Vered RomKedar
Department
of computer
science and applied mathematics
Weizmann Institute
Research interests and
publications:
To
chronological publication
list
Steep
billiardlike potentials 
Transport
theory 
Chaotic fluid mixing 
Exponentially
small splitting Abstract
Publications 
Forced NLS
& Parabolic resonances 
Modeling & medicine 
Steep
Billiardlike potentials and steep Impacts
with M. Kloc, L. Lerman, V. Gelfreich, A. Rapoport, K. Shah, D. Turaev
Main
motivation and
results:
One of the underlying assumptions of statistical mechanics is the Boltzmann ergodic
hypothesis
which roughly states that the dynamics of a sufficiently large number
of small
molecules in a box is ergodic. The
mathematical
investigation of this hypothesis had focused mainly on studying the
dynamics of
hard spheres, and lead to fundamental developments in the theory of
dynamical
systems. Studying the motion of n "real" particles, having smooth
steep repelling potentials, is a mathematical challenge with important
implications to Physics.
To
this aim we develop methodologies for analyzing Hamiltonian systems
with steep
repelling potentials by using the singular billiard limit.
Mathematically 
they allow to study far from integrable
systems in
arbitrary dimension. Physically  they allow to
study
when the widely used idealized billiard model works, and, more
surprisingly,
when it fails. These methodologies are being extended to steep impact
systems  here, in the singular limit, particles move under nontrivial
potential in the domain interior and reflect from the domain boundary
as in the billiards.
In
[R15,R17,R21] we establish that in the two
dimensional
case the smooth steep potentials can indeed ruin ergodicity.
These findings lead to an experimental realization of this phenomenon
in an
atomoptic system (Kaplan et al, PRL, 2001). We recently showed that
similar
approach may be utilized to predict and control the form of the chaotic
scattering function of various scatteringbysteeppotentials problems
[R29].
In [R28], we made the multidimensional leap: we proved that elliptic orbits appear in systems with steep smooth potentials that limit to Sinai billiards for arbitrarily large dimension (see figure for effective stability islands in a 20dimensional phase space). Admittedly, the considered special example cannot be directly linked to the smooth many particles case. Nonetheless, this construction, of stabilizing mechanism in multidimensional highly unstable systems, suggests that analogous stabilizing mechanisms may exist in the smooth Boltzmann problem.
More recently [R34,R38, R44], we study classical particles accelerators: billiards with oscillating boundaries. While it has been already established that 2D accelerators can lead to unbounded growth of the averaged energy of ensembles of particles (i.e. to heating), we are the first to construct accelerators in which this growth is exponential in time.
NEWS:
We have demonstrated that the exponential in time energy growth mechanism may be induced by a billiard with mixed phase space (the Oscillating Mushroom). Moreover, we derived adiabatic theory for nonergodic systems from which the exponential growth rate is predicted and is then corroborated numerically [R46].
[R48] 
K.
Shah, V.Gelfreich,
V.
RomKedar and D. Turaev 
Leaky
Fermi Accelerators 
PRE 91, 062920 (2015)  
[R46]  Oscillating mushrooms: adiabatic theory for a nonergodic system  Journal of Physics A : Mathematical and Theoretical, Volume 47 (Number 39). Article number 395101 . ISSN 17518113  
[R44]  V.Gelfreich, V. RomKedar and D. Turaev  Fermi acceleration and adiabatic invariants for nonautonomous billiards  Chaos 22,
033116 (2012); http://dx.doi.org/10.1063/1.4736542 (21 pages). 

[R43]  V. RomKedar and D.Turaev,  Billiards: a singular perturbation limit of smooth Hamiltonian flows"  Chaos 22,
026102 (2012); http://dx.doi.org/10.1063/1.4722010 (21 pages) 

[R38]  V. Gelfreich, V. RomKedar, K. Shah, D. Turaev,  Robust
exponential
accelerators 
PRL 106, 074101, 2011.  
[R34]  K. Shah, D. Turaev and V. RomKedar,  Exponential energy growth in a Fermi accelerator,  Phys. Rev. E 81, 056205, 2010  
[R29] 
A.
Rapoport and V. RomKedar 
Phys. Rev E., 77, 016207 (2008). 

A. Rapoport, V. RomKedar and D. Turaev 
Comm. Math. Phys, 279, 497534, 2008. 


V. RomKedar and D.Turaev 
Physica D, 130:187210, 1999. 

A. Rapoport, V. RomKedar and D. Turaev. 
Approximating
multidimensional Hamiltonian flows by billiards. 
Comm. Math. Phys., 272(3), 567600, 2007. 

[RP8]

D. Turaev and V. RomKedar 
In
Chaos, kinetics and nonlinear dynamics in fluids and Plasmas (CarryLe Rouet, 1997), volume 511 of Lecture Notes in
Phys., pages 1750. Springer, 

[R15] 
D. Turaev and V. RomKedar 
Nonlinearity, 11(3):575600, 1998. 


D. Turaev and V. RomKedar 

J. Stat. Phys., 112(34):765813, 2003 

[R25]

A. Rapoport, V. RomKedar 
Nonergodicity of the motion in three dimensional steep repelling dispersing potentials. 
Chaos 16, 043108, 2006. 
Multidimensional
systems, forced NLS and Parabolic Resonances
with
Main
motivation and
results:
Nearintegrable systems had been
extensively studied
since the days of Poincaré, yet,
the structure
of their chaotic solutions is still mostly unknown. A decade ago we
discovered
a new kind of instability in such systems, the parabolic resonance
instability.
It was discovered while studying a geophysical application
[R9,R11,R12], and
then we studied it in the multidimensional setting [R19,R20,R22]. More
recently, we proposed that the hierarchy of bifurcation framework
[RP11,R23] is
convenient for identifying the dominant instability mechanisms that
appear in
specific multidimensional models.
Notably, applying this framework to the conservatively forced NonLinear Schrödinger (NLS) equation, conditions under which a truncated form of this equation has a parabolic resonance instability were derived [R23]. It was then demonstrated that such instabilities do appear in the original PDE model  the forced NLS  and correspond to a new type of solutions of this extensively studied equation [R24]. As a blind search for such solutions is a formidable numerical task, our analysis is pivotal for identifying what are the interesting parameter values and initial conditions for which new types of solutions  like the parabolic resonant ones  appear.
NEWS:
1. In [R32], we showed that the parabolic resonance mechanism can even cause spatial decoherence of small amplitude nearly flat initial data.
2. We have classified the structure of the unpertubed energy surfaces of a "generic" class of nearintegrable twodegrees of freedom systems [R31] . This classification shows that there are some mechanisms of instability in twod.o.f. systems that have not been studied yet!
3. In [R36] we developed a framework to study the near integrable PDE structure. This paper was selected to be in the Nonlinearity Highly downloaded Collection for 2010.
4. In [R35] we provide a framework for studyng analytically the parabolic resonance instability by using the adiabatic chaos framework. Most importantly, we show that the volume of the chaotic zone depends on the perturbation parameter as a power law, thus it is numerically observable.
[R35]  V. RomKedar and D. Turaev,  The symmetric Parabolic Resonance Instability,  Nonlinearity 23 13251351, 2010. 
[R36]  E. Shlizermann and V RomKedar,  Classification of solutions of the forced periodic nonlinear Schr\"{o}dinger equation,  Nonlinearity 23 (9), 21832218, 2010. 
[R32] 
E.
Shlizermann
and V. RomKedar; 
Parabolic
Resonance: 
Physical
Review Letters, 102, 033901, 2009. 
[R31]  M.
Radnovic and V. RomKedar 
Foliations of Isonergy Surfaces and Singularities of Curves  Regular and Chaotic Dynamics, Vol. 13, No. 6, pp. 645–668, J. Moser80 memorial issue, 2008. 
[R24] 
E. Shlizermann
and V RomKedar 
Three types of chaos in the forced nonlinear Schrodinger
equation. 
Physical Review Letters, 96, 024104,
2006. 
[R23] 
E. Shlizermann
and V. RomKedar 
Hierarchy of bifurcations in the truncated
and forced NLS model 
Chaos 15(1) 013107,
22 pp, 2005. 
[R9] 
V. RomKedar 
Chaos, 7(1):148158, 1997. 

[R22] 
A. LitvakHinenzon
and V. RomKedar 
SIAM J. Appl.
Dyn. Syst. 3(4), 525—573, 2004. 

[R11] 
V. RomKedar and N.
Paldor 
Bull. Amer. Mete. Soc., 78(12):27792784, 1997. 

[R12] 
V.RomKedar, Y. Dvorkin, and N. Paldor;

Chaotic
Hamiltonian dynamics of particle's
horizontal motion in the atmosphere. 
Physica D, 106(34):389431, 1997. 
[RP5] 
V. RomKedar; Y.
Dvorkin, and N.Paldor 
In M.F. Shlesinger,
Zaslavsky, G.M.thinspace
and U.Frisch, editors, Proc. on Le'vy Flights and Related Phenomena, Lecture
Notes in Physics, pages 7287. SpringerVerlag,
1995. 

[RP6] 
V. RomKedar 
Some characteristics of
twodegreesoffreedom Hamiltonian flows. 
In Doveil Benkadda and Elskens,
editors, Transport, Chaos and Plasma Physics, II, Proc., Marseille,
pages 164178, 
[RP7] pdf Abstract 
V. RomKedar 
Transport in a class of nd.o.f. systems. 
In Hamiltonian systems with three or
more degrees of freedom (S'Agar'o,
1995) , pages 538543. Kluwer Acad. Publ., 
[RP9] pdf Abstract 
A. LitvakHinenzon
and V. RomKedar 
Parabolic resonances in near integrable Hamiltonian systems. 
In D.S. Broomhead,
E.A. Luchinskaya, P.V.E. McClintock, and T.Mullin, editors, emStochaos:
Stochastic and Chaotic Dynamics in the Lakes, pages 358368.
American 
[R19] pdf Abstract 
A. LitvakHinenzon
and V. RomKedar 
Resonant tori
and instabilities in Hamiltonian systems. 
Nonlinearity, 15(4):11491177, 2002. 
[R20] pdf Abstract 
A. LitvakHinenzon
and V. RomKedar 
Parabolic resonances in 3 degree of freedom
nearintegrable Hamiltonian systems 
Physica
D, 164(34):213250, 2002. 
[RP12] pdf Abstract 
E. Shlizermann
and V. RomKedar 
Characterization of Orbits in the Truncated
and Forced Nonlinear Shrodinger Model. 
ENOC, 2005 
[RP11] pdf Abstract 
E. Shlizerman
and V. RomKedar 
Energy surfaces and hierarchies of bifurcations  instabilities in the forced truncated NLS. 
In P. Collet
et al. editors, Chaotic Dynamics and Transport in Classical and Quantum
Systems. Kluwer Academic Press in NATO
Science Series C, 2004. 
Modeling
and Medicine: models of the hematopoesis system:
with R. Malka,
E. Shochat
MD/PhD, and Prof. L. Segel (deceased,
2005).
Main
motivation
and results: Why do GCSF (Granulocyte Colony Stimulating Factor)
injections
help some neutropenic patients (patients
with very
low counts of this kind of white blood cells) and not others?
In [R27] we employ dynamical
systems approach to develop a simple model (not Hamiltonian) that is
adequate
for describing the natural dynamics of the neutrophils
and the GCSF levels in the blood on a time scale of several days. In
[R30] we
use a variant of this model to study the response of patients that
received
chemotherapy and thus became neutropenic
to various
protocols of GCSF support.
More precisely, in [R30] we built axiomatically a mathematical model that describes the neutrophils’ dynamics. We showed that patients fall into three groups according to their blood counts, and that these groups respond differently to the growth factor therapy. The first group responds well. The current treatment is ineffective for the second group, but our results imply that a change in the administration protocol may help. In the third group, treatment with growth factors alone is insufficient at any dose, and the patients need an alternative therapy. These conclusions stem directly from the mathematical model, which makes it possible to predict outcomes and improve treatment. While the immune system is exceedingly complex, the model is simple, and that’s precisely its strength. We are proposing to test these recommendations in clinical trials. If these succeed, treatment efficiency would increase, and perhaps some lives could be saved.
NEWS:
We study the development of bacterial infections under neutropenic conditions by studying the neutrophilsbacteria interaction invitro [R33] and by combining this study with the blood neutrophils dynamics [R39].
[R42]  Roy Malka, Baruch Wolach, Ronit Gavrieli, Eliezer Shochat and Vered RomKedar  Evidence for bistable bacterianeutrophil interaction and its clinical implications  J. Clin Invest. doi:10.1172/JCI59832, 2012. See also commentary . 
[R39]  R. Malka and V. RomKedar  BacteriaPhagocytes Dynamics, Axiomatic Modelling and MassAction Kinetics,  Mathematical Biosciences and Engineering, 8(2), 475502, 2011. 
[R33]  R. Malka R, E. Shochat, V. RomKedar;  Bistability and Bacterial Infections.  PLoS ONE 5(5): e10010. doi:10.1371/journal.pone.0010010, 2010. 
[R27] pdf
Abstract 
E. Shochat, V.
RomKedar and L. Segel 
Bull. Math. Biology 69(7), 22992338, 2007. 

[R30] pdf
Abstract 
E. Shochat and
V. RomKedar 
Clinical Cancer Research 14, 63546363, October 15, 2008 2008. 
Chaotic
fluid mixing :
with R. Aharon, D. F. Carlson, E. Fredj, H. Gildor, R. Mundel
A. Poje,
A. Leonard and S. Wiggins.
Main
motivation and
results:
The field of chaotic advection is concerned with the study of mixing of
passive
scalars in fluids with specified velocity fields that are
timedependent yet
nonturbulent. Since the important paper by Aref
in
1984, it attracts a growing community of researchers coming from
diverse areas
of science and engineering and involves both practical problems and
beautiful
mathematical questions.
In my
Ph.D. studies we
considered
time periodic planar flows (the OVP model) and introduced several
dynamical
systems concepts and tools to this field; We identified the unstable
manifold
and KAM tori as the observed objects in
flow
visualizations and as controllers of transport, discovered lobe
dynamics, and
established the direct relation between lobe dynamics and fluid
transport
[R3,RP1]. These ideas influence theoretical and experimental studies of
fluid
mixing problems arising in geophysical applications, traditional lab
scale
experiments and microfluidics (e.g. these
findings
became the main topic of S. Wiggins popular book "Chaotic Transport in
Dynamical systems").
In [R16] we proved that for a large class of twodimensional timeperiodic fluid flows some mixing characteristics have a universal dependence on the flow frequency even in the nonperturbative regime. These results were realized experimentally (see e.g. Horner et al. JFM 02). In [R36] we show that upper bounds on eddy diffusivity may be found by combining LCS computations from RF data and airial photos of separating lines.
In [41] we show that minute 3d convection can untangle the chaotic tangles of the surface mixing.
In [47] we show that extreme values of observables
provide simple, intuitive,
and computationally cheap diagnostics for fluid mixing; Moreover, they
also enable a significant data reduction, since it is possible to
extract from their cumulative distribution functions much of the
relevant information regarding the existence, location, size and motion
of the coherent structures.
R. Mundel, E. Fredj, H. Gildor and V. RomKedar,  "New Lagrangian diagnostics for characterizing fluid flow mixing"  
[R41]  Rotem Aharon, Vered RomKedar, and Hezi Gildor,  "When complexity leads to simplicity: ocean surface mixing simplified by vertical convection"  Phys.
Fluids 24,
056603 (2012); \http://dx.doi.org/10.1063/1.4719147. ON THE COVER. 
[R36]  D. F. Carlson, E. Fredj, H. Gildor, V. RomKedar  Deducing an upper bound to the horizontal eddy diffusivity using a stochastic Lagrangian model,  Environmental Fluid Mechanics, 15677419 (Print) 15731510 (Online), 2010. 
[R3] 
V. RomKedar, A. Leonard, and S. Wiggins 
An
analytical study of transport, mixing and
chaos in an unsteady vortical flow 
J. Fluid Mech., 214:347394, 1990. 
A. Leonard, V. RomKedar, and S.Wiggins 
proc. intl. conf. on the physics of chaos and systems far from equilibrium. In Nuclear Physics B (Proc. Suppl.), volume2, pages 179190, 1987 

[R16] 
V. RomKedar and A.C. Poje 
Universal
properties of chaotic transport in
the presence of diffusion 
Phys. of Fluids, 11(8):20442057, 1999. 
Transport
theory of periodically forced systems :
with A. Litvak Hinenzon and
G.M. Zaslavsky
Main
motivation
and results: The transport theory of periodically forced
Hamiltonian
systems is concerned with characterizing the motion of ensembles of
initial
conditions from one region in phase space to another.
In
[R2] we observed that the geometry of the unstable and stable manifolds
determines exactly all the transport rates between certain phase space
regions
for a general class of forced planar flows. This discovery of the
popular lobe
dynamics lead to a burst of theoretical and even some experimental
studies in
diverse fields of Physics ([R2] was generalized to various settings and
was
followed by numerous publications on its role in fluid mixing, chemical
reactions,
optics, astrophysics, ship stability, electroconvection
etc.). In subsequent works I developed perturbational
methods for calculating topological characteristics of these manifolds'
complex
geometry (called homoclinic tangles).
These studies
lead to analytical estimates of the transport rates for nontrivial
time
scales, to a lower bound on the topological entropy of homoclinic
tangles in open flows [R1,R7,R8], to analytical methods for detecting
multipulse solutions [R8] and to insights regarding the structure of
strange
attractors in slightly dissipative system [R10].
[R1] 
V. RomKedar 
Transport rates of a family of
twodimensional maps and flows 
Physica
D, 43:229268, 1990. 
[R2] 
V. RomKedar and S. Wiggins 
Transport in twodimensional maps 
Archive for Rational Mech. and Anal.,
109(3):239298, 1990. 
[R7] 
V.RomKedar 
Homoclinic
tangles  classification and applications. 
Nonlinearity, 7:441473, 1994. 
[R8] 
V.RomKedar 
Secondary homoclinic
bifurcation theorems 
Chaos, 5(2):385401, 1995. 
[R10] 
A. Litvak Hinenzon
and V. RomKedar 
Symmetry breaking perturbations and strange
attractors. 
Phys. Rev. E, 55(5):49644978, 1997. 
[R18] 
V. RomKedar and
G.M. Zaslavsky 
Islands of accelerator modes and homoclinic tangles.

Chaos, 9 (3):697705, 1999. 
[RP2] 
V. RomKedar 
Homoclinic
structures in open flows. 
In J.D. Fournier and P.L. Sulem, editors, emLarge Scale Structures in Nonlinear Physics, Proc., Lecture Notes in Physics, pages 5072, VillefranchesurMer, 1991. SpringerVerlag. 
[RP3] 
Y. Ponty, A. Pouquet, V. RomKedar, and P.L. Sulem 
Dynamo in a nearly integrable
chaotic flow. 
In M.R.C. Proctor, editor, Proc. NATO
ASI: Theory of Solar and Planetary Dynamos, pages 241248, 
[RP4] 
V. RomKedar 
The topological approximation method. 
In Doveil Benkadda and Elskens,
editors, Transport, Chaos and Plasma Physics, Proc. , pages 3957,
Marseille, 1994. World Scientific 
[RE1] 
V. RomKedar and
G.M. Zaslavsky 
Chaotic kinetics and transport (overview). 
Chaos, 10(1):12, 2000. Focus Issue:
Chaotic kinetics and Transport, eds V.
RomKedar and G. Zaslavsky. Including
papers from the Workshop on Nonlinear Dynamics: Chaos, Transport and
Transition to Turbulence held at the New York University, New York,
November 1998. 
[RP10] 
V. RomKedar 
Frequency spanning homoclinic
families 
Commun.
Nonlinear Sci. Numer.
Simul., 8(34):149169, 2003.
Chaotic transport and complexity in classical and quantum dynamics. 
Exponentially
small splitting :
[R5] 
C.Amick, S.C.E. Ching, L.P. Kadanoff,
and V.RomKedar 
Beyond all orders: Singular perturbations in
a mapping 
J. Nonlinear Sci., 2:967, 1992. 
[R6] 
V.RomKedar,
L.P. Kadanoff, S.C.E. Ching,
and C.Amick 
The breakup of a heteroclinic
connection in a volume preserving mapping. 
Physica D, 62:5165, 1993 
[R13] 
S. Pekarsky and V.
RomKedar 
Uniform stochastic web in lowdimensional
Hamiltonian systems. 
Phys. Lett. A, 225:274286, 1997. 
[R14] 
S. Pekarsky and V.
RomKedar 
Degeneracies and
instabilities of the motion of a charged particle  the 4d webmap
model. 
Nonlinearity, 10(4):949963, 1997. 