dynamic subgrid scale models. But the use we make of them is original, in that we do not present these models with a smooth solution, but rather one with steep concentration, shear and thermal gradients. For convergence of the above mentioned observables, we achieve in this manner what appears to be the efficient calculation of resolved quantities, even with high Schmidt numbers.
Date: Friday, September 12, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Hong Zhao, University of Illinois at Urbana-Champaign
Title: Simulating flow and flexible structure interactions at medium and low Reynolds number
Abstract: Flow and flexible structure interaction is important in many biological phenomena from insect flying to cellular interactions in the microcirculation. We discuss our efforts on simulating such systems in finite Reynolds number flows and in the special case of Stokes flows. The finite-Reynolds-number system consists of complex geometry elastic solid and its surrounding fluid, both of which are incompressible. The effect of the solid, under certain assumptions, is equivalent to a distribution of surface and body forces that are applied to an otherwise purely fluid system. The motion of system is hence governed by the Navier--Stokes equations with the additional forces due to the structure. These equations are discretized and efficiently solved by a fractional step method on the fixed Cartesian mesh, with the solid forces transferred to the Cartesian mesh via a momentum-conserving Galerkin projection. This algorithm is demonstrated by simulation results including the swimming of a model jellyfish. In the Stokes-flow limit, we consider the motion of closely packed red blood cells flowing in microcirculations. The Stokes flow system is solved by using a boundary integral equation method, which evaluates the boundary integrals with an overall computational cost of $O(N \log N)$ by using Ewald sums and subsequently smooth particle-mesh Ewald method. The cell structures are modeled as elastic membranes with finite bending modulus that enclose a more viscous hemoglobin solution relative to plasma. The surface geometries and variables are represented by spherical harmonic expansions, which result in high numerical accuracy and also enable robust stabilization through dealiasing. We present the simulation results for the relaxation time scale for deformed cells and the apparent viscosity of blood flow through narrow cylindrical tubes. These results agree well with the published experimental results.
Date: Friday, September 5, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Gregory C. Burton, Lawrence Livermore National Laboratory
Title: The Nonlinear LES (nLES) Method: A fundamental paradigm shift in turbulence modeling
Abstract: here
Date: Tuesday, August 26, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Jan Nordstrom, Uppsala University
Title: A Hybrid Methodology for Unsteady Compressible Flows
Abstract: Unsteady compressible flow problems exhibit many flow features such as vortices, shocks, turbulence, etc. It is highly unlikely that any single numerical method can be the "best one" for all cases. We discus the concept of hybrid methods and exemplify with problems that deal with complex geometry and shocks.
Date: Tuesday, August 19, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Andrew Keats, Mechanical Engineering, University of Waterloo
Title: Bayesian inference for source determination: application, background and extensions
Abstract: 1. Bayesian inference for source determination (PhD results): Being able to determine the emission source of a contaminant released into the atmosphere is important in many spheres of interest, such as defence, environmental management, emergency response, and risk assessment. This is an ill-posed inverse problem, which we solve using a Bayesian probabilistic framework. Here, Bayesian inference is applied to find the posterior probability density function (PDF) of the source parameters (location and strength) given a set of concentration measurements. It is shown how the source-receptor relationship required to determine the direct probability (likelihood) can be efficiently calculated using the adjoint of the transport equation for the scalar concentration. The posterior distribution of the source parameters is sampled using a Markov chain Monte Carlo (MCMC) method. The inverse source determination method is validated against real data sets acquired in a complex flow field in an urban (built-up) environment. 2. Numerics and expansion of scope: Probability theory, with its rich philosophical background, provides an effective basis for performing the tasks of scientific inference. This part of the talk discusses how probability theory can be used for model selection and decision making (in the first part, it was used for parameter estimation). Numerical methods for performing these tasks (e.g. MCMC, posterior sampling, nested sampling) are also surveyed. 3. Possible application to uncertainty quantification: Bayesian tools in-hand, we are now able to propose ways of quantifying model-based simulation uncertainties. A practical way to do UQ would be through the semi-empirical formulation of an error model. The model parameters would be estimated using MCMC, and possibly improved upon using decision-theoretic techniques.
Date: Monday, August 18, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Sergei Chumakov, Los Alamos National Lab
Title: On a priori testing and development of SGS models for Large Eddy Simulation
Abstract: Several recent results from a priori investigation of the modeled terms in LES will be presented. We focus on behavior of SGS stress, SGS energy dissipation and SGS scalar flux. For the SGS stress, we introduce two parameters that characterize the "state" of the stress and evaluate several models based on these two parameters. For the SGS energy dissipation rate, we investigate its scaling with the SGS kinetic energy and show that the conventional power-law assumption holds only in limited sense. For the SGS scalar flux, we show that it's closely coupled with the state of the SGS stress and evaluate a particular tensor-viscosity model that is free from user-defined constants. We present some preliminary data that indicates a possible alternative to modeling the SGS energy dissipation via Lagrangian approach. Also, we will discuss some a priori tests that are related to effect of turbulence on passive scalar reaction. Our tests indicate that in the case of fully developed turbulence it may be possible to parametrize the effect of turbulence on the gross reaction rate constant using only scalar variance.
Date: Wednesday, August 13, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Chi-Wang Shu, Brown University, Rhode Island
Title: High Order Shock Capturing Schemes --- An Overview
Abstract: In this talk we will give an overview of algorithm development and application, with an emphasis on recent progress, on high order shock capturing schemes. We will mainly discuss the finite difference weighted essentially non-oscillatory (WENO) schemes, finite volume WENO schemes, and discontinuous Galerkin (DG) finite element methods. A comparison of their relevant advantages and disadvantages will be given.
Date: June 20, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Laurent Jacquin, Director of the Fundamental/Experimental Aerodynamics Department of ONERA - Meudon
Title: On vortex dynamics and aerodynamics: from fundamentals to applications
Abstract: An updated review on mechanisms which drive the dynamics of vortices in fluid flows will be presented, with a special emphasis on medium and long wave instabilities which are particularly important for different aerodynamic applications, such as aircraft wake alleviation, jet mixing enhancement, flow separation control. Basic notions useful to appraise these concrete problems are inertial waves, cooperative instabilities, transient growths and turbulence in vortices. These notions will be surveyed and will be illustrated by means of new experimental and theoretical results.
Date: June 13, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Abdellah HADJADJ, Visiting Scholar, National Institute of Applied Sciences, INSA & CORIA - Rouen - France
Title: Progress in the development of high-order methods for supersonic flows
Abstract: This talk presents an overview of the main research developments in the area of internal and external high-speed aerodynamics. The current research is motivated by the desire to develop reliable numerical tools for predicting complex supersonic aerodynamics in real applications, especially for problems including shock waves interacting with instabilities and turbulence.
Date: May 23, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Robert Baurle, NASA Langley Research Center, Hampton, Virginia
Title: Modeling of High Speed Reacting Flows: Established Practices and Future Challenges
Abstract: Computational fluid dynamics (CFD) has proven to be an invaluable tool for the design and analysis of high-speed propulsion devices. Massively parallel computing, together with the maturation of robust CFD codes, has made it possible to perform simulations of complete engine flowpaths. Steady-state Reynolds-Averaged Navier-Stokes simulations are now routinely used in the scramjet engine development cycle to determine optimal fuel injector arrangements, investigate trends noted during testing, and extract various measures of engine efficiency. Unfortunately, the level of fidelity typically used for these analyses relies heavily on existing measurements (at similar flow conditions) to calibrate model coefficients on a case-by-case basis. The present lecture will highlight the current state-of-the-art practices used for scramjet flowpath analysis. This will be followed by a discussion of the obstacles (with regards to high speed computing and physical modeling) that must be overcome in order to expand these simulation practices, allowing more of the fundamental flow physics to be resolved. An overview of recent awards from the Hypersonics Project of the NASA Fundamental Aerodynamics Program (FAP) that addresses some of these modeling challenges will also be presented.
Date: May 16, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Mr. Senthil Radhakrishnan, University of Maryland
Title: Large-eddy Simulation of Non-equilibrium Flows
Abstract: Large-eddy simulation (LES) of wall-bounded flows is limited to moderate Reynolds number flows due to the high computational cost required to resolve the near wall eddies. LES can be extended to high Reynolds number flows by using wall-layer models which bypass the near-wall region and model its effect on the outer region. Wall-layer models based on equilibrium laws yield poor prediction in non-equilibrium flows, in which Wall-Modeled LES (WMLES) that model the near wall region by Reynolds-Averaged Navier-Stokes (RANS) equation and the outer region by LES, has the potential to yield better results. However, in attached equilibrium flows, WMLES under-predicts the skin friction due to slow generation of resolved eddies at the RANS/LES interface; application of stochastic forcing results in faster generation of resolved eddies and improved predictions. In this work, wall-layer models based on equilibrium laws and WMLES are tested for non-equilibrium flows. Flow over a contoured ramp, with a shallow separation followed by a recovery region, was studied. LES using equilibrium laws was unable to resolve the shallow separation. WMLES predicted the mean velocity reasonably well but over-predicted the Reynolds stresses in the separation and recovery region; application of the stochastic forcing corrected this error. Next, the flow past a two-dimensional bump, in which curvature and pressure-gradient effects dictate the flow development, was studied. WMLES predicted the mean velocity accurately but over-predicted the Reynolds stresses in the adverse pressure gradient region; application of the stochastic forcing also corrected this error. In summary, WMLES with stochastic forcing gave more accurate results than LES using equilibrium laws for the non-equilibrium flows studied.
Date: May 2, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Fabrizio Bisetti, CTR Postdoctoral Scholar
Title: Coupled Large Eddy Simulation and Filtered Density Function modeling methodology for turbulent non-premixed combustion
Abstract: As government agencies in the U.S. and around the world tighten the regulations on pollutant emissions and efficiency requirements for combustion devices, turbulent combustion modeling is playing an increasing role in the design of the next generation of internal combustion engines, turbine combustors, furnaces and burners. Among the most arduous tasks is the modeling of finite rate chemistry effects in non-premixed combustion. It is well known that when chemistry and flow time scales overlap, simplifying assumptions such as those invoked in steady-flamelet approaches become unrealistic. In this talk I will present my work on the filtered mass density function (FMDF) modeling approach coupled with large-eddy simulation (LES) as it applies to turbulent, non-equilibrium piloted jet flames. Such approach has been identified as a promising methodology to capture non-equilibrium and finite rate chemistry effects. An overview of the theoretical and practical aspects of the methodology will be presented first. I will then illustrate how the LES/FMDF method was applied to the simulation of a piloted methane-air non-equilibrium jet flame (Sandia/TUD Flame D test case) for which extensive experimental data is available. The obtained numerical results indicate that the LES/FMDF approach is capable of quantitatively reproducing the experimental data. It is also found that the results are sensitive to the choice of mixing model for subgrid-scale mixing, thus providing valuable insight on the relative performance of different models in the context of LES. Finally I will provide some perspective on the challenges involved in applying the LES/FMDF method to near-extinction turbulent jet flames (Sandia/TUD Flame F).
Date: April 18, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Marcel Ilie, Carleton University, Ottawa, Canada
Title: Quasi-3D aerodynamic, aeroacoustic and aeroelastic investigations of helicopter blade-vortex interaction
Abstract: The Blade-Vortex Interaction (BVI) phenomenon is one of the main sources of noise and vibrations for helicopter, and comprises one of the most complex unsteady flow features of helicopter rotor in forward flight. In rotorcraft, BVI produces impulsive, high amplitude, undesired noise (BVI noise) by the unsteady fluctuations in the blade loading due to the rapid changes of the induced velocity. The numerical simulation of BVI has been a challenge for the fact that most of the numerical techniques either tend to alter the characteristics of the vortex, before the blade-vortex interaction occurs or are not able to capture all the fluctuations of the aerodynamic coefficients. In the present study, aerodynamic and aeroacoustic numerical investigations of Airfoil-Vortex Interaction (AVI) have been conducted, using Large Eddy Simulation (LES). The simulations were performed for a subsonic flow, M = 0.3 and Reynolds number, Re = 1.3x106 based on the NACA 0012 airfoil chord length, c = 0.2m and free stream velocity U=100m/s. The influence of airfoil-vortex vertical miss distance, airfoil angle of attack and vortex characteristics on the airfoil-vortex mechanism of interaction were investigated. In the present study the influence of the aeroelastic response of airfoil on the aerodynamic coefficients was investigated as well. From the present study it was observed that the aeroelastic response of the airfoil, the airfoil-vortex vertical miss distance, the angle of attack and the vortex characteristics have a significant influence on the airfoil-vortex mechanism of interaction and implicitly on the aerodynamic coefficients and acoustic field. The results of the present study will be presented in detail.
Date: April 4, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Eric Johnsen, CTR Postdoctoral Scholar, Stanford University
Title: Numerical simulations of shock-interface interactions with applications to shockwave lithotripsy
Abstract: Interactions between shockwaves and interfaces play an important role in a wide range of applications, such as supernova explosions, mitigation of blast waves, hypersonic propulsion, inertial confinement fusion and cavitation events. However, physical and computational difficulties pose formidable challenges for accurate numerical simulations of such flow phenomena. In the first part of the talk, a numerical framework capable of capturing interface deformation and shockwave propagation is presented in the context of shockwave lithotripsy (SWL). SWL is a non-invasive medical procedure in which shockwaves are focused on kidney stones in an attempt to break them. Because the stones are usually immersed in liquid, cavitation occurs and contributes to stone comminution. The general properties of the non-spherical collapse of a single gas bubble are examined and the pressure along a nearby solid surface is computed as a measure of potential damage. In the second part of the talk, specific aspects of the simulation of turbulent flows with strong shocks and density variations are presented. In particular, recent findings regarding the problem of oscillations generated by slowly moving shocks are discussed. Finally, several multifluid algorithms are examined.
Date: March 28, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Mr. Markus Peer Rumpfkeil, University of Toronto Institute for Aerospace Studies
Title: Airfoil Optimization for Unsteady Flows with Application to High-Lift Noise Reduction
Abstract: The use of steady-state aerodynamic optimization methods in the CFD community is fairly well established. In particular, the use of adjoint methods has proved to be very beneficial because its cost is independent of the number of design variables. The application of numerical optimization to airframe-generated noise, however, has not received as much attention, but with the significant quieting of modern engines, airframe noise now competes with engine noise. Optimal control techniques for unsteady flows are needed in order to be able to reduce airframe-generated noise. In this talk, I present a general framework to calculate the gradient in a nonlinear unsteady flow environment via the discrete adjoint method thereby utilizing a Newton-Krylov approach. I then show two applications of this framework: the drag minimization of viscous flow around a rotating cylinder, and the minimization of the total radiated acoustic power in an unsteady laminar trailing-edge flow. In the second part of the talk, I present validation cases for an acoustic wave propagation program based on the Ffowcs Williams and Hawkings (FW-H) formulation as well as for a two-dimensional hybrid URANS/FW-H code. The general framework is then used to derive the adjointds in the CFD community is fairly well established. In particular, the use of adjoint methods has proved to be very beneficial because its cost is independent of the number of design variables. The application of numerical optimization to airframe-generated noise, however, has not received as much attention, but with the significant quieting of modern engines, airframe noise now competes with engine noise. Optimal control techniques for unsteady flows are needed in order to be able to reduce airframe-generated noise. In this talk, I present a general framework to calculate the gradient in a nonlinear unsteady flow environment via the discrete adjoint method thereby utilizing a Newton-Krylov approach. I then show two applications of this framework: the drag minimization of viscous flow around a rotating cylinder, and the minimization of the total radiated acoustic power in an unsteady laminar trailing-edge flow. In the second part of the talk, I present validation cases for an acoustic wave propagation program based on the Ffowcs Williams and Hawkings (FW-H) formulation as well as for a two-dimensional hybrid URANS/FW-H code. The general framework is then used to derive the adjoint equations for the hybrid URANS/FW-H code in order to be able to optimize the shape of airfoils based on their calculated far-field pressure fluctuations. I show validation results for this novel hybrid URANS/FW-H optimization algorithm using a remote inverse shape design problem involving laminar flow around a NACA0012 airfoil at a high angle of attack. Lastly, I present the results of the minimization of far-field pressure fluctuations of a turbulent blunt trailing edge flow. All the presented results show that it is possible to optimize the shape of an airfoids in the CFD community is fairly well established. In particular, the use of adjoint methods has proved to be very beneficial because its cost is independent of the number of design variables. The application of numerical optimization to airframe-generated noise, however, has not received as much attention, but with the significant quieting of modern engines, airframe noise now competes with engine noise. Optimal control techniques for unsteady flows are needed in order to be able to reduce airframe-generated noise. In this talk, I present a general framework to calculate the gradient in a nonlinear unsteady flow environment via the discrete adjoint method thereby utilizing a Newton-Krylov approach. I then show two applications of this framework: the drag minimization of viscous flow around a rotating cylinder, and the minimization of the total radiated acoustic power in an unsteady laminar trailing-edge flow. In the second part of the talk, I present validation cases for an acoustic wave propagation program based on the Ffowcs Williams and Hawkings (FW-H) formulation as well as for a two-dimensional hybrid URANS/FW-H code. The general framework is then used to derive the adjointds in the CFD community is fairly well established. In particular, the use of adjoint methods has proved to be very beneficial because its cost is independent of the number of design variables. The application of numerical optimization to airframe-generated noise, however, has not received as much attention, but with the significant quieting of modern engines, airframe noise now competes with engine noise. Optimal control techniques for unsteady flows are needed in order to be able to reduce airframe-generated noise. In this talk, I present a general framework to calculate the gradient in a nonlinear unsteady flow environment via the discrete adjoint method thereby utilizing a Newton-Krylov approach. I then show two applications of this framework: the drag minimization of viscous flow around a rotating cylinder, and the minimization of the total radiated acoustic power in an unsteady laminar trailing-edge flow. In the second part of the talk, I present validation cases for an acoustic wave propagation program based on the Ffowcs Williams and Hawkings (FW-H) formulation as well as for a two-dimensional hybrid URANS/FW-H code. The general framework is then used to derive the adjoint equations for the hybrid URANS/FW-H code in order to be able to optimize the shape of airfoils based on their calculated far-field pressure fluctuations. I show validation results for this novel hybrid URANS/FW-H optimization algorithm using a remote inverse shape design problem involving laminar flow around a NACA0012 airfoil at a high angle of attack. Lastly, I present the results of the minimization of far-field pressure fluctuations of a turbulent blunt trailing edge flow. All the presented results show that it is possible to optimize the shape of an airfoil in an unsteady flow environment to minimize its radiated far-field noise while maintaining good aerodynamic performance.
Date: February 22, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Prof. Robert Stein, Physics and Astronomy Department, Michigan State University
Title: Solar Surface Convection Simulations
Abstract: I will report on the current state of our solar surface convection simulations, both large, supergranulation scale (48 and 96 Mm wide by 20 Mm deep) hydrodynamic convection and high resolution mesogranule scale (6 Mm wide by 2.5 Mm deep) magneto-convection calculations. The large scale simulations have been and will continue to be useful for testing and refining various local helioseismic techniques. Once magnetic fields are added they will also be used to study the emergence of magnetic flux, the formation and evolution of the magnetic network, and the evolution of an active region. The small scale, very high resolution, calculation will be used to study the role of magnetic instabilities in the formation and evolution of faculae.
Date: February 15, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Jean Lachaud, NASA Postdoctoral Scholar
Title: Modeling of the ablation of carbon-based composites
Abstract: Context
Carbon-based composites are notably used as thermal protection systems (TPS) of planetary probes for atmospheric entry in the hypersonic regime. In such conditions, TPS surface is submitted to a high heat flux leading to a high surface temperature. Carbon-based materials help to protect the inner structure because they undergo some endothermic gasification phenomena (oxidation and/or sublimation). However, the resulting mass loss is accompanied by a surface recession, called ablation, which has to be controlled. The presented work is a contribution to the improvement of ablation understanding, the focus being set on the near-wall material/fluid interaction in laminar regime.
Presented work (PhD results - Bordeaux University, France)
First, the complex ablative behavior of carbon-based composites is experimentally analyzed. During ablation, these materials develop a multiscale surface roughness that follows their heterogeneous architecture. Surprisingly, their effective ablation velocity is shown to be higher than the intrinsic ablation velocities of their constitutive components. In order to explain these observations, a multiscale modeling strategy is set up; it follows the characteristic scales of the composites: nanoscopic (carbon nanostructure), microscopic (fiber, inter-fiber matrix), mesoscopic (yarn, inter-yarn matrix), and macroscopic (homogenized composite). At each scale, the proposed models notably account for the 3D recession of the wall, the heterogeneous gasification reactions, and mass transfer. A numerical simulation tool, based on Monte-Carlo Random Walks (reaction/mass transfer) and on a simplified Marching Cubes approach (front tracking), has been implemented to solve these models. The onset of surface roughness is correctly reproduced by numerical simulation at each scale. Using some numerically validated hypotheses, an analytical solution is obtained and used to infer by homogenization the ablation velocity of the composite from the properties of the composite components. These analytical results explain the surprising experimental behavior.
Outlook
In the future, this kind of phenomenological multiscale approach could be used to predict ablation velocity with an improved reliability, and to provide guidelines for the optimization of composite materials.
![[image]](/seminars/image022508.jpg){kind=link}
Date: January 25, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Panagiotis Stinis, CTR Postdoctoral Scholar, Stanford University
Title: Model reduction in physical and probability space
Abstract: The purpose of the talk is to give an overview of methods for the reduction of the complexity of computational problems in physical and/or probability space. In general, problem reduction leads to memory effects. I will present a brief introduction to the Mori-Zwanzig formalism and how it can account for theme mory effects. Illustrative examples will be provided.
Date: January 18, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Qianlong Liu, Department of Mechanical Engineering, University of Colorado at Boulder
Title: An Integrated Modeling and Simulation Approach for Flow-Generated Sound Prediction
Abstract: In order to develop accurate, efficient, and reliable numerical models of flow-generated sound generation and propagation in a complex physical environment, the following various challenges need to be comprehensively addressed: efficient representation of complex geometries, development of artificial and outflow boundary conditions that minimize unphysical reflections from artificial boundaries, adaptive grid generation methodologies, adequate resolution of energy containing/noise generating vortical structures, accurate modeling of vortex interaction with surfaces/edges for subsequent noise generation, and use of high-order numerical algorithms. A novel approach is developed to tackle all of these challenges by integrating the Brinkman penalization method, nonreflecting boundary conditions, Adaptive Wavelet Collocation Method, and Ffowcs Williams and Hawkings analogy into a unified framework for modeling and simulation of flow-generated sound prediction. In particular, a Brinkman penalization method is developed for compressible flows around solid obstacles of complex geometries. It is the first consistent Immersed Boundary Method for compressible flows. This method is based on a physically sound mathematical model for compressible flows through porous media. In addition, for numerical efficiency and affordability, nonreflecting boundary conditions based on nonlinear multidimensional characteristics are developed for two- and general flows governed by compressible Navier-Stokes equations with/without scalar transport equations. It is the first artificial boundary conditions for multidimensional compressible flows based on nonlinear multidimensional characteristics. This method substantially improves the classical one-dimensional characteristics-based nonreflecting boundary conditions for complex general flows.
Date: January 11, 2008
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Yoshiyuki Fujitsuna, General Manager, Eng. Dept. Engineering Research Association for Supersonic Transport Propulsion System
Title: Hypersonic/Supersonic Transports Propulsion System Research in Japan
Abstract: See here [Word Document]
Date: January 10, 2008 (Thursday)
Time: 4:00pm
Location: CTR Conference Room
Speaker: Dr. Carlos Bettencourt da Silva, Technical University of Lisbon, Portugal
Title: Dynamics, topology and geometry of the flow near the turbulent/non-turbulent interface in jets
Abstract: The present seminar focuses in the turbulent entrainment mechanism that exists in free shear flows such as mixing layers, wakes, and jets. In these flows the flow field can be divided into two regions. In one region the flow is turbulent (T) and its vorticity content is high, while in the other region the flow consists of largely irrotational (nonturbulent - NT) flow. The two flow regions are divided by the turbulent/nonturbulent (T/NT) interface where the turbulent entrainment mechanism takes place, by which a given fluid element from the irrotational zone becomes turbulent. This T/NT interface is very sharp and is continually deformed over a wide range of scales. Its thickness is of the order of the Taylor micro-scale. The mechanism of turbulent entrainment is still involved in a great deal of mystery, despite the great number of works devoted to it. Understanding of the physical mechanisms taking place at the T/NT interface is important in many natural and engineering flows since important exchanges of mass, momentum and passive or active scalar quantities take place across the T/NT interface. In the past it was assumed that the turbulent entrainment mechanism is mainly driven by "engulfing" motions caused by the large scale flow vortices, but recent experimental and numerical works give more support to the original model of Corrsin and Kistler (1954,1955) where the entrainment is primarily associated with small scale ("nibbling") eddy motions. Nevertheless, it is still argued that the entrainment and mixing rates are largely determined by the large scales of motion. The present seminar discusses the dynamics, topology and geometry of the fluid elements during the turbulent entrainment process using direct numerical simulations of turbulent plane jets at Re_lambda 120. Here we analyze four particular points in this context: (a) the geometry of the dissipation, (b) the geometry of the fluid elements, (c) the dynamics of the enstrophy and strain, and (d) the large scale coherent vortices. In the end, new light is shed upon the detailed physical processes involving both the large and the small scales of motion of turbulence within this important and complex problem.