Skip to main content Skip to secondary navigation

CTR Seminars Archive 2019

Main content start

Interface dynamics: mechanisms of stabilization and destabilization and structure of flow fields

Date and Time: Friday, December 13, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Professor and Chair Snezhana I. Abarzhi, Professor and Chair of Applied Mathematics University of Western Australia

Interfacial mixing and transport are nonequilibrium processes coupling kinetic to macroscopic scales. They occur in plasmas, fluids, and materials over celestial events to atoms. Grasping their fundamentals can advance a broad range of disciplines in science, mathematics, and engineering. This work focuses on the long-standing classic problem of stability of a phase boundary - a fluid interface that has a mass flow across it. We briefly review the recent advances and challenges in theoretical and experimental studies, develop the general theoretical framework directly linking the microscopic interfacial transport to the macroscopic flow fields, discover the new mechanisms of interface stabilization and destabilization for both inertial and accelerated dynamics, and chart perspectives for future research.

The work is supported by the University of Western Australia (AUS) via project grant 10101047, and the National Science Foundation (USA) via award 1404449.

Abarzhi et al 2019 Proc Natl Acad Sci USA 116 (37) 18218. https://doi.org/10.1073/pnas.1714500115.
Ilyin et al 2019 Europhysics Letters 125, 14002. https://doi.org/10.1209/0295-5075/125/14002.
Ilyin et al 2018 Physics of Plasmas 25, 112105. https://doi.org/10.1063/1.5008648.
Ilyin et al 2018 http://arxiv.org/abs/1901.04575.

Speaker Bio: 

Snezhana Abarzhi works at the University of Western Australia as Professor and Chair of Applied Mathematics. Before the University of Western Australia, she worked at Carnegie Mellon University, University of Chicago, Stanford University and State University of New York at Stony Brook in the USA, as well as internationally (Osaka University in Japan, University of Bayreuth in Germany, Landau Institute for Theoretical Physics in Russia). Snezhana Abarzhi's research interests are in Applied Mathematics (applied analysis, partial differential equations, boundary value problems) and in Theoretical Applied and Physics (fluids, plasmas, materials). The focus of Dr. Abarzhi’s research is on fluid instabilities and interfacial mixing. Her contribution to this field is in rigorous physics-based theory for fundamentals of non-equilibrium dynamics, interfaces and mixing. 2019 Special Feature Issue on ‘Interfaces and Mixing’ of the Proc Natl Acad Sci USA https://www.pnas.org/page/about/special-features. http://onlinedigeditions.com/publication/?i=629516#{%22issue_id%22:629516,%22page%22:0}.

-----------------------------------------

A Computationally Efficient “Turnkey” Approach Turbulent Combustion Modeling

Date and Time: Friday, November 15, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Professor Michael E. Mueller, Associate Professor in the Department of Mechanical and Aerospace Engineering at Princeton University

Turbulent combustion is an extremely challenging “multi-multi” problem: multi-physics, multi-scale, and multi-species. Since not all scales of turbulence and combustion can be resolved in DNS for practical conditions of interest, models are required for the unresolved turbulent combustion processes in LES and RANS. However, the large number of thermochemical scalars required to describe combustion chemistry (potentially hundreds or thousands of chemical species) means that the unresolved state-space that needs to be modeled is extremely high-dimensional. Turbulent combustion models can generally be divided into two distinct classes based on how this dimensionality challenge is addressed. In the first class of models, no attempt is made to reduce the dimensionality of the unresolved state-space (“brute-force” models). While very general, these approaches are extremely computationally intensive and realistically impractical. Conversely, in the second class of models, the dimensionality of the unresolved state-space is reduced by a priori presuming that combustion occurs in one of the asymptotic “modes” of nonpremixed combustion, premixed combustion, or homogeneous autoignition, each of which can be described by simple one-dimensional manifold equations. While this results in a substantial reduction in computational cost, these models are not generally applicable to “multi-modal” combustion processes characteristic of practical systems. In this seminar, recent efforts to overcome this fundamental modeling trade-off will be discussed. Our new turbulent combustion modeling framework is both computationally efficient and extremely general, requiring no a priori knowledge about the underlying combustion processes in order to reduce the dimensionality of the unresolved state-space. The new approach relies on two game-changing components: (1) generalized two-dimensional manifold equations capable of describing arbitrary “multi-modal” combustion processes and (2) sensible computational algorithms that shift away from unnecessary precomputation and high-dimensional pretabulation toward ‘just-in-time’ computation and adaptive tabulation. Preliminary application with LES to canonical turbulent flames will be briefly discussed.

Speaker Bio: 

Michael E. Mueller is an Associate Professor in the Department of Mechanical and Aerospace Engineering at Princeton University, an associated faculty member in the Princeton Institute for Computational Science and Engineering, an associated faculty member in the Andlinger Center for Energy and the Environment, and the Director of the Graduate Certificate in Computational Science and Engineering. He received a BS degree in mechanical engineering from The University of Texas at Austin in 2007, a MS degree in mechanical engineering from Stanford University in 2009, and a PhD degree in mechanical engineering from Stanford University in 2012 before moving to Princeton in 2012. In 2017 he was recognized with an award through the Young Investigator Program (YIP) of the Army Research Office (ARO), and he currently serves as Associate Editor for the Journal of Engineering for Gas Turbines and Power. His expertise is the computational modeling of turbulent reacting flows. Current research interests within his group includes multi-modal turbulent combustion, combustion-influenced turbulence, pollutant emissions, and uncertainty quantification.

-----------------------------------

Dynamics of swept shock-wave/turbulent-boundary-layer interactions

Date and Time: Friday, November 1, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Michael Adler, CTR Postdoctoral Fellow

Swept shock-wave/turbulent-boundary-layer interactions (STBLIs) exhibit key dynamical differences from their spanwise-homogeneous counterparts, including the suppression of a dominant mechanism of low-frequency unsteadiness. An extensive database of wall-resolved simulations is constructed to examine the differences between the properties of spanwise-homogeneous, swept, and compound swept interaction classes. A novel method for calculating the dynamic linear response of these unsteady flows is also presented, which allows for the identification and characterization of absolute instabilities in the time-resolved turbulent flow, and the ramifications for STBLI dynamics are discussed. The spanwise homogeneous interaction exhibits an absolute instability that is well-correlated with the prominent band of low-frequency unsteadiness, whereas an absolute instability is not present in the simple swept interaction, which exhibits a muted low-frequency band. Further, a relationship between interaction symmetry, separation topology, and low-frequency unsteadiness is described, in which the prominent low-frequency unsteadiness is spatially associated with surface-flow singular points that function to topologically close the separation. For swept interactions, the ramifications of quasi-conical interaction symmetry on the frequency scaling of the shear layer bands are also discussed. In essence, the shear layers of the swept interactions exhibit a mix between classical and conical free-interaction scaling in both mean flow and frequency content.

Speaker Bio: 

Dr. Michael Adler received his PhD in Aeronautical and Astronautical Engineering from The Ohio State University in 2019, where his research examined many aspects of shock-wave/turbulent-boundary-layer interactions with Prof. Datta Gaitonde. He has recently joined CTR as a postdoctoral fellow and works with Prof. Sanjiva Lele on numerical methods for shock/material-interface interactions and elastic-plastic flow phenomena, including the Richtmyer-Meshkov and other instabilities of elastic-plastic flows.

------------------------------

Attached eddies in turbulent channel flow: scale interactions and spectral energy transfer

Date and Time: Friday, October 25, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Hyun Ji (Jane) Bae, Postdoctoral Scholar at the Department of Aerospace at Caltech

The resolvent formulation of McKeon & Sharma (2010) is applied to supersonic turbulent boundary layer flows in order to study the validity of Morkovin’s hypothesis, which postulates that high-speed turbulence structure in zero pressure-gradient turbulent boundary layers remains largely the same as its incompressible counterpart. The resolvent analysis highlights two distinct regions of the supersonic turbulent boundary layer in the wave parameter space: the relatively supersonic region and the relatively subsonic region. The relatively supersonic region, where the flow is supersonic relative to the freestream, contains resolvent modes that display structures consistent with the eddy Mach wave radiation that are absent in the incompressible regime. In the relatively subsonic region, the low-rank approximation of the resolvent operator is effective and the model exhibits a universal and geometrically self-similar behavior via a transformation given by the semi-local scaling. Moreover, with the semi-local scaling, the resolvent modes follow the same scaling law as its incompressible counterparts in this region, which has implications for modeling and the prediction of turbulent high-speed wall-bounded flows.

Speaker Bio: 

Dr. Hyun Ji Bae is a postdoctoral scholar at the Department of Aerospace at Caltech. She received her Ph.D. from Stanford University in 2018 in Computational and Mathematical Engineering. Her main research focuses on computational fluid mechanics, in particular on modeling, understanding, and control of wall-bounded turbulence using reduced order modeling.

--------------------------------

Scaling laws for compressible wall turbulence

Date and Time: Tuesday, September 3, 2019 - 10:00

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Andrew Trettel

A recent mean velocity transformation (Trettel and Larsson 2016) converts compressible mean velocities into equivalent incompressible mean velocities.  This transformation works properly for channel flows but does not work properly for boundary layers.  A two-dimensional extension of this transformation reveals the inherent limitations in the transformation as an extension of the incompressible law-of-the-wall.  The transformation cannot properly transform the outer layer coordinate in boundary layers, and the error associated with this correlates highly with the error in the transformation itself.  Despite these limitations, the two-dimensional extension reveals additional scaling laws for compressible wall turbulence, including scalings for the streamwise coordinate and production rates.

Speaker Bio: 

Dr. Andrew Trettel received his PhD in mechanical engineering from the University of California, Los Angeles in 2019. Trettel’s recent research has been in developing and improving scaling laws for compressible wall turbulence.

-----------------------

Stable and Accurate Filtering Procedures

Date and Time: Friday, August 23, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Professor Jan Nordström

The study of so called transmission problems in  [2] revealed that successful numerical filtering may include a delicate balance between the need to remove high frequency oscillations (filter often for accuracy) and the need to avoid possible growth  (filter seldom for stability).  In this talk we investigate this contradiction, and propose different avenues for improved functionality.

The filter operators derived in [1] are the basic building  blocks.  We demonstrate that explicit use of the basic filter operators guarantee accuracy but lead to instabilities, while an implicit implementation preserve stability but degrade accuracy.  We also prove that a specific accuracy condition is necessary for stability, and that the basic operators do not satisfy that condition.

Finally, new modified filter operators that satisfy the specific accuracy conditions, are stable in combination with summation-by-parts operators [3], and can be used both explicitly and implicitly are constructed. The new operators are shown to efficiently damp the highest frequencies on the grid, including the π-mode.
References

[1]  Christopher A Kennedy and Mark H Carpenter. “Comparison of several numerical methods for simulation of compressible shear layers.” In: NASA Technical Paper 3484 (1997).
[2]  Jan Nordström and Viktor Linders. “Well-posed and stable transmission problems.” In: Journal of Computational Physics 364 (2018), pp. 95–110.
[3]  Magnus Svärd and Jan Nordström. “Review of summation-by-parts schemes for initial–boundary- value problems.” In: Journal of Computational Physics 268 (2014), pp. 17–38.

Speaker Bio: 

Since 2010 Dr. Jan Nordström is a Professor in Scientific Computing and since 2012 he is the Head of Division of Computational Mathematics, in the Department of Mathematics, at Linköping University (LiU) in Sweden. In his research they develop numerical methods for multi-physics applications governed by partial differential equations. They consider various kinds of uncertainties in the data or parameters of the problem and aim for a computational methodology that delivers an answer with error bars. Professor Nordström is interested in initial boundary value problems, and in particular the fundamental effect of boundary and interface conditions on well-posedness and stability. The applications include flow, wave, heat and uncertainty propagation.

------------------------------

Fractal dimension of transitional boundary layer spot interfaces detected by a self- organizing map

Date and Time: Thursday, August 22, 2019 - 10:00

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Zhao Wu

An unsupervised machine-learning algorithm, the self-organizing map (SOM), is used to identify the turbulent boundary layer (TBL) and non-TBL regions in bypass transition. The data employed for the analysis are from an archived direct simulation publicly available in the Johns Hopkins Turbulence Databases (JHTDB, http://turbulence.pha.jhu.edu). The data points in the entire flow domain are automatically classified into TBL and non-TBL regions by the SOM, based on their standardized velocity, velocity fluctuations, velocity gradients and their spatial locations. Thus the SOM identifies the turbulent-boundary-layer interface (TBLI) without the usual need for choosing thresholds on e.g. vorticity or velocity fluctuations. The TBLI is found to be a hyperplane in the input space. The SOM distinguishes the streaks in the laminar region and the weak free-stream turbulence from TBL region. Results from our approach are shown to be consistent with threshold-based methods in the special cases when those are applicable.

This approach is then used to study the turbulent spots. The nature of turbulent spots in transitional boundary layers, and whether their internal structure shares characteristics of equilibrium turbulence, remain open questions of considerable interest. Here we study scaling properties of the interface separating the spots from the outside flow. For high-Reynolds-number turbulence, such interfaces are known to display fractal scaling with a fractal dimension near D=2+1/3, where the 1/3 can be related to the Kolmogorov scaling of velocity fluctuations (e.g. de Silva et al. PRL 2013). We measure the volume-area fractal scaling of the naturally triggered turbulent spots. Results from the volume-area fractal dimension confirm D=7/3, i.e. trends consistent with fully developed turbulence. Applying an alternative area-perimeter analysis on planar cuts at various heights shows D decreasing then increasing. It is argued that these trends could be associated to changes in the thickness of the interface at different heights from the wall.

Speaker Bio: 

Dr. Zhao Wu received his doctoral degree in Mechanical Engineering from The University of Manchester, UK in 2017. During his PhD, his research was focused on direct numerical simulations of fluid flow and conjugate heat transfer. He then joined Johns Hopkins University as a postdoctoral researcher working on the Johns Hopkins Turbulence Databases (JHTDB). His research interests include machine learning, data compression, high-performance computing and computational fluid dynamics.

--------------------------------

Hydrodynamic stability of a premixed flame subjected to transverse shear

Date and Time: Wednesday, August 21, 2019 - 10:00

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Mr. Xiaoyi Lu, Ph.D. Candidate

Lu presents findings on the hydrodynamic stability of a premixed flame subjected to transverse shear. The problem configuration is a situation of interest for laminar and turbulent flames when they travel into a region of shear.

The linear stability problem is first analytically solved, and the dispersion relation is determined. The effects of the transverse shear and thermal expansion are examined. Lu’s research then carried out a weakly nonlinear analysis in the weak thermal expansion limit and derived the modified Michelson-Sivashinsky (MS) equation, which describes the evolution of the flame surface. Numerical solutions of the MS equation show that due to the transverse shear, the flame develops a skewed cusp-like structure, that steadily propagates into the unburned gases and simultaneously translates along the transverse direction. The fully nonlinear evolution of premixed flames with a realistic density jump is then investigated using the Direct Numerical Simulation (DNS) approach.

Speaker Bio: 

Mr. Xiaoyi Lu is a Ph.D. Candidate in Theoretical and Applied Mechanics at the University of Illinois at Urbana-Champaign. He received his B.S. degree in Mechanical Engineering from the University of Texas at Austin.

--------------------------------

Data-assisted Reduced-order Modeling of High-dimensional and Fully Turbulent Flows: Forecast and System Identification

Date and Time: Friday, August 16, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Mohammad Amin Khodkar

One-Dimensional (1D) Reduced-Order Models (ROMs) are developed for a 3D highly turbulent Rayleigh Bènard convection (RBC) system with Rayleigh number Ra=106, which is nearly 600 times larger than its critical value, while three general objectives are pursued: 1) Predicting the time-mean response to external forcings, 2) Identification of underlying dynamics, and 3) Short-term prediction of spatiotemporal evolution of the flow. Towards the first aim, Linear Response Function (LRF) of the system is obtained via Green’s Function (GRF) method, which is an equation-dependent method for system identification involving applying many localized, weak forcings to the system and calculating its long time-mean response from direct numerical simulations. An alternative model-free approach based on using the Fluctuation-Dissipation Theorem (FDT) in the subspace of approximated Koopman modes is also employed to calculate LRF. The predictions of both methods for the changes in the flow variables in response to external forcings are compared to DNS results. Using Koopman modes as the basis functions, rather than the commonly used Proper Orthogonal Decomposition (POD) modes, resolves a previously identified problem in applying FDT to systems with non-normality. We furthermore present a Koopman-based framework which relies on delay-embedding of vector-valued observables and treats the nonlinearities as exogenous forcings to predict the spatiotemporal evolution of high-dimensional and chaotic dynamical systems. The excellent performance of this model is demonstrated for some well-known prototypes of chaotic dynamics and a fluid example.

Bio: 

Dr. Amin Khodkar is a Postdoctoral Research Associate at the Environmental Fluid Dynamics Group of Mechanical Engineering Department of Rice University. He received his B.Sc. and Ph.D. in Mechanical Engineering from the University of Tehran and University of California at Santa Barbara in 2013 and 2017, respectively. His research focuses on developing semi-analytical and numerical models for the stratified turbulent flows of high complexity and dimensions, with emphasis on data-driven reduced-order modeling of geophysical turbulence.

--------------------------------

Machine Learning for Multiphase Flows

Date and Time: Tuesday, August 6, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Professor Sivaramakrishnan Balachandar

No Abstract.

Bio: 

http://www.mae.ufl.edu/people/balachandar

----------------------------------

High-Fidelity Simulation, Active Flow Control and Reduced Order Modeling of a Plunging Airfoil under Deep Dynamic Stall

Date and Time: Friday, August 2, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Professor William R. Wolf, Founder and PI of the Laboratory of Aeronautical Sciences in the School of Mechanical Engineering at University of Campinas, Brazil.

Unsteady flows over plunging and pitching airfoils with large excursions in effective angle of attack exhibit the phenomenon of dynamic stall. This process is characterized by unsteady separation and formation of a large leading-edge vortex that exerts high amplitude fluctuations in aerodynamic loads. Although several studies have been conducted for pitching airfoils at high Reynolds numbers, research on dynamic stall for plunging airfoils is more scarce, especially at low and moderate Reynolds numbers.

We employ large eddy simulations to study the flow physics of deep dynamic stall over a plunging SD7003 airfoil at Reynolds number Re=60,000. The current study of an airfoil in plunging motion finds application in design and operation of small unmanned air vehicles and micro air vehicles. An assessment of different arrangements of flow actuators is also presented aiming to reduce the overall drag through modification of the dynamic stall vortex mechanism. Finally, reduced order modeling strategies based on deep neural networks are discussed for turbulent flows involving dynamic stall.

Speaker Bio: 

William R. Wolf is the founder and PI of the Laboratory of Aeronautical Sciences in the School of Mechanical Engineering at University of Campinas, Brazil. He received his BSc in Mechanical Engineering from University of Sao Paulo, Brazil, in 2003 and his PhD in Aeronautics and Astronautics from Stanford University, in 2011. Since 2013, he has been a faculty in the Department of Energy at University of Campinas and a CNPq research fellow since 2016. His group develops research in computational aeroacoustics and aerodynamics focusing on both fundamental research and technological applications. He received a FAPESP Young Investigator award and has coordinated several projects funded by the US Army Research Office, Boeing and Embraer.

-------------------------------

Multi-scale modeling and simulation of cloud cavitation with application to medical ultrasound

Date and Time: Tuesday, June 11, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Kazuki Maeda

Cavitation and bubble dynamics are of critical importance in various industrial systems. In ultrasound therapy, cloud cavitation that is nucleated in the human body is crucial for treatment outcomes; bubble clouds can coherently collapse to cause injury as well as scatter the wave energy to shield therapy targets. In this talk, I will present a method for multi-scale modeling and simulation of intense cloud cavitation in a compressible liquid, and its application for medical ultrasound. The method employs an Eulerian-Lagrangian approach to solve mixture- averaged equations of motion, in that the continuous phase is discretized on an Eulerian grid, while the gas phase is modeled as spherical, Lagrangian point-bubbles at the sub-grid scale. A stochastic-based closure model is introduced to accurately capture the effect of subgrid-scale pressure fluctuations that are induced by the bubbles. The simulation is used to identify scaling laws that dictate the coherent dynamics of bubble clouds and the cloud-induced, far-field acoustic waves. I will also discuss recent efforts in development of a system for real-time acoustic monitoring and feedback control of cavitation toward optimization of the therapy treatment.

Speaker Bio: 

Dr. Kazuki Maeda is an Acting Assistant Professor of Mechanical Engineering at the University of Washington. He received his BS degree in Mechanical Engineering from the University of Tokyo in 2013, and PhD degree in Mechanical Engineering from California Institute of Technology in 2018 with the Richard Bruce Chapman Memorial Award. Kazuki also received the Overseas Scholarship Award from the Funai Foundation for Information Technology from 2013 to 2018. Kazuki currently serves as a sub-contract PI on NIH’s Program Project Grant that aims to improve outcomes of nephrolithiasis. Kazuki’s research interests are in simulation, experiments, and modeling of multiphase flows and their engineering applications.

---------------------------------------

Towards sensitivity analysis on turbulence via shadowing methods

Date and Time: Friday, May 24, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Mr. Angxiu Ni

Turbulence is an important example of chaotic dynamical systems, and sensitivity analysis is a powerful tool for design via computation.  However, traditional sensitivity methods explode for chaotic systems, posing the challenge we will address in this talk.  We first review how to compute meaningful derivatives of long-time-averaged objectives in chaotic systems via the shadowing method, which we then reformulate as a 'non-intrusive' minimization problem on the unstable subspace. Then we show a recent adjoint shadowing theorem, based on which we develop an adjoint sensitivity algorithm, NILSAS, whose implementation requires only minor modifications to existing adjoint solvers.   Finally, we show an application of NILSAS on a chaotic flow over a three-dimensional cylinder at Re=1100, with cost similar to simulation.

Speaker Bio: 

Mr. Angxiu Ni (pronounces as ang-shyou-knee) got his BS in Mechanical Engineering from Tsinghua University. He then got two MS, one at Tsinghua University working with Professor Haixin Chen, one at Massachusetts Institute of Technology working with Professor Qiqi Wang. Since 2017, Angxiu's is a PhD Candidate at the Department of Mathematics of UC Berkeley, supervised by Professor John Strain. Angxiu's research interest is numerical methods for dynamical systems. His main works are the non-intrusive formulation for shadowing methods and an adjoint shadowing theorem.

------------------------------------

A Novel Diffuse Interface Method for Two-phase Flows

Date and Time: Friday, May 3, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Mr. Seyedshahabaddin (Shahab) Mirjalili

From oceanic breaking waves to atomization of liquid fuels for combustion, two-phase flows are omni-present in natural and industrial settings. Despite decades of numerical method development, due to the many challenges involved in simulation of realistic two-phase flows, no gold-standard has yet emerged in the literature. In this study, we present a novel diffuse interface method that addresses various challenges for simulation of incompressible, immiscible two-phase flows. The boundedness of this mass-conserving interface-capturing method is proven analytically. Then, a comparison of the fully coupled solver with a state-of-the-art VoF solver is provided, demonstrating the cost-efficiency and advantages of our approach. In order to handle turbulent, high density ratio two-phase flows, we show how the momentum transport equation must be modified to achieve consistency with mass transport and conservation of momentum and kinetic energy. Finally, a robust method for representation of surface tension forces that utilizes discrete surface energy definition is presented.

Bio: 

Shahab Mirjalili defended his PhD in Mechanical Engineering at Stanford University in March 2019. During his PhD under the supervision of Professor Ali Mani, he focused on development of numerical methods for simulation of two-phase flows and application to studying micro-bubble generation. Shahab Mirjalili received his MS from Stanford University, and BS from Sharif University of Technology, both in Mechanical Engineering.

-----------------------------------

Big data vs big computation

Date and Time: Friday, April 26, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): ICME / CTR Visiting Associate Professor Qiqi Wang

What is the future of computing? Some believe it's big data. For others, it's big computation. Supporters of big data believe that most problems can be solved by gathering huge amounts of data and applying machine learning. Those who believe in big computation, however, postulates that all phenomena in the world can be explained by solving simple physical equations with sufficient computational power. Who is right? In this talk, I will give my personal view on this question.

Bio: 

Qiqi Wang (pronounces as Chi-chi Wong) got his BS in mathematics from University of Science and Technology at Hefei, China. He then got his Ph.D. in Computational and Mathematical Engineering at Stanford, under the supervision of Parviz Moin and Gianluca Iaccarino. From 2009 to 2015, Qiqi was an assistant professor of Aeronautics and Astronautics at MIT. Since 2015, he has been an associate professor in the same department. He got tenure in 2018 and is currently a visiting professor at Stanford University.

------------------------------------

DNS and WMLES for shock-induced aerodynamic heating in hypersonic boundary layers at Mach 6

Date and Time: Friday, April 5, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr.-Ing. Lin Fu

Shock-wave/boundary-layer interactions are typical flow phenomena in hypersonic flows. In this study, large-scale DNS and equilibrium wall-modeled LES (WMLES) are employed to investigate and predict the physical processes responsible for the intense overheating downstream from the shock-impingement region, where the boundary layer suddenly transitions to turbulence.  (1) At low shock incidence angles, the post-shock transition strongly depends on the inlet disturbances, e.g. the amplitude and the random phase. WMLES has difficulties to replicate the DNS statistics due to the fact that coarse meshes cannot support the propagation of inlet disturbance along the laminar portion of the boundary layer. A critical incidence angle is predicted above which the boundary layer transitions without inlet disturbances. For these high incidence angles, the streamwise streaks are generated after the shock-induced separation bubble and the transition occurs when the streaks break up. WMLES under-predicts the separation bubble size and the discrepancy increases significantly with a larger incidence angle. For very high incidence angles, however, the Stanton number and skin friction coefficient can be predicted well by WMLES.         (2) For the compressible turbulence regions, the classical scaling laws and hypotheses, e.g. the weak Reynolds analogy, the mean temperature-velocity relation, the strong Reynolds analogy, the Morkovin’s hypothesis on turbulence intensities and the velocity scaling laws, are revisited by our DNS data. The performances of WMLES in capturing these turbulence statistics are also examined. It is found that the established transformations fail to collapse the velocity profiles to the log-law assumption for cooling wall conditions while other hypotheses are still valid.

Bio: 

Dr.-Ing. Lin Fu is a Postdoctoral Fellow of CTR at Stanford University since January 2018. He is now working on hypersonic flows in Prof. Moin’s group. Before he joined CTR, he obtained his Ph.D. degree with a grade of Summa Cum Laude in Technical University of Munich, Germany. He is interested in wall-modeled large-eddy simulations of high-speed flows, and high-order numerical methods for complex flows.

--------------------------------------

Acoustic boundary conditions for high-fidelity simulations of combustion instabilities

Date and Time: Friday, March 15, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Quentin Douasbin

Combustion devices are prone to combustion instabilities (CIs). They classically occur when heat release rate oscillations and acoustic fluctuations become coherent, creating constructive interferences. To predict the occurrence of CIs in combustion devices, one must account for 1) the acoustic waves in the entire combustion system and 2) the flame response to these fluctuations. The former can only be obtained if the acoustic properties of the boundaries are properly prescribed. During this seminar, a Time Domain Impedance Boundary Condition (TDIBC) method will be presented as well as a modeling strategy allowing accounting for truncated portions of the physical domain.

Bio: 

Before joining the CTR, Dr. Quentin Douasbin was a Postdoctoral Fellow at CERFACS, Toulouse, France, where he worked on the development of CFD software for industrial partners. He received his Ph.D. from the Institut de Mécanique des Fluides de Toulouse (IMFT), France, under the supervision of Prof. Thierry Poinsot and Dr. Laurent Selle. He also obtained a degree in Mechanical Engineering from INSA Rouen, France, as well as an MSc in Computational Fluid Dynamics from Cranfield University, UK. His research interests are thermoacoustics, supercritical combustion, time-domain impedance boundary conditions and modal decomposition techniques for acoustic fields’ analysis and reconstruction.

-----------------------------------

On the effects of electric fields on counterflow diffusion flames

Date and Time: Friday, March 1, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Mario Di Renzo

The impingement of electric fields on flames is known to have potential for mitigating combustion instabilities, enhancing flame propagation, and decreasing pollutant emissions. In particular, electric fields can be used to steer ions produced inside the flame in opposite directions depending on their sign. These ions collide and exchange momentum with the surrounding neutral molecules, which leads to modifications in the flow field. In this talk, steady axisymmetric numerical simulations of methane/air counterflow laminar diffusion flames are employed in order to analyze the effects of the incident electric field on the flow field. The simulations include detailed chemistry and multicomponent transport, and extend classic second-order methods to tackle the stiffness associated with the transport of charged species. The results indicate a strong coupling between the electric and aerothermochemical fields that nonetheless has significant consequences on the self-similar character of the flow.

Bio: 

Before joining the CTR as a Postdoctoral fellow, Dr. Di Renzo was a researcher at Politecnico di Bari, Italy, where he received his BS and MS in Mechanical Engineering as well as his Ph.D. under the supervision of Prof. G. Pascazio and Prof. M. D. de Tullio. He also received an MSc from the Department of Power and Propulsion of Cranfield University. His research interests are hypersonic and supersonic flows, reacting flows, particle-laden flows and their interaction with electric fields.

-----------------------------------------

Information transfer in energy-eddies of different sizes and why this is useful

Date and Time: Friday, February 15, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Adrián Lozano-Durán

Turbulent flows in the presence of bounding surfaces, as those occurring in oceanic and atmospheric currents, around vehicles, or inside pipes, may be apprehended as a collection of whirls or eddies. These eddies follow a regeneration cycle, i.e, existing eddies are seeds for the origin of new ones and so forth.  Understanding this process is critical for the modeling and control of geophysical and industrial flows where a non-negligible fraction of the energy is dissipated by turbulence in the immediate vicinity of walls. In the present work, we examine the causal interactions among energy-containing eddies in wall-bounded turbulence by measuring how the knowledge of past states of the eddies reduces the uncertainty of future states. Our approach unveils, in a simple manner, that causality of energy-eddies at a given scale is essentially universal and independent of the eddy-size. This observation is accompanied by striking implications for control and modeling of turbulent flows. We show, using neural networks as an example, that novel eddy prediction techniques can be devised for the computationally more affordable smaller eddies, while still being a faithful approach applicable to larger eddies which are intractable even with the current state-of-the-art supercomputers.

Bio: 

Dr. Adrian Lozano-Duran received his PhD from the Technical University of Madrid in 2015 at the Computational Fluid Mechanics Lab. headed by Prof. Jiménez. His main research has focused on Computational Fluid Mechanics and the fundamental physics of wall-bounded turbulence. Currently, he is a Postdoctoral Fellow at the Center for Turbulence Research at Stanford University working on the fundamentals of wall turbulence, large-eddy simulation, wall modeling, and reduced-order-modeling.

--------------------------------------

A fast pressure-correction method for incompressible flows over curved walls

Date and Time: Friday, January 25, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Visiting Associate Professor Antonino Ferrante

We have developed a new pressure-correction method for simulating incompressible flows over curved walls. The methodology is applicable to DNS, LES and RANS. We chose the orthogonal formulation of the Navier-Stokes equations in curvilinear coordinates vs the generalized curvilinear coordinates because the computational cost of advancing the numerical solution of the governing equations in time is substantially reduced as the orthogonal formulation does not contain cross-derivatives in the advection, diffusion, Laplacian, and gradient operators. As a result, the numerical stencils of the finite difference approximations to these operators mirror that of the Cartesian formulation. This also allows us to develop an FFT-based Poisson solver for pressure, called FastPoc, for the cases where the grid is uniform in one spatial direction: surfaces of revolution (e.g., axisymmetric ramps) and surfaces of linear translation (e.g., curved ramps, and curved bumps). Further, we have developed an explicit, three-stage, third-order accurate Runge-Kutta based projection method to advance the velocity field in time which requires solving the Poisson equation for pressure only once per time step. Thus, given that the computational mesh satisfies the property of orthogonality, our numerical method can simulate flows over curved wall surfaces with second-order accuracy in space and third-order accuracy in time. The new FFT-based Poisson solver, FastPoc, is thirty to sixty times faster than multigrid (depending on the tolerance set for multigrid), and the new flow solver, called FastRK3, is overall four to seven times faster when we used FastPoc instead of multigrid. Verification and validation test-cases and applications of separated flows will be presented.

Bio: 

Professor Antonino Ferrante is an Associate Professor of the William E. Boeing Department of Aeronautics & Astronautics at the University of Washington (UW), Seattle, where he leads the Computational Fluid Mechanics Lab. In 1996, he received the B.S. in Aeronautical Engineering from Universita’ di Napoli, Federico II (Italy). In 1997, he received the M.S. in Aeronautics & Astronautics and the Belgian Government Prize from the von Karman Institute. In 2004, he received the Ph.D. in Mechanical and Aerospace Engineering from the University of California, Irvine, where he continued his research as Postdoctoral Scholar until 2007. From 2007 to 2009, he was Postdoctoral Scholar at the Graduate Aeronautical Laboratories of the California Institute of Technology (GALCIT). In 2009, he joined the UW as Assistant Professor where was tenured in 2015. Ferrante is recipient of the NSF CAREER Award (2011). Ferrante’s research is mainly focused on understanding the physical mechanisms and modeling of complex flows, e.g. multiphase and wall-bounded turbulent flows, by developing new computational methodologies for DNS and LES apt to performing simulations on high-performance supercomputers. 

-------------------------------------

Computational modeling of space re-entry aerothermodynamics and magnetized plasmas with COOLFluiD

Date and Time: Friday, January 18, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Andrea Lani

The COOLFluiD (Computational Object Oriented Libraries for Fluid Dynamics) project, which started in 2002 as a joint effort between the Von Karman Institute for Fluid Dynamics (VKI) and the Center for mathematical Plasma Astrophysics (CmPA) at the University of Leuven (KUL), has led to the development of a world-class open source platform for HPC and multi-physics modelling. Within this framework, research efforts have been devoted particularly to the modeling of space re-entry aerothermodynamics and magnetized plasmas. The former includes the development of models/algorithms for simulating chemically reacting hypersonic flows and radiation in re-entry flight conditions and experiments in plasma wind tunnels. The latter focuses on models/algorithms for characterizing space weather phenomena such as solar wind/Earth magnetosphere interactions (by means of ideal magnetohydrodynamics) and a fundamental ubiquitous process known as magnetic reconnection (by means of new generation multi-fluid/Maxwell models). This seminar will present computational modeling aspects and a gallery of numerical results for both the above mentioned applications, while also providing a concise overview of CmPA activities, COOLFluiD and the Leuven Computational Modeling Center (LCMC).

Bio: 

Dr. Andrea Lani is currently Research Expert at the Center for Mathematical Plasma Astrophysics (CmPA) at KU Leuven and Director of the Leuven Computational Modeling Center (LCMC). Previously, he has been Senior Research Engineer in the Aeronautics and Aerospace Department at the Von Karman Institute (VKI) in Belgium and Postdoctoral Fellow at the NASA Ames / Stanford CTR. He received his M.Sc. in Aerospace Engineering from the Polytechnic of Turin (Italy) and his Ph.D. in Engineering Sciences from the University of Brussels (ULB). He has also served as technical team member for the NATO Science and Technology Organization (STO). Dr. Lani is main developer and project leader of COOLFluiD, an open source simulation platform whose development has involved 100+ international contributors so far. His research interests include numerical algorithms, modelling of space re-entry aerothermodynamics, plasma flows and HPC.

---------------------------------------

Input-output analysis of high-speed turbulent jet noise

Date and Time: Friday, January 11, 2019 - 16:30

Location: CTR Conference Room 103

Event Sponsor: Parviz Moin, Director of Center for Turbulence Research

Speaker(s): Dr. Jinah Jeun

In this talk we investigate amplifying behavior of small perturbations about Reynolds-averaged Navier-Stokes solutions of high-speed turbulent jets using input-output (I/O) analysis.  Inspired by control theory in electrical engineering, our method investigates how input forcing (jet turbulence) leads to output noise.  Singular value decomposition of the resolvent of the linearized Navier-Stokes equations forms an orthonormal set of I/O mode pairs, sorted in descending order by the magnitude of the corresponding singular values.  In this way we find that the input modes capture coherent structures in the near-field of turbulent jets.  The optimal input modes correspond to wavepackets represented by asymmetric pseudo-Gaussian envelope functions at a given forcing frequency.  These wavepackets remain similar in shape over a range of frequencies for St > 0.5.  While the optimal mode is a wavepacket, sub-optimal modes represent decoherence of the optimal input mode.  Furthermore, by projecting high-fidelity large eddy simulation (LES) data onto the basis of input modes, we find that input modes do indeed capture the acoustically relevant dynamics in the jet.  The far-field acoustics predicted by the LES are recovered using only a few number of I/O modes.

Bio: 

Dr. Jinah Jeun was a Postdoctoral Associate in the Department of Aerospace Engineering and Mechanics at the University of Minnesota. She received her BS degree in Aerospace Engineering from Korea Advanced Institute of Science and Technology; MS degree in Mechanical Engineering from the University of California, Los Angeles; and PhD degree in Aerospace Engineering and Mechanics from the University of Minnesota in 2018 under the supervision of Professor Joseph W. Nichols. She was the recipient of the Amelia Earhart Fellowship in 2016. Jeun’s research interests are in the areas of computational aeroacoustics, flow stability, and reduced-order modeling.