Computations in Science Seminars

January 4, 2012: Leo Kadanoff, University of Chicago; e-mail:; Innovations in Statistical Physics; In 1965-71, a group of people, myself included, formulated and perfected a new approach to physics problems, under the names of scaling, universality and renormalization. This work became the basis of a wide variety of theories ranging from particle physics and relativity, through condensed matter physics, and into economics and biology.; This work was of transcendental beauty and of considerable intellectual importance.; This left me with a personal problem. What next? Constructing the answer to that question would dominate the next 45 years of my professional life.; * The most important work came from Cyril Domb, Michael Fisher, Benjamin Widom, A. Pashinski, V. Pokrovskii, and Kenneth Wilson.
January 11, 2012 (^): Taehun Lee, City College of New York; Faculty contact: Wendy Zhang,; High-order Lattice Boltzmann Simulations of Drops and Bubbles; The lattice Boltzmann method (LBM) is a mesoscale approach, which can accommodate coarse-grained, molecular-level information into the macroscopic description of complex interfacial phenomena. This is achieved by introducing a phase field function into a single-phase lattice Boltzmann formulation to distinguish between phases (i.e. liquid/vapor, liquid/liquid), together with a phenomenological free energy functional of the solid-liquid-vapor system whose dissipative minimization constrains the temporal evolution of the phase field. LB equation is generally derived from the discrete Boltzmann equation by discretizing it on uniform rectangular mesh and usually comprises collision and streaming steps. While this greatly facilitates numerical procedure, it limits shapes of the computational domain that LBM can be applied to. This limitation could substantially increase computational effort for flows of boundary-layer type and in complex geometries with strong interactions between solid surface and contact line. To overcome geometric constraint of LBM and to improve its numerical stability at high Reynolds number, we have recently proposed high-order Galerkin/Discontinuous Galerkin Spectral/Finite Element LBM. In these computational frameworks, LB equation is regarded as a special space-time discretization of the discrete Boltzmann equation in the characteristic direction, and is solved by higher-order accurate schemes on unstructured mesh. In this presentation, a brief introduction to the temporal and spatial discretizations of the discrete Boltzmann equation will be given, with emphasis on the Galerkin/Discontinuous Galerkin approximations on unstructured mesh. Applications of the new LBM will be discussed in the simulations of single- and two-phase flows including flow past a cylinder, drop coalescence, and drop impact on thin liquid layer and flat/heterogeneous substrates.
January 18, 2012: Jack Cowan, University of Chicago; e-mail: $email address$ , Faculty contact: Leo Kadanoff,; Stochastic Wilson-Cowan equations for networks of excitatory and inhibitory neurons; We have recently found a way to describe the large-scale statistical dynamics of neocortical neural activity in terms of (a) the equilibria of the mean-field Wilson-Cowan equations, and (b) the fluctuations about such equilibria due to intrinsic noise, as modeled by a stochastic version of such equations. Major results of this formulation include a role for critical branching, and the demonstration that there exists a nonequilibrium phase transition in the statistical dynamics which is in the same universality class as directed percolation (DP). Here we show how the mean-field dynamics of interacting excitatory and inhibitory neural populations is organized around a Bogdanov-Takens bifurcation, and how this property is related to the DP phase transition in the statistical dynamics. The resulting theory can be used to explain the origins and properties of random bursts of synchronous activity (avalanches), population oscillations (quasi-cycles), synchronous oscillations (limit-cycles) and fluctuation-driven spatial patterns (quasi-patterns). If time permits we will also show how such a system of interacting neural populations can be made to self-organize to a state near the Bogdanov-Takens bifurcation, if the coupling constants (synaptic weights) are activity-dependent, and follow the rules of spike-time dependent synaptic plasticity (STDP).
January 25, 2012: Greg Grason, Umass Amherst; e-mail: , Faculty contact: David Biron,; Topological Defects in Perfect Packings of Twisted Filament Bundles; In this talk, I will discuss recent progress in our understanding of a fundamental and non-linear coupling between in-plane order and out-of-plane geometry in twisted assemblies of filamentous molecules, key structural motifs in cells and living tissues. Not unlike the coupling of in-plane stresses and out-of-plane geometry in thin elastic sheets, we find that certain textures of filament tilt generate intrinsic stresses that frustrate the cross-sectional packing of bundles. Surprisingly this problem is formally akin to crystalline order on spherically-curved membranes, and sufficiently twisted textures of filament bundles favor the incorporation of one or more topological defects in the otherwise regular cross section of the bundle. Based on the non-linear continuum theory of filament arrays, we explore the complex spectrum of topological defects -- disclinations, dislocations and grain boundaries -- that thread through the ground-state packing of these materials and show that the structure of these highly-irregular packings is primarily governed by two geometric parameters relating to the degree of twist and the lateral size of bundles. Finally, I discuss a simple approach to exposing the hidden, non-Euclidean geometry of twisted bundles that underlies the frustration of in-plane packing in these materials.
February 1, 2012: John Reppy, Cornell; e-mail: , Faculty contact: Leo Kadanoff,; Does the Supersolid Exist ?; The possible existence of a Bose-condensed supersolid state in solid 4He was first suggested over 40 years ago by Chester, Andreev and Lifshitz, and Leggett. The first experimental evidence for the supersolid has appeared only recently with the torsional oscillator experiments of Kim and Chan where they observed, at low temperature, a decrease in the period of an oscillator containing a solid 4He sample. They interpreted this decrease in period as superfluid-like decoupling of a fraction of the moment of inertia of the 4He sample from the motion of the torsional oscillator. This discovery created great excitement in the low temperature community and has been followed by a flurry of activity in many laboratories around the world. The Kim-Chan discovery was followed closely by the observation by Day and Beamish of an anomalous increase in the shear modulus of solid 4He in the same temperature range as the supersolid observation. It has developed that these two phenomena share many common features and appear to be closely related. The possibility exists that the period shifts seen in the torsional oscillator experiments may, at least in some cases, be a consequence of the shear modulus anomaly. At Cornell, we have constructed multiple-frequency torsional oscillators in an attempt to delineate the effects of the supersolid and shear modulus anomaly. This approach takes advantage of the expectation that the supersolid phenomenon, as in the case for superfluidity in liquid 4He, is relatively insensitive to frequency, while the effect of changes in the effect on the oscillator period will have pronounced frequency dependence. These measurements are currently in progress, however, we have been able to establish that in certain cases the shear anomaly is can produce period shift and dissipation signals identical in form to the supersolid signals reported by Kim and Chan. J.D.R would like to acknowledge the collaboration of Xiao Mi and Erich Mueller in the work and also the support of the NSF through Grants DMR-06586, PHY-0758104, and CCMR Grant DMR-0520404.
February 8, 2012: Yossi Cohen, The Weizmann institute; e-mail: , Faculty contact: Wendy Zhang,; The dynamics of crack in torn thin sheets; The prediction of crack path as it propagates into a brittle material is one of the main challenge in fracture mechanics. In the framework of Linear Elasticity Fracture Mechanics, the most important information governing the dynamics of crack growth can be found in the stress eld close to its edge. In the vicinity of the crack, the stress eld under a mode III shear tearing of a thin plate has a universal form but with a non-universal amplitude known as the Stress Intensity Factor. All the non-universal aspects of the stress distribution are collected in the Stress Intensity Factor which depends on everything, including the crack length, the boundary conditions and the history of the loads that drive the crack evolution. Although the equations of elasticity for thin plates are well known, there remains the question of selection of a path for a propagating crack. We invoke a generalization of the principle of local symmetry to provide a criterion for path selection and demonstrate the qualitative agreement of our results with the experimental ndings. We also ana- lyze the nature of the singularity at the crack tip with and without the nonlinear elastic contributions. Finally we present an exact analytic results for the stress intensity factor to the linear approximation for the crack developing in thin sheets.
February 15, 2012 (^) Hosted by Margaret Gardel: Fred MacKintosh, Vrije University, Amsterdam; e-mail:; Mechanics and dynamics of fiber networks: criticality, mechanical integrity; Much like the bones in our bodies, the cytoskeleton consisting of filamentous proteins largely determines the mechanical response and stability of cells. These biopolymers form fiber networks, whose mechanical stability relies on the fibers' bending resistance, in contrast to rubbers that are governed by entropic stretching of polymer segments.Thus, the elastic and dynamic properties of such semi-flexible polymers are very different from conventional polymeric materials. We show that these networks exhibit both a low-connectivity rigidity threshold governed by fiber bending, as well as a high-connectivity threshold governed by fibre-stretching elasticity. We show that the latter exhibits rich zero-temperature critical behavior, including a crossover between various mechanical regimes along with diverging strain fluctuations and a concomitant diverging correlation length. Inspired by both intra- and extracellular networks, we describe recent theoretical modelling and experiments on simplified fiber networks in vitro. Among the more striking material properties of these networks is their nonlinear elasticity, with a strong stiffening response to stress. Unlike passive materials, however, living cells are kept far out of equilibrium by metabolic processes and energy-consuming molecular motors that generate forces to drive the machinery behind various cellular processes. We show how such internal force generation by motors can lead to dramatic mechanical effects, including a strong stiffening of cytoskeletal networks. Furthermore, stochastic motor activity can give rise to diffusive-like motion in elastic networks, as has been observed in living cells.
February 22, 2012 (^): Charles Sykes, Tufts University; e-mail: , Faculty contact: David Biron,; Turning a Single Molecule into an Electric Motor; In stark contrast to nature, current manmade devices, with the exception of liquid crystals, make no use of nanoscale molecular motion. In order for molecules to be used as components in molecular machines, methods are required to couple individual molecules to external energy sources and to selectively excite motion in a given direction. Significant progress has been made in the construction of molecular motors powered by light and by chemical reactions, but electrically-driven motors have not been demonstrated yet, despite a number of theoretical proposals for such motors. Studying the rotation of molecules bound to surfaces offers the advantage that a single layer can be assembled, monitored and manipulated using the tools of surface science. Thioether molecules constitute a simple, robust system with which to study molecular rotation as a function of temperature, electron energy, applied fields, and proximity of neighboring molecules. A butyl methyl sulphide (BuSMe) molecule adsorbed on a copper surface can be operated as a single-molecule electric motor. Electrons from a scanning tunneling microscope are used to drive directional motion of the BuSMe molecule in a two terminal setup. Moreover, the temperature and electron flux can be adjusted to allow each rotational event to be monitored at the molecular-scale in real time. The direction and rate of the rotation are related to the chiralities of the molecule and the tip of the microscope (which serves as the electrode), which illustrates the importance of the symmetry of the metal contacts in atomic-scale electrical devices.
February 29, 2012: Anette (Peko) Hosoi, MIT; e-mail: , Faculty contact: Leo Kadanoff,; Low temperature solvent annealing in organic thin films; We examine solidification in thin liquid films produced by annealing amorphous films in a solvent vapor. Micrographs captured during annealing reveal the nucleation and growth of single-crystal needles. The needle lengths scale like power laws in time where the growth exponent depends on the thickness of the deposited film. The evolution of the thin film is modeled by a lubrication equation, and an advection-diffusion equation captures the transport of material and solvent within the film. We define a dimensionless transport parameter which describes the relative effects of diffusion and coarsening-driven advection. For large values of this parameter, needle growth matches the theory of 1D, diffusion-driven solidification. For low values, the collapse of droplets -- i.e. coarsening -- drives flow and regulates the growth of needles. Within this regime, we identify and analyze two asymptotic limits: needles that are small compared to the typical drop size, and those that are large.
March 7, 2012 (^): Elisha Moses, Weizmann Institute; e-mail: , Faculty contact: David Biron,; Computing with living neuronal networks; Neurons explanted from the brain will grow on the bottom of a dish and form a highly connected, electrically active neural network. We show that its computational abilities are determined by collective effects in a new kind of percolation system, and are limited due to random connectivity. Geometrical guidance, along with redundancy and multiplexing, reproduce some minimal yet reliable computation functions of the network.
March 14, 2012: Robert Schroll, University of Massachusetts; e-mail: , Faculty contact: Wendy Zhang,; The wrinkling behavior of highly bendable thin sheets; The behavior of a thin elastic sheet can be characterized by its 'bendability', a number that compares bending and stretching forces applied to the sheet. Highly bendability sheets are so thin that bending energies are essentially negligible. Because of this, highly bendable sheets wrinkle easily when subject to confinement. Traditionally, such wrinkles have been described by a 'post-buckling analysis' that describes the wrinkled state as a perturbation of the flat, un-buckled state. We argue that this is inappropriate for highly bendable sheets, since wrinkles are able to reduce the compressive stress essentially to zero. Instead, a 'far-from-threshold' analysis, in which wrinkles are treated as a singular perturbarion of a collapsed compressive stress state, must be used. A simple planar problem is used to illustrate this method, which is then used to analyze the behavior of thin sheets on liquid drops. Experiments demonstrate the need for this far-from-threshold analysis, but they reveal additional unexpected behavior.
March 28, 2012: Michael Rust, University of Chicago; e-mail: , Faculty contact: David Biron,; It's about Time: A Three-Protein Clock from Photosynthetic Bacteria; Despite being composed of molecular components subject to intense thermal fluctuations, living cells routinely display strikingly precise and coherent behavior. A recently discovered example of this phenomenon is a ~24-hour oscillator found in the photosynthetic cyanobacterium Synechococcus elongatus. In a realistic environment, this oscillator is phase-locked to the daily rhythms of light and dark experienced by the organism, but precise oscillations will continue even if the organism is deprived of rhythmic cues. Thus, it is similar in function to the circadian clocks found in animals and plants, familiar to anyone who has become jet-lagged following a cross-country flight. Surprisingly, three purified protein components from this organism, KaiA, KaiB and KaiC, can be mixed in a test tube with ATP to reconstitute stable biochemical oscillations outside of the cell. Though the phase of this oscillation can be quite responsive to the environment, the period remains close to 24 hours over a broad physiological range of temperatures, protein concentrations and nucleotide conditions. I will describe a combination of biochemical experimental work and dynamical systems analysis in our attempt to understand both the emergence of robust oscillations and phase shifting in this simple system.
April 11, 2012: Sonja Schmid, Virginia Tech; e-mail: , Faculty contact: Leo Kadanoff,; Choosing the Right Reactor for the Job: Chernobyl, Fukushima, and Beyond; When Soviet nuclear scientists and engineers developed the RBMK (the 'Chernobyl-type' reactor), they were convinced that this design surpassed its rivals in every regard: it was easy to assemble, economical, and so safe it didn't need a containment structure. An experienced pool of operators was ready, having been trained on the RBMK's military and dual-use cousins. The first RBMK was built less than 50 miles from Leningrad, and by 1986 fourteen more were up and running, delivering a total of 15.5 GWe. After the Chernobyl disaster, critics portrayed the RBMK as technically flawed, incompetently operated, and part of a corrupt, mismanaged industry. Very similar charges were mounted in the aftermath of the Fukushima disaster: concerns about the Mark I containment allegedly date back to the early 1970s; Tepco, the utility, had attempted to cover up safety violations in the past; and the Japanese nuclear industry in general maintained too cozy a relationship with the regulatory agency. Poised to learn the lessons of Fukushima, the U.S. government has ordered a safety review of American reactors and simultaneously granted a license to a new, inherently safe reactor design; several more designs are under review. This talk will discuss what we would gain from taking the long view instead of focusing only on the immediate aftermath of a serious accident. I will argue that safety is more than technical reliability, and that it needs to be understood in the context of complex, messy historical, organizational, and cultural dimensions that defy standardization. Finally, the talk will raise a few related questions about the role of small modular reactors for our future energy policy.
April 18, 2012: Ariel Amir, Harvard; e-mail: , Faculty contact: William Irvine,; Relaxations in glasses - full aging and beyond; Glassy systems are ubiquitous in nature, from window glasses, through the anomalous magnetic properties of spin-glasses, to memory effects observed in electronic systems. Among their key properties are slow relaxations to equilibrium without a typical timescale and aging, the dependence of relaxation on the system's age. Understanding these phenomena is a long-standing problem in physics. In this talk I will show that the particular example of electron glasses is a useful case study to understand the generic mechanisms involved, leading to aging. I will describe our approach to the problem, and show that it generally leads to a particular form of aging, which we found to agree well with data on electron glasses, as well as various other systems such as disordered semiconductors and structural glasses. I will also show results on the expected deviations from the universal form, and what we think can be learnt from them.
April 25, 2012: Humphrey Maris, Brown University; e-mail: , Faculty contact: Leo Kadanoff,; Experiments with Electrons in Superfluid Helium; In this talk I will describe experiments to study the properties of electrons immersed in liquid helium. By using an ultrasonic technique it has been possible to make movies showing the motion of individual electrons. I will describe the details of the experiments and show the results that have been obtained.
May 2, 2012 (^): Cristina Marchetti, Syracuse; e-mail: , Faculty contact: William Irvine,; Collective dynamics of active matter: from self-propelled particles to migrating cell layers.; Bacterial suspensions, extracts of cytoskeletal filaments and motor proteins, and cell colonies are examples of assemblies of interacting self-driven units that form a new type of active soft matter with intriguing collective behavior. In this talk I will discuss the theoretical modeling of active systems. Specific examples will include bacterial swarming and the collective migration of confluent layers of epithelial cells that have been shown to exhibit glassy dynamics at high density.
May 3, 2012 (^) Special seminar time: Thursday 3 p.m. at CIS EB041: Marc Fermigier, PMMH-ESPCI; Faculty contact: Wendy Zhang,; Windblown droplets; This is an experimental study on the conditions in which an airstream can displace liquid drops deposited on a solid surface. We vary the size of the drops, the wettability conditions and the viscosity of the liquid. The drift speed of the droplets is interpreted by balancing the viscous dissipation within the liquid and the work of the aerodynamic force. Larger drops, flattened by gravity, have drift speeds different from small droplets which remain nearly spherical.
May 9, 2012 (^): Vinothan Manoharan, Harvard; e-mail: , Faculty contact: William Irvine,; A particle walks into an interface...; When small solid particles encounter liquid interfaces, they can assemble into a variety of structures, including crystals, clusters, and gels. But the dynamics of assembly and the interactions that drive it are still not well-understood. We use digital holographic microscopy and confocal microscopy to directly observe colloidal particles in the early stages of self-assembly. These experiments have revealed unexpected dynamics in seemingly simple phenomena, such as the binding of a single colloidal particle to an interface. We find that a particle takes a surprisingly long time -- weeks or even months -- to relax to equilibrium. This behavior can be understood in terms of a dynamic wetting mechanism involving thermally-activated hopping of the contact line over surface defects. The results call into question the validity of models of colloidal interactions that assume the particles have reached equilibrium with the interface. They also suggest new ways to control these interactions and the resulting self-assembled structures.

(&) : When Wendy Zhang is unavailable for the seminar.

(^) : When Leo Kadanoff is unavailable for the seminar.

(#) : When David Biron is unavailable for the seminar.

(*) : When William Irvine is unavailable for the seminar.

Back to JFI home page.