Thursday, October 19
Joseph Katz - Johns Hopkins University
On the Interaction of Turbulent Boundary Layers with Compliant Surfaces
The presentation has two parts, starting with a little bit of history of the fluids program at JHU over the last 35 years, including collaborations with Charles Meneveau in studies evaluating subgrid stress models for LES. The second part describes recent experimental studies of the interactions of a compliant wall with a turbulent boundary layer. These experiments involve simultaneous measurements of the time-resolved, three-dimensional flow field and the two-dimensional surface deformation at friction Reynolds numbers varying between 2,300 to 9,000. The optical setup integrates high speed tomographic PIV/PTV for measuring the flow with Mach-Zehnder interferometry for mapping the deformation. The time-resolved 3D pressure field is calculated by spatially integrating the material acceleration. The process involves interpolation of the unstructured PTV data to a regular grid using a constraint-cost minimization code that forces the velocity field to be divergence free and the material acceleration, curl free. Integration of the acceleration is performed using a virtual boundary, parallel-line, omni-directional algorithm. Early experiments involved stiff compliant walls with deformation amplitudes too small to affect the flow, resulting in one-way coupling between the flow and deformation. Subsequent studies have been performed using softer material, resulting in two-way coupling, including momentum deficit in the inner part of the boundary layer, and a substantial increase in turbulence level. Combining data obtained from several references, trends of the deformation amplitude scaled by the compliant wall thickness collapse when plotted vs. pressure fluctuations scaled by the compliant material shear modulus. The deformation wave speed varies between 53% to 80% of the free stream velocity (53% in recent data), and the preferred wavelength is about three times the wall thickness, the latter being consistent with theoretical models. Conditional averaging and correlations reveal the characteristic 3D flow structure that affect the wall deformation, e.g., a spanwise vortex with a laterally inward flow above a surface bump inducing a sweeping diverging flow above a dimple located downstream of it. Adopting insight derived from atmospheric wind-wave interactions, the pressure-deformation correlation peak at the ‘critical layer’, where the mean flow speed is equal to the surface wave speed. The critical layer is located within the log layer, increasing in elevation with increasing Reynolds number. For the entire region below this layer, wavenumber-frequency spectra of pressure and vertical velocity fluctuations indicate that the turbulence is phase locked and travel with the deformation even for deformation amplitudes much smaller than a wall unit. In contrast, above the critical layer, the turbulence is advected at the local mean streamwise velocity, and its coherence with the deformation decays rapidly. These findings indicate that the height of the zone dominated by flow-deformation interactions is determined by the surface wave speed.
_____________________________________________________________________________________________
Colm-cille Caulfield - University of Cambridge
Shaken by Physics or Stirred by Machine: Modelling Stratified Mixing (with a Twist)
Richard Feynman acknowledged that ‘turbulence is the most important unsolved problem of classical physics’, and it is always important to remember that he was referring to the simplest case of a fluid of constant density. An even more challenging class of problems arise when the turbulent fluid has a variable density, as turbulent mixing can then convert injected kinetic energy into both viscous dissipation and potential energy. Of course, the earth's oceans are just such variable-density stratified fluids, and the larger scale effect of such stratified turbulence is one of the key areas of uncertainty in modelling the global climate system. As human activity strongly perturbs that system's boundary conditions, it is critical to understand better how stratified turbulence is born, lives and dies within the world's oceans. Fortunately, enormous advances in data availability from both observation and numerical simulation have led to breakthroughs in our fundamental understanding of turbulence in stratified fluids. In this talk I briefly review some of these recent breakthroughs made by my collaborators, where the use of physics-informed, data-driven techniques have proved (somewhat surprisingly) useful.
_____________________________________________________________________________________________
Luciano Castillo - Purdue University
Potential of Wind Turbines on the Altercation of Carbon Dioxide Concentration
Anthropogenic carbon dioxide (CO2) is a major contributor to global warming, necessitating significant reductions to mitigate climate change. One potential solution for achieving these reduction goals is the direct air capture (DAC) of CO2. However, current DAC methods are expensive due to low incident CO2 concentrations, and the devices themselves consume fossil fuel-based energy. In this article, we propose a sustainable approach that explores the role of wind turbines in modifying local CO2 concentrations. To investigate this concept, large eddy simulations were conducted on two commercial-scale 5 MW wind turbines. Realistic CO2 profiles from 13 different global locations during various seasons were considered. The simulations were performed under neutral atmospheric boundary layer conditions. The results demonstrate that the wake recovery mechanism of a wind turbine promotes rapid mixing of CO2 both above and below the turbine blade tips. In cases where the initial concentrations of CO2 were elevated above the turbine, downward entrainment of CO2 occurred. Conversely, when high concentrations of CO2 were present in the lower atmosphere, wind turbines facilitated a decrease in concentration by up to 138 kg/m within the intermediate wake (7D, D is the diameter of wind turbine) of the second turbine, T2. These findings provide motivation for further exploration of the synergistic use of wind turbines in conjunction with Direct Air Capture devices or local CO2 pollutant diverters to reduce the local CO2. This combined approach has the potential to mitigate local high concentrations of CO2 in the lower atmosphere.
_____________________________________________________________________________________________
Gregory Eyink - Johns Hopkins University
Onsager's "Ideal Turbulence" Theory and Large-Eddy Simulation
Lars Onsager in 1949 proposed a novel theory of high Reynolds-turbulence as described by singular or "weak" solutions of Euler equations. This theory has seen huge progress in the past decade in both the mathematics and the physics literature. It is less well known to the engineering community, despite providing foundations for the large-eddy simulation (LES) modeling method. Recently, mathematicians have begun to develop the Onsager theory for wall-bounded turbulent flows, thus making connection with the critical engineering problem of wall-modeled LES. This talk will review concisely these recent developments and point out, in particular, some rather striking conclusions.
_____________________________________________________________________________________________
Rajat Mittal - Johns Hopkins University
The Force Partitioning Method: A Data-Enabled Method for Dissecting Vortex Dominated Flows
Pressure on a body immersed in a flow is induced simultaneously by vortices, acceleration reaction (a.k.a. added mass) effects associated with body and/or flow acceleration, and viscous diffusion of momentum, and determining the relative importance of these different mechanisms remains one of the most important and fundamental issues in fluid dynamics. Pressure-induced drag and lift are key to the performance of wings, rotors and propellers; undulating fins and flapping wings generate pressure-induced forces that are key to locomotion in fish, birds and insects; time-varying fluid dynamic forces drive flutter and flow-induced vibrations of flexible structures in engineering and biology, and these same forces enable the extraction of energy from flow via devices such as wind-turbines. I will describe the force partitioning method (FPM), a data-enabled method that partitions pressure forces into components due to vorticity, acceleration reaction and viscous diffusion. FPM has been used to gain new insights into a variety of vortex dominated flows including dynamic stall in pitching foils, vortex-induced vibration of bluff-bodies, locomotion of carangiform swimmers and rough-wall boundary layers, and results from these analyses will be presented. Finally, FPM has been extended to aeroacoustics, and applications of the aeroacoustic partitioning method (APM) to dissect aeroacoustic noise in engineering and biological flows will be presented.
_____________________________________________________________________________________________
Laurent Chevillard - ENS-Lyon
MultifraCharles
I will review several aspects of the contributions of Charles and co-workers concerning the statistical modeling of fluid turbulence.
_____________________________________________________________________________________________
Javier Jiménez - Universidad Politécnica de Madrid
The dimensions of turbulence
The dimension in phase-space of the attracting manifold of turbulence is a key constraint to the possibility of data-driven models of its evolution, since it is unlikely that enough data can be generated to train any such model for even moderately high dimensions. Previous attempts to compute the dimension of the full attractor suggest that it is of the order of hundreds, but it may be that the high-probability core of the projection of the system on some reduced set of variables can be approximated as a low-dimensional dynamical system. With this in mind, the generalised dimensions and multifractal spectrum of several thousands of low-dimensional projections of the large scales of a turbulent channel flow are computed, using 100,000 flow fields at Ret»950. It is found that most of their multifractality can be traced to a few individual variables participating in the projections, which are identified. However, low dimensionality is not automatically equivalent to representability as a dynamical system, as shown by examples. Supported by the ERC.
_____________________________________________________________________________________________
Elie Bou-Zeid - Princeton University
Heat transfer over rough walls: fundamental physics, numerical simulations, and bulk parametrizations
Understanding the physical processes modulating the transport of scalars, such as heat, over very rough surfaces is essential for understanding the thermal environment of cities, how wind and solar farms modify heat and water exchanges between the atmosphere and Earth surface, and parameterizing surface physics in coarser Earth systems models. This talk examines this problem in the urban context. We review some of the main flow features from the building to the city scales, comparing to vegetated canopies. We then then examine similarities and dissimilarities between scalar and momentum transport, and the role of dispersive fluxes. While the momentum exchange reaches an Re-independent regimes, the heat transfer never does. Finally, we chart new theoretical pathways to link the momentum and scalar roughness lengths based on surface renewal theory. These lengths are essential for simulations of turbulent flows over rough surfaces from the wall-modeled large eddy simulation to the climate simulation scales.
_____________________________________________________________________________________________
Ronald Joslin and Ben Hallissy - National Science Foundation and Department of Energy
NSF-DoE Partnership in Wind Energy
The presentation will briefly give an Overview of NSF and the NSF/DoE Partnership. An Overview of the DoE Wind Energy Technologies Office will then be presented followed by NSF/DoE Research Funding Opportunities, which include Internships, the Engineering Research Initiation (ERI) program, the Fluid Dynamics program, and Broader Areas of Interest to NSF/DoE Partnership.
_____________________________________________________________________________________________
Richard Stevens - University of Twente
Advancing Wind Farm Modeling through Fluid Physics and High Performance Computing
Wind turbines interact with the environment across a broad range of length scales, spanning from millimeters (viscous scales) and meters (wakes and tip vortices) to "geophysical scales" of hundreds of meters (inter-turbine spacing) up to tens of kilometers (windfarms). This wide range of scales presents significant challenges in theoretical analysis and numerical simulations of windfarm dynamics. Understanding these interactions is crucial to improve windfarm design and operation. Modeling windfarm-atmosphere interactions is an illustrative example of a multiscale physics problem crucial for the renewable energy transition. Successfully addressing this challenge necessitates a synergistic approach involving the development of novel physical models and high-performance computing strategies.
The performance of large windfarms depends on the development of turbulent wind turbine wakes and the interaction between these wakes. Turbulence is crucial in transporting kinetic energy from the large-scale geostrophic winds in the atmosphere towards heights where windfarms can harvest this energy. High-fidelity simulations offer insights into the interaction between windfarms and the atmosphere. While early studies centered on 'ideal' scenarios, recent efforts consider terrain and atmospheric stability. This presentation addresses challenges in modeling and computing challenges in understanding the multi-scale modeling challenges of capturing the windfarm-atmosphere interactions.
_____________________________________________________________________________________________
Huidan Yu - University of Indianapolis-Purdue University Indianapolis
Image-based Computational and Experimental Fluid Dynamics for Pulsatile Flows
The increasing attention within the computational fluid dynamics (CFD) community towards solving pulsatile flows using imaging data has led to a growing need for advanced methods. However, the current numerical approaches, which involve a combination of image processing, CFD, and their interconnection, are plagued by labor-intensive processes, complexity, and a propensity for errors. Often, an ad-hoc amalgamation of software packages is required, making the entire endeavor even more challenging. The confluence of multidisciplinary tasks, encompassing image processing, computational fluid modeling, and high-performance computing, coupled with their intricate integration, can be overwhelming. This presentation will introduce an innovative and robust computational platform designed specifically for image-based CFD. This platform is fortified by concurrent laboratory experimentation and is primarily oriented towards medical applications, particularly for noninvasive and patient-specific diagnoses and therapeutic decision-making in the realm of cardiovascular diseases.
_____________________________________________________________________________________________
Mickael Wilczek - University of Bayreuth
Systematic construction of velocity gradient models for turbulence
The dynamics and statistics of small-scale turbulence can be described in terms of velocity gradients, which makes them an appealing starting point for low-dimensional modeling approaches. Modeling velocity gradients in turbulence requires formulating closures for nonlocal pressure contributions and viscous effects. Reduced-order models for velocity gradients are typically derived by making modeling hypotheses about the small-scale dynamics and statistics of turbulence. Clearly, the fidelity of the resulting models depends on the accuracy of the underlying assumptions. Here, we discuss an alternative, data-driven approach to derive a velocity gradient model that captures given velocity gradient statistics by construction. By analyzing the velocity gradient PDF equation, we distinguish contributions to the single-time statistics from those that impact temporal correlations. We then systematically construct a closure comprising the pressure and viscous contributions to reproduce a given velocity gradient PDF by design. We use the `normalizing flow' machine learning approach to estimate the full eight-dimensional velocity gradient PDF from direct numerical simulation (DNS) data. To assess the fidelity of our model, we benchmark our approach against Lagrangian velocity gradient data from DNS. This contribution presents joint work with Maurizio Carbone, Vincent Peterhans, and Alexander Ecker and is part of a project that has received funding from the European Research Council (ERC) under the European Union's Horizon 2020 research and innovation programme (Grant agreement No. 101001081).
_____________________________________________________________________________________________