Fluid Mechanics Seminars
Abstracts 2000-2001
* * * * * * ** * * * * * *
Monday, June 18, 2001
2:15 - 3:15 pm
479 EBU-II
Dr. Bill Ashurst
Sandia National Laboratories
Livermore, California
"One Dimensional Turbulence: vector formulation and application to
free shear flows"
One Dimensional Turbulence (ODT) is a stochastic model, which
tries to capture the essence of turbulent combustion: fully resolved molecular
fluxes within a turbulent flow. To do
this task a three-dimensional velocity vector is described along a line
(assumed to be the inhomogeneous direction in a temporal or spatially
developing flow). Molecular processes
are solved numerically and the turbulent flow effects of compressive strain
rate and entrainment are accomplished with rearrangement events, which are
called, eddies. Upon random selection
of an eddy length and location, guided by an assumed probability distribution,
the proposed eddy energy is used to determine the eddy time scale, or
frequency. This time scale is compared with
an event rate distribution using a Bernoulli trial process; the eddy acceptance
probability is kept low by the choice of the eddy time step so that the
probability of more than one eddy occurring at one time vanishes. The action of the eddy is to rearrange all
quantities within the eddy length. This
is done by compressing the spatial scale to one-third of the eddy length,
making two copies of the compressed result and placing them to fill out the
eddy domain, the middle copy is then reversed.
This triplet map will make a saw-tooth curve out of a region with a
linear gradient. Repeated application
of the triplet map, at random locations, will generate a lognormal distribution
of gradients.
Using available direct numerical simulations (DNS) of Navier-Stokes
turbulence in two shear flows, planar shear layer and planar wake, we have
calibrated the ODT model by setting one rate term in order to match the
temporal growth rate found in the DNS simulations. Comparison of first, second and third order turbulence quantities
reveals nice agreement considering the simplicity of the ODT assumptions.
New ODT work considers the temporal planar mixing layer in which
the density of one stream may be a thousand times larger than the density of
the other stream. A desired future
application of this work is a model of atomization in a liquid fuel jet. Work done in collaboration with Alan
Kerstein, Scott Wunsch, Vebjorn Nilsen, Tom Dreeben and Rod Schmidt.
* * * * * * ** * *
* * * * *
Monday, June 11, 2001
2:15 - 3:15 pm
479 EBU-II
Dr. Trish Sur
Naval Hydrodynamics Division
Science Applications International Corporation (SAIC)
"Microfluidics"
In this seminar, a brief introduction to the field of
microfluidics and a description of basic microfluidics system components will
be followed by a discussion of a specific application-a MEMS based liquid
cooling system for future micro/nano spacecraft, which is being developed at
the Jet Propulsion Laboratory (JPL).
Since the electronic and other payload power densities in future
micro/nano spacecraft are expected to exceed 25 Watts/cm2, advanced thermal
control concepts and technologies are required to keep their payload within
allowable temperature limits. The
mechanically pumped cooling system consists of a working fluid circulated
through microchannels by a micropump; the microchannel heat sinks have been
fabricated in silicon at JPL. Numerical
and experimental results from this ongoing investigation will be presented.
* * * * * * ** * *
* * * * *
Monday, June 4, 2001
2:15 – 3:15 pm
479 EBU-II
Professor Patrick McMurtry
University of Utah, Salt Lake City
“Turbulent Mixing in a Verrry Long Pipe”
Results from modeling and experimental studies of passive scalar mixing
in a turbulent pipe flow will be presented.
Results support a recently predicted similarity scaling of the
concentration spectra in flows that are unbounded in one direction. Reflecting this scaling, the scalar variance
exhibits a power-law rather than exponential decay, indicating that the
traditional plug-flow reactor picture of turbulent pipe-flow mixing omits key
physical mechanisms. Details of the
experimental technique, which utilizes a caged-fluorescene dye to generate
non-intrusive scalar field initialization, will be provided.
* * * * * * ** * * *
* * * *
Wednesday, May 23, 2001
11:00 – 12:00 pm
479 EBU-II
Professor V.C. Patel
Director, Iowa Institute of Hydraulic Research and
University of Iowa Foundation Distinguished Professor of Mechanical Engineering
The University of Iowa
“Large-Eddy
Simulation of Turbulent Flow over Rough Surfaces”
Large-eddy simulation (LES) with a dynamic sub-grid scale model is employed to study turbulent flow over rough surfaces. Fully-developed flow in two-dimensional channels with one of the walls roughened by sinusoidal waves of varying steepness or rectangular ribs with different spacing is considered to investigate the effect of roughness on the mean flow and eddy structure. The illusive roughness function and other empirically established mean-flow correlations are predicted for each type of roughness. The mean and instantaneous flows show that different types of surface roughness produce quite different turbulence transport characteristics. Turbulence intensities and Reynolds stresses are significantly different with different roughness although the mean velocity distributions are essentially the same. A new approach to modeling of turbulent flow over rough surfaces in the framework of LES is proposed. Surface roughness is decomposed into a resolved-scale roughness, which does not require modeling, and a sub-grid scale roughness, that has to be modeled. Preliminary results show great potential of this approach.
* * * * * * ** * *
* * * *
Monday, May 21, 2001
2:15 – 3:15 pm
479 EBU-II
Haris J. Catrakis
Mechanical and Aerospace Engineering
University of California, Irvine
“Mixing, Fluid Interfaces, and Optical Wavefronts in Turbulent Shear Flows”
Progress and issues in our understanding of turbulent shear flows are discussed with emphasis on the mixing efficiency of turbulence, the structure and dynamics of fluid interfaces, and the propagation of optical wavefronts through turbulent shear flows. In particular, new three-dimensional space-time measurements of turbulence-generated fluid interfaces recorded in a novel flow facility are found to exhibit scale-dependent behavior even at flow conditions above the mixing transition, at least at the Reynolds numbers investigated. Turbulence-degraded optical wavefronts are also found to exhibit scale-dependent behavior. Prospects for developing predictive quantitative descriptions of such behavior are discussed.
* * * * * * ** * * * * * *
Monday, May 14, 2001
2:15 – 3:15 pm
479 EBU-II
Professor G. R. Spedding
Department of Aerospace & Mechanical Engineering
University of Southern California
“Applications and
Implications of Stratified Wakes Research"
When a patch of turbulence is free to evolve in a
stably-stratified environment (such as the earth's atmosphere, or ocean
thermocline), it does so in ways that are quantitatively and qualitatively
different from its counterpart in a homogenous fluid. The initially-turbulent
wake behind a bluff body is one of the canonical flows in stratified turbulence
studies, and accurate, whole-field, quantitative velocity measurement techniques
reveal certain startling characteristics with potential practical
significance. Partly because
stratified wakes research can inhabit a domain that balances the tension
between basic and applied research communities and requirements, the research
program sponsored by ONR (Office of Naval Research) in stratified wakes has
been brisk and vigorous during the last 8 years. A number of questions with rather practical motivation have
provoked research and answers that have unexpectedly broad application. At the same time, many fundamental questions
remain. This talk will summarize the
progress so far, including comparisons with recent numerical experiments.
* *
* * * * ** * * * * * *
Monday, May 7, 2001
2:15 – 3:15 pm
EBUII - Room 479
Prof. L. Gary Leal
Department of Chemical Engineering
University of California, Santa Barbara
“Flow-induced coalescence in viscous fluids”
The problem of flow-induced coalescence is a key to understanding a
number of technological problems. However, it is difficult to study either
experimentally or theoretically, and our understanding of it is surprisingly
limited. In this talk, I describe recent studies from my group, based primarily
on experimental investigations of the interaction of two drops in flow of a
viscous fluid. These studies were carried out in a miniaturized version of the
4-roll mill, and lead to the first quantitative data (so far as we are aware)
on the critical conditions for coalescence.
I will consider systems that have a clean fluid interface, as well as the
effects of large surfactants that are relevant to modeling of the process of
reactive blending of polymeric fluids.
The results expose a number of unresolved questions that are the subject
of current investigation.
* *
* * * * ** * * * * * *
Wednesday, May 2, 2001
2:00 - 3:00 pm
479 EBU-II
Professor Andrew W. Woods
BP Institute, University of Cambridge
“Gravity Driven Flows in Porous Rocks"
The talk will describe a series of problems involving liquid infiltration into porous media in which the dominant force driving the flow is gravity. Flows through heterogeneous porous layers, flows with stratified physical properties, flows which involve liquid-vapour phase change and flows which react with the medium will be described. The presentation will describe mathematical models of the flows, a range of similarity solutions for these flows, together with a series of laboratory experiments, which have been carried out to test the models. Applications to oil recovery, geothermal power, groundwater transport and geological processes will be discussed.
* * * * * * ** * * * * * *
Monday, April 30, 2001
2:15 – 3:15 pm
479 EBU-II
James J. Riley
University of Washington
"Dynamics of Turbulence Strongly Influenced by Buoyancy”
When turbulence is generated in a stably-stratified fluid, e.g., in the open ocean or in stable regions of the atmosphere, it often occurs locally without a continuous source of energy. Such turbulence will ultimately decay, and at some point in its decay the effects of stable stratification will become important and even dominant. Laboratory experiments indicate that, in thelater times of decay, when the effects of stratification become dominant, the characteristics of the flow change dramatically, and the flow appears to be dominated by quasi-horizontal vortical motions with considerable vertical structure. Theoretical arguments suggest that, at high Reynolds numbers, such flows may be subject to several instabilities. These instabilities, however, may not be possible at the low Reynolds numbers characteristic of the late times in the laboratory experiments. Therefore the behavior of such flows at the higher Reynolds numbers characteristic of the oceans and atmosphere is still an open issue.
In this
seminar the results of some preliminary high resolution direct numerical
simulations which address this problem will be discussed. The simulations are initialized in the late
stage, i.e., stratification dominated, regime; cases for several Froude numbers
and Reynolds numbers are investigated.
It is found that quasi-horizontal vortices develop, and are subject to
two types of instabilities. The first
is related to differential growth in scale of the vortices in horizontal
planes, leading to strong vertical shearing of the horizontal motion, while the
second is local and develops in regions of high vertical shear. The implications of these results regarding
the scalability of the laboratory data will be discussed.
* * * * * * ** * * * * * *
Friday, April 20, 2001
11:00 – 12:00 pm
479 EBU-II
Linda Hedges
Principal Engineer
Boeing Aircraft
“Unsteady Simulations of Massively Separated Flow”
Airframers now compete very intensely in terms of community noise. Today’s industrial noise-prediction methods are highly empirical. With an impending step in community-noise requirements there is urgency for developing new tools to predict and reduce aerodynamic noise generated by complex geometries such as an extended airplane landing gear. Physics-based methods for predicting noise around an aerodynamic source couple a detailed model of the flow field immediately around the source with a linear acoustic model, such as Lighthill’s equations, for the prediction of the radiated noise that propagates to the far field. The detailed model of the flow field must include a comprehensive unsteady description of the turbulent flow around the source, including information such as fluctuations in velocities and wall pressures.
Detached-Eddy Simulation (DES) is a recently developed hybridization of Large-Eddy Simulation (LES) and Reynolds-Averaged Navier-Stokes (RANS). The unsteady and/or massively separated regions of the flow are modeled with LES, so that the bulk of the flow field is as detailed as those flow fields obtained with LES. Since the major features of the flow involve length and time scales much larger than those of the boundary-layer turbulence, the boundary layer modeling is performed with the RANS equations with little loss in the description. The computational requirements for DES are similar to unsteady RANS, whereas an LES of the entire boundary layer at useful Reynolds numbers is far out of reach, and the accuracy of the complete flow field is potentially similar to LES. Results on airfoils, cylinders and spheres have been highly encouraging. Very few complex geometries have been attempted with DES or LES to date. Additionally, DES has yet to be applied to noise prediction. The method is promising, but still very new.
The methods of DES and
Unsteady RANS are applied to the flow behind a landing gear truck. The geometry
is based on a 31% scale Boeing 757 main landing bogie tested by Lazos at the
NASA Langley Research Center in the Basic Aerodynamic Research Tunnel (BART).
The DES flow calculations display a wide range of scales, and appear to provide
a sound basis for noise generation, structural vibrations, and drag.
* * * * * * ** * * * * * *
Monday, April 9, 2001
2:15 p.m.
479 EBU-II
Professor Hassan M. Nagib
Illinois Institute of Technology
Chicago, IL
“The Overlap Region
of Wall-Bounded Turbulent Flows and the Log-Law Versus Power-Law Debate”
The two experiments of Oysterlund and Hites
(e.g., see note in Phys. Fluids, vol. 12, no. 1, p. 1, 2000) have been recently
extended to Reynolds numbers based on momentum thickness exceeding 70,000. The present measurements are carried out on a
flat plate in zero-pressure gradient in the National Diagnostic Facility
(NDF). As in our previous work, the
wall shear stress is directly measured using oil-film interferometry and
hot-wire anemometry is utilized to measure the velocity. Contrary to the recent conclusions of
Barenblatt, Chorin and Prostokishin (JFM, vol. 410, p. 263, 2000, and Phys.
Fluids, vol. 12, no. 9, p. 2159, 2000; see also our comment on page 2360 of the
same issue), the mean velocity distribution in the overlap region for these higher Reynolds number boundary layers
continues to be very accurately described by the Reynolds-number-independent log law. Our work has established that the subtle differences between the
two relations representing the overlap region, and the data, can only be
revealed with the aid of the derivatives of the velocity with respect to the
distance from the wall; particularly when only low to moderate Reynolds number
experiments are available. As concluded
from the earlier two experiments, the values of the log-law coefficients that
best represent the data are k =
0.38 and B = 4.1. In addition, the data from four independent
investigations of turbulent flow through two-dimensional channels and pipes
have been examined using the same techniques to establish which of the two
relations is more representative. These
investigations include DNS, LDV and PIV of the channel flow, and the
Pitot-probe measurements in the super-pipe experiment. Finally, brief comments will also be
included on asymptotic analyses of turbulent boundary layer and channel flows
with an emphasis on the wall region and the various ways of scaling the outer
flow.
* * * * * * ** * * * * * *
Monday, March 12, 2001
2:15 p.m.
479 EBU-II
Dr. Tony Leonard
California Institute of Technology
"Flow Induced Vibration: Lessons Learned from Numerical
Experiments"
In the past several years we have been working to improve our understanding of the relation between the unsteady forces, near-wake vorticity field and the structural motion in the context of flow-induced vibration. A combined experimental-computational approach has been used. In this talk the emphasis will be on the numerical experiments.
* *
* * * * ** * * * * * *
Monday, March 5, 2001
2:15 p.m.
479 EBU-II
Dr. Carlos Haertel
ETHZ
Zurich, Switzerland
"Analyzing Structure and Stability Properties of Gravity Currents by
Direct Numerical Simulation"
Results are presented from an ongoing study, which is concerned with the computational analysis of density-driven and particle-driven gravity currents. The study deals with generic flow configurations, like lock-exchange flows or finite-volume releases in plane and axisymmetric settings. It concentrates on the case of small density differences where the Boussinesq approximations apply. To solve the governing equations numerically, high-order methods based on spectral and spectral-element discretizations are used. In the discussion of the results, special attention will be given to the pronounced frontal instability commonly observed with gravity currents, and on details of the energy budget of such flows. Previous theoretical and experimental findings concerning the structure and dynamics of gravity currents will be discussed in light of the present simulation results.
* *
* * * * ** * * * * * *
Monday, February 26, 2001
2:15 – 3:15 pm
EBUII - Room 479
Professor C. Pozrikidis
Mechanical and Aerospace Engineering Department
University of California, San Diego
"Mathematical Modeling and Numerical Simulation of The
Biofluid-Dynamics of Red Blood Cells"
Red blood cells are liquid capsules containing a viscous fluid that is enclosed by a biological membrane consisting of a lipid bilayer and a supporting network of proteins. In the absence of flow, the cells assume the shape of a biconcave disk. When subjected to flow, the cells deform in a way that is determined by the type and strength of the flow and by the mechanical properties of the membrane which is known to behave like a viscoelastic sheet. In this talk, an integrated mathematical description of the equations governing the fluid dynamics and membrane mechanics is presented working under the auspices of low-Reynolds-number hydrodynamics coupled with the nonlinear theory of thin shells. The governing equations are solved using a novel implementation of the boundary element method that accounts for the membrane elasticity and bending stiffness. Parametric investigations demonstrate the significance of the membrane properties on the cell deformability and on the magnitude of the developing membrane tensions, and thereby establish a relationship between membrane structure and cell behavior in large-scale or capillary blood flow.
* *
* * * * ** * * * * * *
Tuesday, February 20, 2001
2:15 – 3:15 pm
EBUII - Room 479
Dr. George Carnevale
Scripps Institution of Oceanography
"Overturning and Mixing in the Thermocline Forced by Wave-Wave
Interactions"
The oceanic thermocline is a highly stratified transition zone from warm well-mixed surface water to the much colder deep water. Recent observations [Alford and Pinkel, 2000] of temperature fluctuations in this zone have found strong overturning and mixing events with scales of about 2 m in the vertical. There is a correlation between these overturns and the passage of internal waves of vertical scale on the order of 10 m that produce regions of high vertical strain rate. These scales are in the buoyancy range, a region of the spectrum that represents a transition from large-scale wave-dominated motions to small-scale turbulence. We investigate this transition through simulations in which regions of high strain rate are established by forcing a 20 m standing internal-wave. Breaking and mixing are found in both the regions of high strain rate and the regions of steep isopycnal slope that are produced by this forcing. The observations also suggest that propagating wave packets of vertical extent of about 50 m with internal vertical length scale of about 10 m may be important in causing overturning in the thermocline. For a typical set of environmental and wave-packet parameters, we find packets that can travel over approximately 200 m in the vertical, as in the observations. The waves in these packets break and generate turbulence that forms a continuous wake of small-scale perturbations as perdicted by Thorpe (1999).
* *
* * * * ** * * * * * *
Monday, February 12, 2001
2:15 – 3:15 pm
479 EBU-II
Dr. Tihomir Hristov
Department of Mechanical and Aerospace Engineering
University of California, Irvine
"Critical Layer Effects in the Wind above Ocean Waves"
Ocean surface waves are perceived to modify
profoundly the structure of the atmospheric boundary layer over the ocean,
although the physical mechanism behind the phenomenon is still incompletely
understood. Starting with the critical layer theory (Miles, 1957), large body
of theoretical and numerical work has built on the concept of resonant
wind-wave coupling. However, insufficient information from experiments
precluded the verification of this concept. In particular, the paradigm of
resonant wind-wave interaction has been questioned by early experiments, which
have found no wave-induced fields in the wind.
This presentation discusses analysis of data acquired in the open ocean from
the stable instrument platform FLIP. By employing a provably optimal filtering
technique, sizeable wave-induced fields are detected, thus supporting the
concept of resonant wind-wave coupling. Also, we select distinct features of
the wave-induced flow, predicted by the critical layer theory, such as
discontinuity of the phase and minimum of the amplitude at the critical height.
Data show signatures of these features clearly and persistently, indicating
that for the resolved wave scales the critical layer mechanism is active in the
open ocean.
* * * * * * ** * * * * * *
Monday, February 5, 2001
2:15 – 3:15 pm
479 EBU-II
Professor Andrew J. Bernoff
Harvey Mudd College
"Dynamics & Stability of Van der Waals Driven Thin Film
Rupture"
When manufacturing optical coatings and insulating layers in micro-circuits, it is desirable to create thin fluid films with thicknesses of less than a micron. For these films, van der Waals forces and surface tension dominate. Williams & Davis (1982) first derived a lubrication model for films under these forces. More recently, Zhang \& Lister (1999) showed that this model has a countable infinite set of similarity solutions corresponding to finite-time film rupture. In this talk, we will examine the dynamics and stability of thin film rupture in both planar and axisymmetric geometries.
In particular, as the film thickness decreases, the flat state loses stability as the attractive van der Waals forces of the substrate overcome the stabilizing influence of surface tension. This loss of stability generally leads to finite-time rupture of the thin film. We describe a systematic technique for calculating self-similar rupture solutions and determining their linear stability. The dynamically stable similarity solutions produce observable rupture behavior as localized, finite-time singularities in the models of the flow. For the problem of axisymmetric van der Waals driven rupture, we identify a unique stable similarity solution for point rupture of a thin film and a second mode of singularity formation corresponding to annular ``ring rupture.'' We use an energy argument to relate the initial perturbations of the interface and the bifurcation structure to the mode of rupture observed. Finally, we show both analytically and numerically that although axisymmetric ring rupture is observable as a long-lived transient in the full system, it is eventually subject to a transverse instability favoring rupture at a single point on the annulus.
* *
* * * * * * * * * * *
Monday, January 29, 2001
2:15 – 3:15 pm
479 EBU-II
Professor Paul Roberts
Department of Mathematics, UCLA
"Topics in Super Fluid Mechanics"
Many of the fascinating properties of superfluids have been known for nearly a century, and early models, such as that of Landau for superfluid Helium, were strikingly successful in describing their behavior. It was first realized in the mid 50s that vorticity is quantized, and this was demonstrated in the beautiful experiments of Packard and Williams. Turbulence in a superfluid is therefore a kind of "vortex tangle" of quantized vortex lines. Although the vorticity in a superfluid is discrete, it "tries", in a sense, to mimic the continuous vorticity of a classical fluid in similar situations. For example, an impurity moving sufficiently rapidly through a superfluid tends to shed vorticity in much the same way as a classical fluid would. One way of describing such processes in a qualitatively satisfactory way is through a particular type of nonlinear Schroedinger equation, the Gross-Pitaevskii equation. Such solutions will be presented in this talk. This model is used to elucidate different aspects of superfluid behavior: the motion, interactions, annihilations, nucleation and reconnections of vortex lines, vortex rings, and vortex loops; the motion of impurities; flow through apertures; superfluid turbulence and the capture of impurities by vortex lines. The GP equation has also been used with success to describe the behavior of the recently discovered low density condensates.
* *
* * * * * * * * * * *
Monday January 22, 2001
2:15 – 3:15 pm
479 EBU-II
Professor Edward J. Kerschen
University of Arizona, Tucson
"Sound Generated by Gust-Airfoil Interactions"
Gust-airfoil interactions are responsible for much of the noise generated by the fans, compressors and turbines of aircraft engines, and are an important source of noise in many other applications as well. Typical convected disturbances (gusts) are atmospheric turbulence, viscous wakes of upstream bodies, trailing vorticity shed from three-dimensional lifting surfaces, and fluid temperature variations in the case of flow exiting a combustor. The physics of the phenomenon depends crucially on the source compactness--the ratio of the airfoil chord to the wavelength of the generated sound. Interactions in low Mach number flows are generally compact, while at high subsonic Mach numbers non-compact effects become very important. An overview of the author's work in this field is presented. Attention is placed on the non-compact limit, where scattering phenomena at the leading and trailing edges of the airfoil become very important. In the non-compact limit, the airfoil shape and angle of attack significantly affect the radiated sound field, in ways that are best elucidated by singular perturbation analysis. Prospects for the active control of sound generated by gust-airfoil interactions are also briefly discussed.
* *
* * * * * * * * * * *
Tuesday January 16, 2001
1:45 – 2:45 pm
479 EBU-II
Professor Francesco Paparella
Scripps Institution of Oceanography
"Excitation of basin modes by air-sea coupling"
A conceptual model of the coupling between the upper-ocean wind-driven circulation and the mid-latitude atmospheric wind-stress illustrates that large-scale basin-wide oscillations with decadal period can be excited. These oceanic modes are also found in the absence of ocean-atmosphere feedback, but they are damped. The period of the oscillation and the spatial structure of the modes are essentially the same with and without coupling. These oscillations are distinct from the coupled modes of variability arising from a delayed negative feedback between the wind-driven flow and the wind-stress. They are ocean-only linear basin modes which, become sustained by ocean-atmosphere coupling.
* *
* * * * * * * * * * * *
Special Seminar
Tuesday January 9, 2001
2:00 - 3:00 pm
479 EBU-II
D.D. Joseph
Regents and Russell J. Penrose Professor
Department of Aerospace Engineering and Mechanics
University of Minnesota, Twin Cities
"Lift Correlation's from Direct Simulation of Solid-Liquid Flow"
Lift acting on a fluidized particle plays a central role in many important applications, such as removal of drill cuttings in horizontal drill holes in the petroleum industry, sand transport in fractured petroleum reservoirs and the cleaning of particles from surfaces. The levitation of 300 particles in a Poiseuille flow was numerically studied, from which a data bank was obtained. The method of correlations provides a direct link between direct simulation and engineering application. The roles of slip and slip-angular velocity in the lift-off of particles will be elucidated in this lecture.
* *
* * * * * * * * * * * *
Monday, January 8, 2001
2:15 - 3:15 p.m.
479 EBU-II
Professor Adam Fincham
LEGI-UJF-INPG-CNRS
Grenoble, France
"Vortices and vorticity measurement in stratified and rotating
fluids"
Late time stratified flows are characterized by extreme anisotropy where the vertical component of velocity w, plays little part in the vortex dynamics and is best described as part of an associated internal wave field. Though quasi-2D in nature these flows exhibit fully 3D vorticity fields as the tendency for layering creates strong vertical shear and promotes a horizontal alignment of the vorticity vector. Standard full field imaging velocimetry measurement techniques capable of providing the component of vorticity perpendicular to the plane of the measurement, are inadequate for diagnosing the complex 3D vortex topology of these flows. Careful exploitation of the inherent anisotropy, and the use of relatively large experimental tanks where the Reynolds number is obtained primarily from the length scale, allows for time resolved, full 3D measurement of all three components of vorticity. Application of the 3D measurement technique to a variety of flows including stratified dipolar vortices and stratified wakes reveals interesting vortex topologies where closed vortex lines force a compactness of these net zero circulation structures. Comparison between the global energy decay and the internal viscous dissipation demonstrated the accuracy of the measured deformation tensor and allows determination of energy loss from other sources such as internal wave radiation.
* *
* * * * * * * * * * * *
Monday, November 27, 2000
2:15 - 3:15 p.m.
479 EBU-II
Professor Niel Balmforth
Department of Applied Mathematics
University of California Santa Cruz
"Meandering of a Barotropic Jet"
I consider the dynamics of an unstable jet of two-dimensional, incompressible fluid on the beta-plane. In the inviscid limit, the instabilities correspond to discrete eigenmodes emerging from a continuous spectrum, which has important repercussions on the weakly nonlinear theory of the modes. In particular, standard dimensional reduction techniques fail to give a low-order description of this problem, partly because the simple shape of the unstable normal mode is insufficient to capture the structure of the forming pattern. That pattern takes the form of "cat's eyes" in the vorticity distribution which develop inside the modal "critical layers" (slender regions to either side of the jet's axis surrounding the levels where the modal wavespeed matches the mean flow). Asymptotic expansions are used to generalize the weakly nonlinear theory.
* *
* * * * * * * * * * * * *
Monday, November 13, 2000
EBU-II - Room 479
2:15 p.m. - 3:15 p.m.
Professor T. Maxworthy
Department of Aerospace and Mechanical Engineering
University of Southern California
A history of the types of laboratory experiments used to study oceanographic phenomena will be reviewed briefly. Then emphasis will be placed on two or three examples of personal interest that will be presented in a little more detail. These will most probably be chosen from: Turbulence in Stratified Fluids, Particle Laden Gravity Currents, Gravity Current Propagation through a Linearly Stratified Fluid, The Dynamics of Partially enclosed Seas.
* *
* * * * * * * * * * * * *
Monday, November 6, 2000
EBU-II - Room 479
2:15 p.m. - 3:15 p.m.
Professor William Young
Scripps Institution of Oceanography
University of California, San Diego
In this talk I introduce a pedagogical model of eddy diffusion and use a combination of simulation and analytic solution to illustrate the distinctions between ensemble averages and spatial averages. If the spatial average of a single realization is approximated by an ensemble average then a flow is said to be "self-averaging". The mixing parameterizations used in ocean circulation models plausibly assume the self-averaging property for physical and chemical tracers. But, because of reproduction, biological tracers (plankton) present additional difficulties which may defeat the self-averaging assumption: because birth is always adjacent to a living individual there is a uniquely biological source of correlation's between pairs of organisms (parent and progeny). Consequently, when a planktonic species reproduces in a turbulent fluid, filamentary patches of plankton, form as a result of an inverse cascade which starts on the length scale of an individual plankter. the implications of this for advection-reaction-diffusion modeling of plankton population dynamics will be discussed.
* *
* * * * * * * * * * * * *
Monday October 30, 2000
EBU-II - Room 479
2:15 p.m. - 3:15 p.m.
Dr. Evgeny Novikov
Institute for Nonlinear Science
University California San Diego
The dynamics and statistics of turbulent flows is better understood in terms of characteristics of motion which are local in physical space and have a nonlinear mechanism of amplification. For the three-dimensional turbulence the primary local characteristic is the vorticity field. For the two-dimensional turbulence the corresponding local characteristics provide a natural way to measure the intermittency effects. A description of turbulence in terms of local characteristics naturally leads to the conditional averaging with a fixed local characteristic at a point. It also leads to a new scaling, which has a experimental support. Conditional averaging and the new scaling provide a basis for a statistical enslavement of the small-scale turbulence, which can be used for the large-eddy simulations in various applications. Conditional averaging statistical modeling of turbulence. Among other components are Markove modeling, transformations from homogeneous to inhomogeneous turbulent flows and a variational approach to turbulent boundary layers.
* *
* * * * * * * * * * * * *
Monday, October 23, 2000
EBU-II 479
2:15 p.m.-3:15p.m.
Professor J.C.R. Hunt
Department of Geological Sciences & Department of Space & Climate
Physics
University College London
"The Relative Movements and Coalescence of Small Inertial Particles in
Vortices and Turbulence"
Recent studies of the relative movement of small inertial particles around vortices show that over a narrow range of particle size, corresponding to a range of relaxation time and settling velocity, and circulation of line vortices the particles tend to be concentrated outside the vortices and on average to fall faster. This leads to greater coalescence as larger particles collide with smaller ones. Experimental and numerical evidence is provided to show that this mechanism changes the Lagrangian spectra and bulk settling velocity of particles in turbulence. Since the kolmogorov microscale eddies have a circulation of the order of the molecular viscosity, this explains how turbulence causes cloud droplets to form by enhanced coalescence at diameters of the order of 20 microns, independent of the rate of dissipation. Thence a theoretical model from first principles provides particle size distributions in clouds. Some curious effects of very small aerosols on clouds can also be explained.
* *
* * * * * * * * * * * * *
Monday, October 16, 2000
2:15 -3:15 p.m.
EBU-II 479
Professor W. D. Smyth
Oregon Sate University
"Mixing in Turbulent Overturns: Direct Simulations and Oceanographic
Observations"
Stratified turbulence is often assumed to mix with efficiency near 0.2, i.e. about one-fifth of the energy supplied to the turbulence is used to raise the gravitational potential energy of the fluid, while the rest is dissipated as heat. However, recent measurements in the ocean suggest that mixing efficiency actually varies quite significantly, ranging from near zero to near unity. One possible explanation for the variability is time dependence: mixing efficiency varies over the lifetime of a turbulent event.
In this talk, I'll describe direct numerical simulation of shear-driven overturns in a stratified fluid with Prandtl number exceeding unity. Comparison with observations show that these events provide a useful model for turbulent overturns in the ocean. We find that the pre-turbulent, rollup phase is characterized by very high mixing efficiency. Mixing efficiency then decreases as the overturn breaks and becomes turbulent. As also allow us to develop methods for assessing time dependence of mixing efficiency from in situ measurements of ocean turbulence. The results are consistent with the simulations, i.e. newly-created overturns mix more efficiently than those which have become fully turbulent.
These results challenge our basic understanding of the relationship between mixing and turbulence. How is it that a laminar flow, despite its relatively low strain rate, is able to mix so effectively in comparison with fully turbulent flow? A physical explanation for this surprising result will be proposed.
* *
* * * * ** * * * * * *
Monday, October 9, 2000
EBU-II 479
2:15 -3:15 p.m.
Dr. Stan Luckhardt
Mechanical Aerospace & Engineering
University California San Diego
Plasma Turbulence
"Boundary Plasma Physics and Surface Interaction Experiments"
Present day experiments with magnetic fusion plasmas generate large heat fluxes (1-100MW/m-2) and particle fluxes (1023-1024 ions m-2 s-1) creating a plasma interaction region with nearby surface materials. Understanding the resulting plasma boundary layer is generally critical to the success of these experiments, and encompasses the disciplines of plasma physics and surface materials interactions sciences. A range of time dependent phenomena are typically encountered in the plasma boundary including coherent plasma modes such magneto-hydrodynamic (MHD) and plasma wave modes, transient events such as ELMs and disruptions, and broadband plasma turbulence. These phenomena exhibit complex two and three-dimensional structure and time evolution. A major challenge for fusion experiments in the coming years will be to developments for systematic measurement of the full 2 and 3D nature of these phenomena. Such measurements should provide direct images or accurate reconstruction's of these coherent plasma modes and turbulence. Here at the UCSD PISCES laboratory, in collaboration with the Princeton Plasma Physics Laboratory, we have developed a novel plasma density imaging technique based on cathodoluminescent phosphor tecchnology.* Fluctuations in plasma electron density are converted into a visible two dimensional image by means of a Zn doped ZnO phosphor coated plate inserted into the plasma. The important features of this technique are its fast time response (microsecond range), sensitivity to low energy boundary plasma electrons (several eV energy range), millimeter range spatial region of the plasma. the resulting images provide unique insights into plasma turbulence and flows.
A. Liebscher, S.C. Luckhardt, s. Zweben, 10th
International Conference on High Temperature Plasma Diagnostics, (July 2000)
accepted for publication in Review of Scientific
Instruments.
* * * * * * * ** * * * * * *
Monday, October 2, 2000
EBU-II 479
2:15 - 3:15 p.m.
Professor G.R. Tynan
Mechanical Aerospace and Engineering Department
University California San Diego
"From Fusion Reactors to Planetary Atmospheres:
Searching for the cause of reduced turbulent mixing and mean shear flow
generation in 2D fluids"
Turbulent mixing in 2D fluids (such as magnetized plasmas and planetary atmospheres) can be substantially reduced under appropriate circumstances. the consequences of these reductions can be important (e.g. reduced heat losses from prospective fusion reactors, reduced mass or heat transfer planetary equatorial and or polar regions) and thus the process is of significant interest. In this talk we present several second-order correlation measurements of density, fluid velocity, and temperature fluctuations obtained in tokamaks to provide insight into the mechanism of reduced turbulent mixing. Next, we examine the spontaneous generation of strong mean sheared flow, which is spatio-temporally correlated with the turbulent transport reduction. Such large-scaled flows are thought to be generated via the turbulent Reynolds stress, and can be viewed as an increase in the inverse-cascade rate of the fluctuation energy. Thus measurement of the inverse energy cascade rate directly(using a transient perturbation) or indirectly (using bispectral and higher-order techniques) should permit the underlying nonlinear dynamics of the shear-flow generation to be experimentally addressed. Preliminary observations of increased nonlinear coupling between turbulent-scales and mean-flow scales in tokamak devices will be presented, and plans for a more controlled study of mean-flow/turbulance interactions in a quiescent plasma device under construction at UCSD will be discussed. Finally, the similarities between magnetized plasma fluid dynamics and the atmospheric dynamics of rotating planets and stars suggest that similar search be carried out using data from a rotating water tank and/or planetary atmospheric measurements. An approach for this work is proposed.
* * * * * * ** * * * * * *
Monday, September 25, 2000
2:15 -3:15 p.m.
EBU-II 479
Professor Yuji Hattori
Kyushu Institute of Technology (Japan
"Direct Simulation of Sound Generated by Collision of Vortex
Pairs"
The sound generated by head-on collision of two-dimensional vortex pairs is studied by direct numerical simulation. A high-precision numerical scheme is used to capture the sound, which is much smaller than the ambient pressure. Two types of vortex motion are considered: symmetric case and asymmetric case. In the symmetric case, the vortex pairs have the sam translational velocity and the combination of vortices is changed after the collision. The sound has two clear peaks and the quadrupole component is dominant. In the asymmetric case asymmetric case, the vortex pairs have different translational velocity and the orbit of the faster pair has a loop. The sound has four clear peaks. The largest component of the sound is the quadrupole one, but the sound is asymmetric, or in other words, the octupole components cannot be neglected in spite of low Mach number. The sound obtained numerically is in good agreement with the prediction by acoustic analogies. The mechanism of sound generation is shown in some details.
For more information call 534-6029