CTR Seminars Archive 2017
Direct numerical simulation of a turbulent boundary layer with separation and reattachment over a wide range of Reynolds numbers
Date and Time: Friday, November 10, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Hiroyuki Abe
Separation and reattachment of a turbulent boundary layer are crucial issues in aeronautical and engineering applications since they are associated with upper bound of efficiency for the devices. Understanding of the underlying physics and the accurate prediction however may not be still sufficient especially for pressure-induced separated flows. In the present work, we have performed a series of direct numerical simulations (DNSs) for a pressure-induced turbulent separation bubble on a flat plate. Suction and blowing are imposed at the upper boundary for producing a separation bubble. The inlet Reynolds number Re_theta based on the freestream velocity and the momentum thickness is equal to 300, 600 and 900, the latter value being three times larger than that of the seminal DNS works (Spalart and Coleman 1997; Na and Moin 1998). In this talk, we first describe effects of varying suction and blowing (i.e. pressure gradients) and compare with the earlier DNS works. We then discuss the Reynolds-number dependence in a large separation bubble. Also discussed are possible scaling laws of the root-mean-square value of wall-pressure fluctuations established on the basis of the relationship with local maximum Reynolds stresses. Finally, the extended DNS work for Re_theta=1500 will be introduced briefly.
Bio:
Dr. Hiroyuki Abe is an Associate Senior Researcher in Japan Aerospace Exploration Agency. He received his PhD in Mechanical Engineering from Tokyo University of Science in 2002. His research interests include direct numerical simulation of wall turbulence (channel and boundary layer), flow physics regarding both large and small scales, scalar transport, flow separation, and turbulence modeling.
-----------------------------------
Optimal Spectral Approximation Methods for Parameterized PDEs with Application to Flow Problems
Date and Time: Friday, November 3, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Kookjin Lee
We consider the numerical solution of parameterized linear systems where the system matrix, the solution, and the right-hand side are parameterized by a set of uncertain input parameters. We explore spectral methods in which the solutions are approximated in a chosen finite-dimensional subspace. It has been shown that the stochastic Galerkin projection technique typically does not minimize any measure of the solution error. As a remedy for this, we propose a novel stochastic least-squares Petrov–Galerkin (LSPG) method. The proposed method is optimal in the sense that it produces the solution that minimizes a weighted l2-norm of the residual over all solutions in a given finite-dimensional subspace. Moreover, the method can be adapted to minimize the solution error in different weighted l2-norms by simply applying a weighting function within the least-squares formulation. In addition, a goal-oriented semi-norm induced by an output quantity of interest can be minimized by defining a weighting function as a linear functional of the solution. Extensive numerical experiments show that the weighted LSPG methods outperforms other spectral methods in minimizing corresponding target weighted norms.
Bio:
Kookjin Lee is a Ph.D. candidate in the Computer Science department at the University of Maryland College Park, working with Professor Howard Elman. Kookjin’s primary research interest lies in uncertainty quantification, designing and developing efficient and optimal solution algorithms for parameterized PDEs. He has developed fast iterative low-rank solvers for linear and nonlinear parameterized PDEs, and optimal spectral approximation methods for parameterized PDEs. He received his B.S. and M.S. in Computer Science and Engineering from Seoul National University.
------------------------------------
Detailed modeling of radiative heat transfer in combustion systems
Date and Time: Friday, October 20, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Xinyu Zhao
Radiative heat transfer process in combustion systems has received relatively little attention to date. Recently, it starts to generate increasing interest given the current trends of engine designs for both internal combustion engines and aeronautical engines. Meanwhile, the need to properly model radiative heat transfer in fire-related scenarios is another driver of the interest. Radiation inside combustion systems is a complex process involving the interactions between spectral gases, soot, droplets, turbulence and the enclosure geometry. Gasses and soot/wall have distinct emission and absorption characteristics, and particles such as spray droplets or water mists have strong scattering effects that might alter the distribution of the heat flux. The turbulent fluctuations in temperature and composition add further complexity to the problem by affecting the production and destruction of soot in a significant manner. The complex geometries encountered in conventional engine or enclosure fires create difficulties in measuring and modeling the heat transfer processes, which hinders the understanding of the physical processes. In this seminar, first-principle-based high-fidelity models are introduced to study the radiative heat transfer process under engine-relevant conditions. Details of the high-fidelity models will be explained, and comparison with popular reduced-order models will be made. The physics will be discussed by means of case studies of recent simulations.
Bio:
Dr. Xinyu Zhao is an Assistant Professor at the University of Connecticut. She joined the Mechanical Engineering Department in 2015 Spring, and prior to that, she was a postdoctoral research fellow in Combustion Energy Frontier Research Center at Princeton (2014). She received her Ph.D. degree in Mechanical Engineering from Pennsylvania State University (2014), and she received her Bachelor’s and Master’s degrees in Thermal Engineering from Tsinghua University (2006 and 2008).
--------------------------------
Modeling velocity gradients along particle trajectories in turbulence
Date and Time: Friday, September 29, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Perry Johnson
The dynamics of velocity gradients in turbulent flows not only represent an attractive way to explore theoretical issues such as intermittency, but also prove important for a number of micro-physical processes that occur in turbulent environments. Inspired by the qualitative success of the restricted Euler model for Lagrangian evolution of velocity gradients, this talk will introduce a stochastic Lagrangian model carefully constructed for homogeneous isotropic turbulence. Using the local isotropy hypothesis, it will be demonstrated how such a model can be applied in large-eddy simulations (LES) to capture important sub-grid micro-physics. Extensions to higher Reynolds numbers (intermittency) as well as inertial particle trajectories will also be considered.
Bio:
Dr. Perry Johnson recently finished his PhD in Mechanical Engineering at Johns Hopkins and joined CTR as a postdoctoral scholar this September 2017. His research interests include small-scale turbulence, particle-laden flows, and near-wall dynamics.
---------------------------------
Transition to turbulence in wall-bounded shear Flows
Date and Time: Friday, September 22, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Michael Karp
In this talk several aspects of transition to turbulence in wall-bounded shear flows are addressed. One aspect discusses the formation and evolution of coherent structures, such as counter-rotating vortex pairs (CVPs) and hairpins, observed in various transitional as well as turbulent flows. The vortex dynamics is followed using a novel analytical-based numerical method for the evolution of localized disturbances in homogenous shear base-flows. Using insights gained from the evolution of localized disturbances, a minimal element model, capable of following the evolution of packets of hairpins, is developed. The model is compared successfully with experiments and the excellent agreement in all cases demonstrates the universality and robustness of the model.
The other aspect discusses transition from an instability point of view. An analytical model for subcritical transition via the transient growth mechanism is developed. The linear transient growth mechanism is represented analytically by four decaying normal modes and their nonlinear interactions. The model utilizes separation of scales between the slowly evolving base flow and the rapidly evolving secondary disturbance to capture most transition stages using the multiple time scales method. The model predictions are verified by comparison with direct numerical simulations. It is shown that the most dangerous secondary disturbances are associated with spanwise wavenumbers, which generate the strongest inflection points, i.e. those having maximal shear, rather than with those maximizing the energy gain during the transient growth phase.
Bio:
Dr. Michael Karp is a Postdoctoral Fellow in the Center for Turbulence Research at Stanford University. Karp received all of his degrees from the Faculty of Aerospace Engineering, Technion (BSc 2007, Msc 2013, PhD 2017). His dissertation combined analytical and numerical methods for understanding transition in wall-bounded shear flows. Karp’s research interests include Aerodynamics, Fluid Mechanics, Flow Instabilities, Transition to turbulence, Flow Control and Flight Mechanics.
---------------------------------
Sub-grid Scale Closures for LES from a Statistical Mechanics Perspective
Date and Time: Friday, September 8, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Professor Karthik Duraisamy
This talk will address the issue of sub-grid closure in large eddy simulations, leveraging ideas from non-equilibrium statistical mechanics. The approach is based on the Mori-Zwanzig (M-Z) formalism, which provides a framework to re-cast a high-dimensional dynamical system into an equivalent, lower-dimensional system. In this reduced system, which is in the form of a generalized Langevin equation (GLE), the effect of the unresolved modes on the resolved modes appears as a convolution integral (which is sometimes referred to as memory). The appearance of the memory term in the GLE demonstrates that, coarse-graining non-scale-separated systems (such as turbulence) leads to non-local effects. The M-Z formalism alone does not lead to a reduction in computational complexity as it requires the solution of the orthogonal (unresolved) dynamics equation. A model for the memory is constructed by assuming that memory effects have a finite temporal support and by exploiting the Germano identity. The appeal of the proposed model, is that it is parameter-free and has a structural form imposed by the mathematics of the coarse-graining process (rather than the phenomenological assumptions made by the modeler, such as in classical subgrid scale models). Demonstrations are presented in the context of predictive LES for rotating and non-rotating turbulence, and turbulent channel flow. Recent results on extending these techniques in the context of Discontinuous Galerkin discretizations and will also be presented.
Relevant Papers:
1. Parish, E. and Duraisamy, K., “A Dynamic Sub-grid Scale Model for Large Eddy Simulations based on the Mori-Zwanzig formalism," Journal of Computational Physics, 2017.
2. Gouasmi, A., Parish, E., and Duraisamy, K., “Characterizing Memory Effects in Coarse-Grained Nonlinear Systems Using the Mori-Zwanzig formalism," Under revision, Proceedings of Royal Society Series A, 2017. arXiv:1611.06277.
3. Parish, E. and Duraisamy, K., “Non-local closure models for Large Eddy Simulations based on the Mori-Zwanzig formalism," Physical Review Fluids, 2017.
Bio:
Karthik Duraisamy is an Associate Professor of Aerospace Engineering at the University of Michigan, Ann Arbor. He obtained a doctorate in aerospace engineering and masters in applied mathematics from the University of Maryland, College Park. Prior to his current position, he was at Stanford University (2009-2013). He is the director of the Center for Data-driven Computational Physics at the University of Michigan. His research interests are in turbulence modeling and simulations, data-driven and reduced order modeling and numerical methods for PDEs.
--------------------------------------
Energy Stable Boundary Conditions for the Nonlinear Incompressible Navier-Stokes Equations
Date and Time: Friday, September 1, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Jan Nordström
We derive boundary conditions for the nonlinear incompressible Navier-Stokes equations following the general recipie given in [1]. We present two formulations stemming from different techniques to diagonalize the boundary terms. Both formulations lead to an energy estimate.
In the first formulation, the boundary conditions are obtained through a suitable set of rotations. In the second formulation, the boundary conditions are derived directly by a standard eigenvalue decomposition [2, 3]. The two formulations differ in character and have different pro’s and con’s.
The rotational technique lead to more natural formulations, but the formulation must be changed depending on whether there is inflow or outflow. The rotation formulation does not always lead to a nonlinear bound.
The characteristic technique leads to more involved formulations, but the formulation retains the same form independent of whether there is inflow or outflow. The characteristic formulation provides a nonlinear bound for the velocity field for both solid wall and far field boundary conditions.
The continuous problem is approximated by using finite differences on Summation-By-Parts (SBP) form. The solid wall boundary conditions are weakly imposed with the Simultaneous-Approximation-Term (SAT) procedure [4]. It is shown that by mimicking the continuous analysis, the resulting nonlinear SBP-SAT scheme is provably energy stable, divergence free and high-order accurate.
[1] J. Nordström, “A Roadmap to Well Posed and Stable Problems in Computational Physics,” Accepted in Journal of Scientific Computing.
[2] J. Nordström, N. Nordin, and D. Henningson, “The Fringe Region Technique and the Fourier-method Used in the Direct Numerical Simulation of Spatially Evolving Viscous Flows,” SIAM Journal of Scientific Computing, Vol. 20, No. 4, pp.1365-1393, 1999.
[3] J. Nordström, K. Mattsson, and C. Swanson, “Boundary Conditions for a Divergence Free Velocity-Pressure Formulation of the NavierStokes Equations,” Journal of Computational Physics, Vol. 225, Issue 1, pp. 874-8901, 2007.
[4] M. Svärd and J. Nordström, “Review of summation-by-parts schemes for initial-boundary-value problems,” Journal of Computational Physics, Vol. 268, pp. 17-38, 2014.
Bio:
Since 2010 Jan Nordström is a Professor in Scientific Computing and since 2012 serves as the Head of the Division of Computational Mathematics, Department of Mathematics, Linköping University. He was a Research Scientist at the Aeronautical Research Institute of Sweden (FFA) from 1980 to 1995. Nordström serves on the board of Linköping Institute of Technology (LiTH) and the National Supercomputer Center (NSC). Education: 1980 MSc (civ ing) Master of Science in Aeronautics, Royal Institute of Technology (KTH) in Stockholm, Sweden; 1993 PhD (Tekn. Dr) in Numerical Analysis, Department of Scientific Computing, Uppsala University (UU), Sweden; 1999 Docent (Habilitation) in Numerical Analysis, UU.
-------------------------------------
Smoothed Profile Method for Modeling Particulate Flows
Date and Time: Friday, August 25, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Fazlolah Mohaghegh (Ehsan)
This study shows recent developments in the Smoothed Profile Method (SPM) as a diffuse-interface approach. SPM is well suited towards simulating the dynamics of dense particulate flows at the meso-scale, i.e. when particles are resolved. Several challenges are addressed: extension of the diffuse interface method for a wide range of Reynolds numbers, tackling the stability of computations in an FSI setting with added mass effects, and developing a parallel implementation with adaptive mesh refinement to handle large ensembles of particles. This research also offers a new particle collision model for arbitrary shape particles. The model is unique in the sense that it is simple, does not change the computational cost and does not have case sensitive parameters.
Bio:
Fazlolah Mohaghegh (Ehsan) is a PhD Candidate at the University of Iowa working under supervision of Professor Udaykumar. He received his BS and MS in Mechanical Engineering from Sharif University of Technology (SUT) and Iran University of Science and Technology (IUST). Before starting his PhD, Mohaghegh was a member of IUST CFD lab and worked on the CFD simulation of multiphase flows for different applications such as spray dryers, spray combustion systems and powder cladding devices. In his PhD, he worked on the development and implementation of diffuse interface immersed boundary methods to model fluid-structure interaction of rigid objects. Most of Mohaghegh’s research focused on the modeling of blood as a particle laden flow in the hinge gap region of a mechanical heart valve.
------------------------------------
Model-based and data-based flow analysis using optimization
Date and Time: Friday, August 18, 2017 - 16:15
Location: Yang & Yamazaki Environment & Energy Building (Y2E2), Room 111, 473 Via Ortega (Science and Engineering Quad, near the corner of Via Ortega and Panama Street)
Event Sponsor: ICME and ME Special Seminar
Speaker(s): Professor Peter Schmid
In recent years, PDE-constrained optimization has become an effective and efficient tool in the analysis of complex fluid systems. Inherent stability, receptivity to external or internal forcing, or sensitivity to uncertainties or imperfections of fluid systems can be treated within this approach. We will present a computational framework based on this concept and demonstrate its ability to extract relevant information from numerical simulations, with examples from aero-acoustics, inertial mixing, roughness-induced receptivity and turbomachinery cascades.
We will also discuss the reformulation of the PDE-constrained into a data-constrained framework and show preliminary steps in the analysis of fluid systems based on data only. We will present work in progress on phase-space clustering, data assimilation and dynamic observers to detect and describe relevant mechanisms and coherent structures in data sequences.
Bio:
Peter Schmid is Chair Professor of Applied Mathematics and Mathematical Physics at Imperial College London. Before joining Imperial he held a research director position at the French National Research Agengy (CNRS) and was Professor (PCC) of Mechanics at Ecole Polytechnique in France. He did his undergraduate and graduate studies in aerospace engineering at the Technical University Munich and obtained his doctoral degree in Mathematics from the Massachusetts Institute of Technology, after which he joined the faculty of the Department of Applied Mathematics at the University of Washington in Seattle, USA. His main research interests lie in theoretical and computational fluid mechanics with an emphasis on hydrodynamic stability theory, flow control, model reduction, system identification and the analysis of a wide range of fluid flows using adjoint, sensitivity and optimisation techniques. He is also interested in quantitative flow analysis using data-driven decomposition methods.
--------------------------------------
A multifractal formulation of the momentum cascade in wall-bounded turbulence
Date and Time: Friday, August 4, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Xiang I. A. Yang
The cascading process of turbulent kinetic energy from large-scale fluid motions to small-scale and lesser-scale fluid motions in isotropic turbulence may be modeled as a hierarchical random multiplicative process according to the multifractal formalism. In this work, we show that the same formalism might also be used to model the cascading process of momentum in wall-bounded turbulent flows. However, instead of being a multiplicative process, the momentum cascade process is additive. The proposed multifractal model is used for describing the flow kinematics of the low-pass filtered streamwise wall-shear stress fluctuation τl’, where l is the filtering length scale. According to the multifractal formalism, <τ’2>~ log(Reτ) and <exp(pτl’) >~(L/l)ζp in the log-region, where Reτ is the friction Reynolds number, p is a real number, L is an outer length scale and ζp is the anomalous exponent of the momentum cascade. These scalings are supported by the data from a direct numerical simulation of channel flow at Reτ=4200.
Bio:
Dr. Xiang I. A. Yang is a Postdoctoral Fellow in the Center for Turbulence Research at Stanford University. Yang received his Bachelor’s degree in Mechanical Engineering from Peking University in 2012 and his Ph.D. in Mechanical Engineering in 2016 from Johns Hopkins University.
----------------------------------------
Characterization of buoyancy-driven turbulent flows over inclines
Date and Time: Friday, July 21, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Marco Giometto, Postdoctoral Fellow, Center for Turbulence Research, Stanford University
Buoyancy driven turbulent flows over inclines will be considered within the conceptual framework of the Prandtl slope-flow model. Such flows are ubiquitous in engineering and the environment, but despite decades of active research, they remain a poorly understood problem. The sensitivity of the system to the model parameters is first characterized via direct numerical simulations, with a focus on variations in mean flow, second order statistics, and budget equations for the mean- and turbulent-kinetic energy. Turbulent flows resulting from upward moving warm fluid and downward moving cold fluid are identical for the vertical wall setup (up to the sign of the along-slope velocity), but undergo a different transition in the mechanisms sustaining turbulence as the sloping angle decreases. It will be shown how such a behavior results in stark differences between the two flows over shallow inclines. An analytic solution is then sought for the mean flow and buoyancy profiles as a function of the imposed surface buoyancy, the (constant) Brunt Väisälä frequency, the roughness height, and the parameters of the eddy viscosity model. Analytic asymptotic expansions are developed for both near-wall and outer regions. Matching across an overlap region allows for the construction of a composite expansion which reveals the structure of the wall jet and the associated buoyancy perturbation. The present asymptotic solution will be compared to existing data as well as numerical and alternative solutions.
Bio:
Dr. Marco Giometto is a Postdoctoral Fellow in the Center for Turbulence Research at Stanford University. He received his B.Sc. and M.Sc. degrees from the University of Padua, a Ph.D. in Civil and Environmental Engineering from Braunschweig TU University (2014), and a second Ph.D. in Mechanical Engineering from École Polytechnique Fédérale de Lausanne (2016). His research interests center around the study of boundary-layer flows that do not conform to classical similarity theories, such as buoyancy driven flows over inclines, canopy flows, and non-equilibrium turbulent boundary layers. He uses a combination of theoretical and numerical approaches to gain insight on such problems and to develop simple predictive models for use in interdisciplinary applications.
-------------------------------
Observation of laminar to turbulent transition in an already turbulent flow
Date and Time: Friday, July 7, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Professor Shuisheng He, Department of Mechanical Engineering, University of Sheffield, United Kingdom
Professor He will present a new perspective of transient turbulent flow and show that in such flows, turbulence does not progressively evolve from one state to another. Instead, the flow is characterised by the development of a laminar boundary layer followed by transition to turbulence. The talk will begin with a review of DNS results, followed by discussing recent laboratory experiments of a flow accelerating from an initially turbulent state following the opening of a valve, together with LES of the experiments and extended Stokes first problem solutions for the early stages of the flow. As previously predicted by DNS simulations of transient flow following a near-step increase in flow rate, the flow has been found to respond to the acceleration in a manner that is closely analogous to a laminar flow accelerating from rest. In both cases, the primary consequence of the acceleration is the temporal growth of a boundary layer from the wall, gradually leading to a strong instability causing transition. In this interpretation, the initial turbulent fluctuations can be regarded as noise in an otherwise well-defined flow behaviour. We observe the spontaneous appearance of turbulent spots and discontinuities in the velocity signals in time and space, revealing rich detail of the transition process, including a striking contrast between streamwise and wall-normal fluctuating velocities.
Bio:
Professor Shuisheng He is the Chair in Thermofluids in Department of Mechanical Engineering at the University of Sheffield in the United Kingdom (UK). He received his BSc and MSc from Huazhong University of Science and Technology (HUST) in China and his PhD from the University of Manchester in the UK. He has spent over three years with British Energy (now EDF Energy) as a reactor thermal hydraulics analyst, before starting an academic career with Robert Gordon University, and later the University of Aberdeen, and since 2011, the University of Sheffield. Professor He’s research focuses on non-equilibrium turbulence, including unsteady and buoyancy-influenced flows, nuclear thermal hydraulics, and computational fluid dynamics using RANS, LES and DNS. He is a Member of UK EPSRC Review College, Member of UK Turbulence Consortium, and Fellow of the Institution of Mechanical Engineers (IMechE).
------------------------------------
Clustering of Inertial Aerosols in Homogeneously Sheared Gas
Date and Time: Friday, June 9, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Mohamed Houssem Kasbaoui, PhD Candidate, Aerospace Engineering, Sibley School of Mechanical and Aerospace Engineering, Cornell University
Particle-laden flows of sedimenting small heavy solid particles or droplets in a carrier gas have strong inter-phase coupling even at low volume fractions. The slip velocity between phases leads to sustained clustering that strongly modulates the overall flow. The analysis of inertial aerosols in homogeneous shear reveals three fundamental mechanisms contributing to the formation of clusters: (1) the preferential concentration of inertial particles in the stretching regions of the flow (2) particle-trajectory crossing (PTC) and (3) a Rayleigh-Taylor instability due to the vertical stacking of particle-rich and particle-depleted regions. Simulations are conducted in Euler-Lagrange and Euler-Euler formalisms. The Euler-Lagrange simulation method, based on particle tracking, capture all three effects but suffer from lack of scalability. Euler-Euler methods, based on a kinetic description, offer better scalability but require extreme care in the presence of PTC.
Bio:
Mohamed Houssem Kasbaoui holds a Diplome d’Ingenieur from Ecole Centrale Paris with an Applied Mathematics Major, and a Master of Science in Physics from University Paris-Sud. He joined Cornell University in 2012 to pursue a PhD Degree in Aerospace Engineering with the goal to characterize the interactions between inertial particles and turbulence in particle-laden flows. His research focuses on the role of cascading fundamental instabilities on the formation of particle clusters. In addition to the modeling efforts, Kasbaoui uses and develops extensive Computational Fluid Dynamics (CFD) tools to provide input to the models. He is co-advised by Professors Olivier Desjardins and Donald L. Koch.
--------------------------------
Direct numerical simulation of droplet-laden isotropic turbulence
Date and Time: Friday, June 2, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Michael S. Dodd, Department of Aeronautics and Astronautics, University of Washington
Understanding how droplets and turbulence interact is important in numerous applications ranging from rain formation to oil spills to spray combustion, yet most of our knowledge of two-phase turbulence is limited to solid particles. In contrast to solid particles, droplets introduce new phenomena and parameters into the flow, e.g., droplets can deform, break up, coalesce, develop internal circulation, and change phase. In this talk, Dodd will review efforts towards improved understanding of droplet-turbulence interaction. Dodd uses direct numerical simulation (DNS) to capture the exchange of momentum, heat, and mass between the droplets and surrounding flow, while fully resolving the turbulence. First, Dodd will show DNS results of non-evaporating droplets of Taylor length scale size in decaying isotropic turbulence. The pathways for turbulent kinetic energy (TKE) exchange between the turbulent carrier flow and the flow inside the droplets will be explained starting from TKE budget equations. Dodd will show how increasing the viscosity ratio between the droplets and carrier fluid increases the decay rate of TKE and explain the underlying physical mechanisms. Second, Dodd will present preliminary results of DNS of evaporating droplets in isotropic turbulence.
Bio:
Michael S. Dodd is a Ph.D. Candidate in the Department of Aeronautics and Astronautics at the University of Washington. He earned his B.S. in Aerospace Engineering from the University of Michigan. Dodd uses direct numerical simulation to gain fundamental understanding of droplet-turbulence interaction. He also works on developing numerical methods for gas-liquid turbulent flows.
-------------------------------------
Simulating solids like fluids: A fully Eulerian approach to fluid-structure interaction
Date and Time: Friday, May 26, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Professor Kenneth M. Kamrin, Department of Mechanical Engineering, Massachusetts Institute of Technology
Fluids and solids tend to be addressed using distinct computational approaches. Solid deformation is most commonly simulated with Lagrangian finite-element methods, whereas fluid flow is amenable to Eulerian-frame approaches such as finite difference and finite volume methods. Problems that mix fluid and solid behaviors simultaneously present interesting numerical challenges. This is true when fluids and solids occupy different regions of space --- i.e. fluid-structure interaction (FSI) --- or in cases where materials behave like a solid but can undergo enormous levels of plastic flow more common of fluids --- i.e. granular materials and yield stress fluids.
Professor Kamrin focuses on FSI, and discusses a new method called the Reference Map Technique, which allows us to simulate deformable solids on a fixed Eulerian grid. The key is to store and update the reference map field on the grid, which tracks the inverse motion. Using this technique to represent the solid phase, we can solve FSI problems on a single fixed grid
using fast update procedures very similar to those used in two-phase Navier-Stokes fluid simulations. Various solid constitutive behaviors can be used, such as nonlinear elasticity and plasticity. Systems of many submerged and interacting solids can be simulated, and, by activating the solids internally, we can simulate systems of "soft swimmers". Incompressibility constraints can be applied in all phases by adopting Eulerian projection approaches commonly used in CFD. The addition of the reference map field to the grid also presents certain benefits when computing level-set interface advection, including a procedure to guarantee mass conservation.
Bio:
Professor Ken Kamrin received a BS in Engineering Physics and a minor in Mathematics at UC Berkeley in 2003, and a PhD in Applied Mathematics at MIT in 2008. He was an NSF Postdoctoral Research Fellow at Harvard University in the School of Engineering and Applied Sciences before joining the Mechanical Engineering faculty at MIT in 2011, where he was appointed the Class of 1956 Career Development Chair. Professor Kamrin's research focuses on constitutive modeling and computational continuum mechanics for large deformation processes, with interests spanning elastic and plastic solid modeling, viscous and non-Newtonian flows, amorphous solid mechanics, upscaling and continuum homogenization, and analytical methods in mechanics. Professor Kamrin has been awarded fellowships from the Hertz Foundation, US Defense Department, and National Science Foundation. Professor Kamrin received the 2010 Nicholas Metropolis Award from APS for work in computational physics, the NSF CAREER Award in 2012, the 2015 ASME Eshelby Mechanics Award for Young Faculty, the Ruth and Joel Spira Teaching Award from the MIT School of Engineering in 2016, and the 2016 ASME Journal of Applied Mechanics Award.
-------------------------------
Geometry Mediated Drag Reduction Using Wrinkled & Textured Surfaces
Date and Time: Friday, May 19, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Shabnam Raayai-Ardakani, PhD Candidate, Department of Mechanical Engineering, Massachusetts Institute of Technology
Previous investigations have suggested that surfaces textured with riblets can reduce the frictional drag force in high Reynolds number laminar and turbulent flow regimes. Shark skin and synthetic manufactured “shark tape” are widely known as examples of such passive drag reduction mechanisms. Inspired by the ribs on the denticles of fast swimming shark species, riblets of different shapes have been studied under laminar and turbulent flow conditions to understand their drag reduction mechanism and to offer guides for designing optimized low-friction bio-inspired surfaces, often with confounding results that show a net frictional drag increase instead of drag reduction. In this talk, Raayai-Ardakani seeks to identify and understand the different physical mechanisms that contribute to the viscous skin friction on textured and wrinkled substrates. First, Raayai-Ardakani uses a rescaling of the Navier-Stokes equations together with conformal mapping to establish a simplified theory for laminar boundary layer flow over V-groove riblets and explore the self-similarity of the velocity profiles as a function of a newly rescaled form of the local Reynolds number and the aspect ratio of the riblets. Then, Raayai-Ardakani uses numerical simulation of boundary layers over sinusoidal riblet surfaces to investigate the use of wrinkled surfaces as a form of riblet to explore more complex textures than the asymptotic theory can describe. Finally, Raayai-Ardakani presents the results of combined computational and experimental exploration of the effects of textured rotors on torque reduction in steady flow between concentric cylinders, widely known as Taylor-Couette Flow. Using 3D printed periodical-textured rotors and a custom designed and built Taylor-Couette cell which can be mounted on a stress-controlled rheometer, Raayai-Ardakani measures the torque on the inner rotor as a function of the Reynolds number of the flow, the aspect ratio of the riblets and the relative size of the riblets (with respect to the gap of the Taylor-Couette cell). Experimental data showing up to 20% torque reduction are then compared with numerical simulations of steady and unsteady axisymmetric Taylor-Couette flow over textured rotors to highlight how the kinematics near the periodic riblets are changed and give rise to a torque reduction as the aspect ratio and riblet wavelength are progressively varied.
Bio:
Shabnam Raayai-Ardakani is a PhD candidate at the Department of Mechanical Engineering at MIT, working at the Hatsopoulos Microfluidics Laboratory (HML) with Professor Gareth McKinley and is interested in understanding the mechanics of surface textures (especially wrinkled surfaces) and their effect on drag force experienced by the surfaces. She has obtained her Master’s degree in Mechanical Engineering from MIT in 2013 and worked with Professor Mary Boyce on mechanics of graded wrinkling. She received her Bachelor’s degree in Mechanical Engineering from Sharif University of Technology in 2011.
----------------------------
Engineering a Small Particle Heat Exchange Receiver for Concentrated Solar Applications
Date and Time: Friday, April 21, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Professor Fletcher J. Miller, Department of Mechanical Engineering, San Diego State University and also in attendance, Dr. Miller’s Ph.D. Advisor, Dr. Arlon Hunt, former Senior Scientist at Lawrence Berkeley Laboratory
The concept of absorbing concentrated solar radiation volumetrically, rather than on a surface, is being researched by several groups with differing designs for high temperature solar receivers. The Small Particle Heat Exchange Receiver (SPHER), one such design, is a gas-cooled central receiver capable of producing pressurized air in excess of 1000 C designed to be directly integrated into a Brayton-cycle power block to generate electricity from solar thermal power. The unique heat transfer fluid used in the SPHER is a low-density suspension of carbon nanoparticles (diameter ~ 250 nm) to absorb highly concentrated solar radiation directly in a gas stream, rather than on a fixed absorber like a tube or ceramic foam. The nanoparticles are created on-demand by pyrolyzing a small flow of natural gas in an inert carrier gas just upstream of the receiver, and the particle stream is mixed with air prior to injection into the receiver. The receiver features a window (or multiple windows, depending on scale) on one end to allow concentrated sunlight into the receiver where it is absorbed by the gas-particle suspension prior to reaching the receiver walls. As they pass through the receiver the carbon nanoparticles oxidize to CO2 resulting a clear gas stream ready to enter a downstream combustor or directly into the turbine. The amount of natural gas consumed or CO2 produced is miniscule (1-2%) compared to what would be produced if the natural gas were burned directly to power a gas turbine.
The idea of a SPHER, first proposed many years ago, has been tested on a kW scale by two different groups. In the recent work, the engineering challenges to developing a multi-MW SPHER is reported. An in-house Monte Carlo model of the radiation heat transfer in the gas-particle mixture has been developed and is coupled to FLUENT to perform the fluid dynamic calculations in the receiver. Particle properties (size distribution and complex index of refraction) are obtained from angular scattering and extinction measurements of natural gas pyrolysis in a lab-scale generator, and these are corroborated using SEM analysis of captured particles. A numerical model of the particle generator has been created to allow for scale-up for a large receiver. We have also designed a new window for the receiver that will allow pressurized operation up to 10 bar with a 2 m diameter window. Calculations and design of a secondary concentrator and overall thermodynamic modeling of the central receiver plant will be discussed if time allows.
Bio:
Fletcher Miller earned a Ph.D. in Mechanical Engineering from UC Berkeley in 1988 in the area of radiation heat transfer in participating media with applications to solar energy. After two years at the German Aerospace Center continuing research in this area, he worked at the NASA Glenn (formerly Lewis) Research Center as a member of the National Center for Microgravity Research for 16 years. During that time he served as Principal Investigator and technical monitor for a variety of experiments on microgravity combustion and fluids in the drop facilities and on sounding rockets. He also served as Team Lead for the Fire Prevention portion of combustion research destined for International Space Station. In 2007 Dr. Miller joined the faculty of San Diego State University in Mechanical Engineering. He continues to work with NASA on fire safety and microgravity combustion research as a Flight Principal Investigator, has initiated a project with NIST on modeling wind effects on wildfire, and has also been PI on several concentrating solar projects including a rival of SPHER. Also in attendance, Dr. Miller’s Ph.D. Advisor, Dr. Arlon Hunt. Dr. Hunt has a Ph.D. in Physics from the University of Arizona (1974) and spent his professional career as a Senior Scientist at Lawrence Berkeley Laboratory (1976 to 2005). He continues to be active in several areas of science and through companies he founded. Dr. Hunt has over 40 years of experience in solar thermal conversion, light scattering, instrumentation, processing of materials, and electro-optics. He was awarded a Senior Fulbright Fellowship to study industrial applications of solar thermal energy in Africa and received the Technology Transfer Award from Federal Laboratory Consortium and Excellence Award from LBNL. He also won the Energy 100 Award for the development of the Diesel Particle Scatterometer as one of the 100 most significant scientific accomplishments in the history of DOE and initiated and carried out a multinational project to build and test a solar receiver based on small particle absorption.
--------------------------------
Statistical Learning of Kinetic Monte Carlo Models of High Temperature Chemistry from Molecular Dynamics
Date and Time: Friday, April 14, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Qian Yang, PhD Candidate, Institute for Computational and Mathematical Engineering, and Professor Evan Reed, Materials Science and Engineering, Stanford University
Complex chemical processes, such as the decomposition of energetic materials and the chemistry of planetary interiors, are typically studied using large-scale molecular dynamics simulations that run for weeks on high performance parallel machines. These computations may involve thousands of atoms forming hundreds of molecular species and undergoing thousands of reactions. It is natural to wonder whether this wealth of data can be utilized to build more efficient, interpretable, and predictive models. In this talk, we will use techniques from statistical learning to develop a framework for constructing Kinetic Monte Carlo (KMC) models from molecular dynamics data.[1] We will show that our KMC models can not only extrapolate the behavior of the chemical system by as much as an order of magnitude in time, but can also be used to study the dynamics of entirely different chemical trajectories with a high degree of fidelity. Then, we will discuss three different methods for reducing our learned KMC models, including a new and efficient data-driven algorithm using L1-regularization. We demonstrate our framework on a system of high-temperature high-pressure liquid methane, thought to be a major component of gas giant planetary interiors. Finally, we discuss how our L1-regularization based algorithm can also be applied to complex systems of reaction rate equations such as those studied in the combustion community, providing a novel data-driven method for reducing nonlinear dynamical systems.
[1] Q. Yang, C. A. Sing-Long, and E. J. Reed, “L1 Regularization-Based Model Reduction of Complex Chemistry Molecular Dynamics for Statistical Learning of Kinetic Monte Carlo Models,” MRS Advances, vol. 1, no. 24, pp. 1767–1772, 2016.
Bio:
Qian Yang is a Ph.D. candidate in the Institute for Computational and Mathematical Engineering at Stanford University. She received her B.A. in Applied Mathematics from Harvard College in 2009, where she studied machine learning algorithms for modeling human balance and was the lead software developer for a startup developing the technology into a medical diagnostic device for the elderly. Qian's current research focuses on model reduction algorithms for chemical reaction networks. She also works on machine learning methods for accelerated materials discovery. Professor Evan Reed is a faculty member in Materials Science and Engineering at Stanford University. He received a B.S. in applied physics from Caltech in 1998 and PhD in physics from MIT in 2003. In 2004, he was an E. O. Lawrence Fellow and staff scientist at Lawrence Livermore National Laboratory before moving to Stanford in 2010. Reed’s recent work focuses on theory and modeling of 2D materials, statistical learning for chemical and energy storage applications, high pressure shock wave compression, and structural phase changes.
----------------------------------
Fast linear algebra algorithms with applications in computational flow-physics
Date and Time: Friday, April 7, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Hadi Pouransari, Mechanical Engineering and Computer Science, Stanford University
In the realm of scientific computing, solving a linear system of equations is often the main bottleneck of the calculations. We extend ideas from the fast multipole method and propose a novel fast linear solver for sparse and dense matrices. The proposed algorithm is fully algebraic and has numerically proved linear complexity with the problem size. Our method relies on the low-rank compression of the new fill-in blocks generated during the elimination process. The compressed fill-ins are computed and stored in a hierarchical tree structure. The proposed solver can be used as a stand-alone direct solver with tunable accuracy determined a priori, or can be employed as a preconditioner in conjunction with an iterative method.
In addition, we present our high performance computational framework developed to simulate heated particle-laden flows. We present various results on the effects of particle preferential concentration. Simulation of the heated particle-laden flows involves solving a variable coefficient Poisson equation. We use this case, as well as many other applications, to benchmark our proposed fast linear solver.
Bio:
Dr. Hadi Pouransari received his B.S. in Computer Science and Mechanical Engineering from Sharif University of Technology in 2011. He joined Stanford University at the same year and received his M.S. in Mechanical Engineering in 2013. In his Ph.D. studies at Stanford, he worked with Professors Eric Darve and Ali Mani on fast linear algebra algorithms with applications in computational flow-physics. He also obtained a Ph.D. minor in Computer Science from Stanford University.
------------------------------------------
Irreversible dispersion of inertial particles in turbulence
Date and Time: Friday, March 31, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Professor Andy Bragg, Department of Civil and Environmental Engineering, Duke University
The question of how particles suspended in turbulence move relative to each other may be addressed from the point of view of forward-in-time (FIT) and backward-in-time (BIT) dispersion. FIT dispersion is physically related to how groups of particles spread out in turbulence, whereas BIT dispersion is physically related to how particles mix together, and is also important for understanding particle collisions in turbulence. When FIT and BIT dispersion are different it signifies irreversibility, and since FIT and BIT dispersion are related to different problems, understanding the irreversibility is of fundamental and practical importance. I will present a new theoretical analysis, along with results from Direct Numerical Simulations, to show that inertial particle dispersion can be very strongly irreversible in turbulence, and that inertial particles can disperse much faster than fluid (interialess) particles. These results significantly advance our understanding of dispersion problems, and lead to new capabilities for predicting the effect of inertia on the rate at which particles spread out and mix together in turbulence, and the rate at which they collide.
Bio:
Dr. Bragg joined Duke University as an Assistant Professor of Civil and Environmental Engineering in Fall 2016. Prior to this, he was a Postdoctoral Associate in the Applied Mathematics and Plasma Physics Group at the Los Alamos National Laboratory, and before that, a Postdoctoral Associate at Cornell University. He obtained his PhD in Theoretical Fluid Dynamics from Newcastle University, England in 2012. Dr. Bragg’s current research interests include theoretical and computational investigations on particle motion in turbulence, motivated by problems including cloud microphysics, organism mixing in oceans, and astrophysics. Recent interests also include problems in theoretical ecohydrology, porous media flows, and geophysical fluid dynamics.
--------------------------------
Adjoint Sensitivity Analysis for Scale-Resolving Turbulent Flow Solvers
Date and Time: Friday, March 24, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Patrick Joseph Blonigan, Postdoctoral Fellow, NASA Ames Research Center
Adjoint-based sensitivity methods are powerful design tools for engineers who use computational fluid dynamics. In recent years, these engineers have started to use scale-resolving simulations like large-eddy simulations (LES) and direct numerical simulations (DNS), which resolve more scales in complex flows with unsteady separation and jets than Reynolds-averaged Navier-Stokes (RANS) methods. However, the conventional adjoint method computes large, unusable sensitivities for scale-resolving simulations, which unlike RANS simulations exhibit the chaotic dynamics inherent in turbulent flows. Sensitivity analysis based on least-squares shadowing (LSS) avoids the issues encountered by conventional adjoint methods, but has a high computational cost even for relatively small simulations. The following talk discusses a new, more computationally efficient formulation of LSS and its application to turbulent flows simulated with Eddy, a discontinuous-Galkerin spectral-element-method LES/DNS solver. First, the new LSS formulation, called “non-intrusive” LSS, is outlined, followed by a cost analysis of the method. Results are presented for the minimal flow unit, a turbulent channel flow with a limited streamwise and spanwise domain.
Bio:
Dr. Blonigan received his B.S. in Mechanical Engineering from Cornell University in 2011. In June 2011 he started his Ph.D. at Massachusetts Institute of Technology in the Department of Aeronautics and Astronautics, and he worked with Professor Qiqi Wang, developing new approaches for sensitivity analysis of chaotic dynamical systems and fluid flow simulations. Dr. Blonigan received his M.S. in 2013 and his Ph.D. in 2016, and he is now a Postdoctoral Fellow at NASA Ames Research Center where he is applying his Ph.D. research on chaotic sensitivity analysis to scale-resolving turbulent flow solvers.
---------------------------------------
An analytical investigation of particle clustering in turbulence
Date and Time: Friday, March 17, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Mahdi Esmaily-Moghadam, CTR Postdoctoral Fellow, Center for Turbulence Research, Stanford University
Clustering of inertial particles plays a key role in the formation of rain to that of the planets in our early solar system. It has been known for decades that heavy tiny particles in a turbulent flow segregate spatially, forming clusters. The degree of clustering highly depends on the particle relaxation time relative to that of the background flow, viz. the Stokes number. While minimal clustering is observed at very large or small Stokes numbers, clustering is maximized when the two time-scales are comparable and St ≈ 1. This nonmonotonic variation, although observed experimentally and reproduced numerically, is yet to be explained theoretically. At the low Stokes number regime, our theoretical knowledge is sufficient for making a quantitative prediction, whereas, for the rest of parameter space, our knowledge is qualitative at its best. Filling this gap is the objective of this study. In this talk, Dr. Mahdi Esmaily-Moghadam will paint a picture that reveals underlying mechanisms leading to cluster formation by putting together pieces obtained from a lower dimensional analysis. This pen-and-paper analysis shows the fundamental response of particles to straining and rotating flows, reveals the existence of a lower bound on the clustering rate, predicts the requirement for trajectory crossing and the asymptotic variation of the Lyapunov exponents for the full regime of Stokes and Reynolds numbers. Quantitative comparison against direct numerical simulations, including particle-laden isotropic turbulence, confirms these findings.
Bio:
Dr. Mahdi Esmaily-Moghadam received his B.Sc. and M.Sc. with high distinction from Sharif University and a second M.Sc. and Ph.D. in Mechanical Engineering from UCSD in 2014. During his doctoral studies, he developed several numerical techniques for multiscale simulation of the circulatory system and proposed a novel surgical design for single ventricle heart patients. Currently, as a postdoctoral scholar at the Center for Turbulence Research at Stanford University, he is studying the interaction of inertial particles with turbulent flows in the context of solar receivers. His research interests center around the study of emerging applications in cardiovascular mechanics and particle-laden flows and development of multiscale computational techniques for such applications.
--------------------------------------
PyFR: High-Order Accurate Cross-Platform Petascale Computational Fluid Dynamics with Python
Date and Time: Friday, March 10, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Dr. Freddie David Witherden, Postdoctoral Scholar, Department of Aeronautics and Astronautics, Stanford University
High-order numerical methods for unstructured grids combine the superior accuracy of high-order spectral or finite difference methods with the geometrical flexibility of low-order finite volume or finite element schemes. The Flux Reconstruction (FR) approach unifies various high-order schemes for unstructured grids within a single framework. Additionally, the FR approach exhibits a significant degree of element locality, and is thus able to run efficiently on modern many-core hardware platforms, such as graphics processing units (GPUs). The aforementioned properties of FR mean it offers a promising route to performing affordable, and hence industrially relevant, scale-resolving simulations of hitherto intractable unsteady flows within the vicinity of real-world engineering geometries. In this talk we will present PyFR an open-source, Python-based framework for solving advection-diffusion type problems using the FR approach. The framework is designed to target a range of hardware platforms via use of a domain specific language. With this PyFR is able to solve the compressible Euler and Navier-Stokes equations on grids of quadrilateral and triangular elements in two dimensions, and hexahedral, tetrahedral, prismatic, and pyramidal elements in three dimensions, targeting clusters of multi-core CPUs, NVIDIA GPUs, AMD GPUs, Intel Xeon Phis, and heterogeneous mixtures thereof. Results will be presented for various benchmark and "real-world" flow problems, and scalability/performance of PyFR will be demonstrated on clusters with thousands of NVIDIA GPUs. Throughout the talk the importance of algorithm-software-hardware co-design, in the context of next-generation computational fluid dynamics, will be highlighted.
Bio:
Dr. Witherden studied Physics with Theoretical Physics at Imperial College London between 2008–2012 earning an MSci degree with first class honours. In September of 2012 he started a PhD in computational fluid dynamics in the department of Aeronautics at Imperial College London under the supervision of Dr. Peter Vincent and graduated in December 2015. Early in 2016 he started a postdoctoral appointment in the department of Aeronautics and Astronautics at Stanford University under the supervision of Professor Antony Jameson. Dr. Witherden's main research interests are in the development new and novel approaches to enable the simulation of hitherto intractable flow problems at extreme scale.
----------------------------------------
Mass transfer effects of drops in a liquid flow
Date and Time: Friday, March 3, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Shigan Chu, Johns Hopkins University, Department of Mechanical Engineering
Even a nominally immiscible drop will grow or dissolve in an ambient liquid due to diffusion. The mass flux at the drop surface is dictated by the drop composition. For a pure drop, this composition is essentially constant in time. The situation is very different for a multicomponent drop in which, due to the mutual interaction of the constituents, their concentrations at the drop surface differ from the respective solubilities and depends on time. The result is a memory term for the mass flux through the interface, which does not exist for a drop of a pure liquid.
In this talk, Shigan Chu will first determine the time-dependent surface concentrations by assuming thermodynamic equilibrium at the drop surface, and their influence on the drop mass transfer in ternary systems. Then the relevance of the new memory term will be evaluated under several situations. The theory of multicomponent drop dissolution will further be applied to multiphase plumes where dispersed phase such as drops and bubbles dissolve into the liquid as they rise.
Bio:
Shigan Chu is currently a Ph.D. candidate in Mechanical Engineering Department of the Johns Hopkins University. He received M.S. in Mechanical Engineering from the Johns Hopkins University and B.S. in Mechanics & Engineering Science from Peking University. His graduate research focuses on mass transfer in multiphase flow and drops/bubbles laden turbulent buoyant plumes.
---------------------------------------
Prediction of pollutants in Gas Turbines using Large Eddy Simulation
Date and Time: Friday, February 10, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Thomas Jaravel, CTR Postdoctoral Fellow, Stanford University
Stringent regulations of pollutant emissions now apply to new-generation combustion devices. To achieve low nitrogen oxides (NOx) and carbon monoxide (CO) emissions simultaneously, a complex optimisation process is required in the development of new concepts for engines. Already efficient for the prediction of turbulent combustion, Large Eddy Simulation (LES) is also a promising tool to better understand the processes of pollutant formation in gas turbine conditions and to provide their quantitative prediction at the design stage. In this work, a new methodology for the prediction with LES of NOx and CO in realistic industrial configurations is developed. It is based on a new strategy for the description of chemistry, using Analytically Reduced Chemistry (ARC) combined with the Thickened Flame model (TFLES). An ARC with accurate CO and NO prediction is derived, validated on canonical laminar flames and implemented in the LES solver. The accuracy of this approach is demonstrated with a highly resolved simulation of the academic turbulent Sandia flame D, for which excellent prediction of NO and CO is obtained. The methodology is then applied to two industrial configurations. The first one is the SGT-100, a lean partially-premixed gas turbine model combustor studied experimentally at DLR. LES of this configuration highlights the chemical processes of pollutant formation and provides qualitative and quantitative understanding of the impact of the operating conditions. The second target configuration corresponds to a mono-sector prototype of an ultralow NOx, staged multipoint injection aeronautical combustor developed in the framework of the LEMCOTEC European project and studied experimentally at ONERA. An ARC for the combustion of a representative jet fuel surrogate is derived and used in the LES of the combustor with an Eulerian formalism to describe the liquid dispersed phase. Results show the excellent performances of the ARC, for both the flame characteristics and the prediction of pollutants.
Bio:
Dr. Jaravel graduated from Ecole Polytechnique (Paris-Saclay) in mechanical engineering and applied mathematics and from Supaéro (Toulouse) in aerospace engineering. He received his PhD in 2016 from Institut National Polytechnique (Toulouse), conducted with CERFACS and Safran Aircraft Engines. His research focuses on Large Eddy Simulation (LES), chemical kinetics and two phase flow combustion. Dr. Jaravel joined the Center for Turbulence Research on January 16, 2017.
------------------------------------
Date and Time: Friday, January 27, 2017 - 16:30
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker: Dr. Philipp Hack
Visualizations of turbulent boundary layers show an abundance of characteristic arcshaped structures whose apparent similarity suggests a common origin in a coherent dynamic process. While the structures have been likened to the hairpin vortices observed in the late stages of transitional flow, a consistent description of the underlying mechanism has remained elusive. Detailed studies of coherent turbulent processes are complicated by the chaotic nature of the flow which modulates each manifestation and renders the isolation of representative samples a challenging task.
The present study harnesses methods developed in the field of computer vision to identify and analyze turbulent flow processes. The algorithm employs morphological operations to distill the topology of the turbulent flow field into a discrete graph. The low-dimensional structural information is stored in a database and enables the identification and analysis of equivalent dynamical processes across multiple scales. Application of the scheme to direct simulations data allows for the first time the time-resolved sampling of the turbulent hairpin process. The analysis attributes the hairpin process to an inviscid instability mechanism related to the inflection points introduced by turbulent streaks.
Bio:
Dr. Philipp Hack received his Ph.D. from Imperial College London in 2014. His theoretical and computational work on the stability of boundary layers was awarded in 2013 with the EPSRC Doctoral Prize Fellowship and in 2015 with a DFG Research Fellowship. Philipp Hack joined the group of Professor Parviz Moin at the beginning of 2015. His current research interests include flow stability, optimization and machine learning.
------------------------------
Model reduction in fluid mechanics: a short tutorial
Date and Time: Friday, January 13, 2017 - 16:15
Location: CTR Conference Room 103
Event Sponsor: Parviz Moin, Director of Center for Turbulence Research
Speaker(s): Aaron Towne, CTR Postdoctoral Fellow, Stanford University
Reduced-order models that capture the essential features of turbulent flows can be used to gain insight into the flow physics, predict quantities of interest, and provide a handle for flow-control efforts. This talk will compare and contrast a selection of empirical (data-based) and non-empirical (equation-based) model-reduction methods, including several variants of proper orthogonal decomposition, dynamic mode decomposition, resolvent analysis, and balanced truncation. The hope is that this introduction will help listeners interpret the work of others and select appropriate approaches for their own work.
Bio:
Aaron Towne is a postdoctoral fellow in the Center for Turbulence Research. He received his B.S. in Engineering Mechanics from the University of Wisconsin-Madison and his M.S. and Ph.D. in Mechanical Engineering from the California Institute of Technology. His research interests include aeroacoustics, reduced-order modeling, and flow instability.