Invited talks

Contributing talks

Posters in the webpage

         Invited talks

Prof. C.K. Hu               (Institute of Physics,  Academia Sinica, Taipei 11529,  Taiwan)

e-mail: huck@phys.sinica.edu.tw

trip: Hangzhou, 4 Nov. 12:05 JL 635M, 8 Nov. 13:15 JL 636M, a double room

Universality, scaling, and super-scaling in critical systems

Universality and scaling are important concepts in the study of critical systems. Here I will briefly review our works on universality, scaling, and superscaling in critical systems. Using the histogram [1] and other Monte Carlo methods, we have found that quite different two dimensional percolation models (including bond and site percolation on three regular lattices, bond percolation on random lattices and their duals, and continuum percolation of soft disks and hard disks) have universal finite-size scaling functions [2] as far as the domains of the studied systems have the same aspect ratios and boundary conditions [3]. Similar results have been found for the Ising model on planar lattices [4].

Very recently, Watanabe, Yukawa, Ito, and myself [5] considered percolation on a RL*L two-dimensional rectangular domains with width L and aspect ratio R. We propose that the existence probability of percolating cluster E_p(L,ε,R) as a function of L, R, and deviation from the critical point ε can be expressed as F(εL^{y_t}R^a), where y_t = 1/ν is the thermal scaling power, a is a new exponent, and F is a scaling function. We use Monte Carlo simulation [1] of bond percolation on square lattices to test our proposal and find that it is well satisfied with a=0.14(1) for R > 2. We also propose super-scaling for other critical quantities, which is confirmed for the percolation probability. Finally, I discuss some problems for further studies. 

[1]  C.-K. Hu, Phys.Rev.Lett. 69, 2739 (1992); Phys. Rev. B46, 6592 (1992).
[2] V. Privman and M. E. Fisher, Phys. Rev. B30, 322 (1984).
[3]  C.-K. Hu, C.-Y. Lin, and J.-A. Chen, Phys.Rev.Lett.75, 193 (1995)
     and 75, 2786(E) (1995), Physica A221, 80 (1995); C.-K. Hu and C.-Y. Lin,      Phys.Rev.Lett.77, 8 (1996); C.-K. Hu and F.-G. Wang, J. Korean Physical       Soc. 31, S271 (1997); H. P. Hsu, S. C. Lin and C.-K. Hu, Phys. Rev.
     E64, 016127 (2001); H. Watanabe, et al., J. Phys. Soc. Japan. 70, 1537      (2001).
[4] Y. Okabe, et al. Phys. Rev. E59, 1585 (1999); Y. Tomita, Y. Okabe, and
    C.-K. Hu, Phys. Rev. E60, 2716 (1999); C.-K. Hu, J.-A. Chen, and
    C.-Y. Lin, Physica A266, 27 (1999).
[5] H.  Watanabe, S. Yukawa, N. Ito, and C.-K. Hu, Phys. Rev. Lett.93,
    in press (2004).

Prof. D.P. Landau   (Center for Simulational Physics, The Univ. of Georgia, Athens, GA 30602, U.S.)

e-mail: dlandau@hal.physast.uga.edu

trip: Shanghai, 28 Oct ALL NIPPON 921G 20:15, 11 Nov ALL NIPPON 960G 8:55

PHOTO

Monte Carlo Simulations of Compressible Ising Models
　

D. P. Landau
Center for Simulational Physics
The University of Georgia, Athens, GA, U.S.A
         and
The Institute for Modern Physics
Zhejiang University, Hangzhou, China
　

Compressible Ising models have long been of interest because of the question of how the Ising transition is influenced by elastic degrees of freedom.  A careful Monte Carlo study showed, somewhat surprisingly, that the Ising ferromagnet at constant pressure had mean field critical behavior.  A recent theoretical investigation has predicted that the effect of elastic interactions depends upon whether the pressure or the volume is held fixed and whether the interactions are ferromagnetic or antiferromagnetic.  We shall describe extensive Monte Carlo simulations of compressible Ising models under all four sets of conditions and present a status report on our understanding of phase transitions in such systems.

Prof. Y. Okabe           (Department of Physics, Tokyo Metropolitan University,
                           1-1 Minami-osawa, Hachioji, Tokyo, Japan 192-0397)

e-mail:  okabe@phys.metro-u.ac.jp

trip: Hangzhou, 1 Nov NH929 12:50, 6 Nov NH930 13:50,   Lily hotel

Study of antiferromagnetic Potts model with polarized field

We study the antiferromagnetic three-state Potts model on the square
lattice.  We use the efficient Monte Carlo algorithm to study
the energy density of states directly.  We pay special attention to
the effect of staggered polarization field.  We determine two phase
boundaries; one of them belongs to the ferromagnetic three-state Potts
criticality, and the other to the Ising type.
　

Prof. J.-S. Wang       (Department of Computational Science, National University of Singapore,
                                   SOC1 #07-21, 3 Science Drive 2, Singapore 117543)

e-mail: wangjs@cz3.nus.edu.sg

trip: Hangzhou, 3 Nov 12:20 CA158, 7 Nov 19:30 CA157,   Lily hotel

Replica Monte Carlo Simulation

Jian-Sheng Wang

Department of Computational Science, National University of Singapore

We introduce the replica Monte Carlo simulation method for Ising spin-glasses.
The method combines the simultaneous simulation of many systems distributed
at different temperatures with cluster moves. In two dimensions, the replica
Monte Carlo is extremely efficient; in three or higher dimensions, it is
equivalent to replica exchange or parallel tempering.  We review its applications
and some recent results.
 

Prof. Peter Young      (Physics Department, University of California, Santa Cruz, CA 95064, USA)

e-mail: peter@bartok.ucsc.edu

Tel: (831) 459-4151,  Fax : (831) 459-3043

trip: Shanghai, United Airlines 0857, 19:30 on Nov. 3, to 13:45 on Nov. 8.

PHOTO

Recent Numerical Studies on Spin Glasses

Results of recent numerical studies of spin glasses will be reviewed. Topics

will include the phase transition in spin glasses with vector spins, and

investigations of the nature of the spin glass state by studying a

one-dimensional model with long-range interactions. A brief discussion of the

parallel tempering Monte Carlo method, which speeds up these simulations at

low temperature, will be also be given.

Prof. B.H. Wang       (University of Science and Technology of China)

e-mail: bhwang@ustc.edu.cn

trip: Hangzhou, 17:00 Nov.4 by HU7691, 8 November 2004

PHOTO

 The minority game of Boolean agent on network

Bing-Hong Wang^{1, 2},  Kan Chen^{2}  and Baosheng Yuan^{2}

         1 Department of Modern Physics, University of Science and Technology of China,

                      Hefei, Anhui, 230026, China

                 2   Department of Computational Science, Faculty of Science,

                      National University of Singapore, Singapore 117543

The dynamics of a complex adaptive system (CAS) can be influenced by the ways the elements

(node or agents) of the network are connected and the ways the elements are interacted. The study

of organization of real-world networks has attracted intensive research interest (Albert, Newman,

Dorogovtsev) ever since the seminal work of Barabasi and Albert. The degree distribution P(k),

which is defined as a probability of  finding a node with exactly k links, has been used as one of the

most essential variables in characterizing complex network. The power law or scale-free (SF) degree

distributions have been found exist in most real-world complex networks. (Barabasi, Strogatz).

We can understand the dynamics of a CAS system through the interactive behavior of competing

agents. One good example is the minority game (MG) model proposed by Challet and Zhang as

a simplification of Arthur's El Farol Bar attendance model (Arthur). The MG model turns out to

be a quite effective statistical tool in study of the dynamics of a multi-agent based system and it

has revealed some quite important characteristics of a system with competing agents, such as the

agent population distributions ( K.Chen, B.-H. Wang, and B. Yuan, Phys. Rev. E. 69, 025102

r.c. (2004) ), the phase structure and phase transition of utilization of the limited resources in

an evolutionary dynamical system (B Yuan, K Chen and B-H Wang, arXiv: cond-mat/0408391).

However in all these studies, the agents are not directly linked each other but they all respond to a

global environment collectively created by all the agents themselves.

One of the earliest study of network dynamics systematically is pioneered by Kauffman who

introduced NK random network and studied its Boolean dynamics. Recently there is renewed

interest in the network dynamics. Aldana and Cluzel demonstrated that the scale-free network

favors robust dynamics. Paczuski et al applied MG model in a random network to study the issue

of system self-organization and the dynamics of system state of order/disorder. Galstyan studied

the networked Minority Game focusing on the impact of different mean connectivity K of a random

network on the system coordination under different system capacities. Anghel et al used MG model

as a tool to investigate how an inter-agent communications across a network lead to the formation of

influence network. A very important question of how different network organizing principles affect

the system dynamics is however not fully explored in general. In this study, we try to address this

question with Boolean dynamics of network based Minority Game.

Here we ask a few questions: 1).how the utilization of the limited resources will be changed if

the agent's behavior is influenced by other agents in a connected network instead of by a global

environment. 2). how the system dynamics and the phase structure will be changed if an evolution

scheme is applied; and 3). how the different network organizations affect the dynamics of the

system. We try to explore these questions in this study with intensive numerical simulations. We

use MG model as a tool to study the Boolean dynamics under the framework of Kauffman Boolean

network. Different network organizations will be explored including growing random network, scale-

free network and a model of growing directed network with link-reversal we proposed recently.

We show that for an evolutionary Minority Game in the Kauffman random NK network the

critical value of mean connection number K is extended to include Kc = 3 in contrast to Kc = 2

in the well-known case of simple Boolean dynamics. An evolutionary approach has been applied

in the study of dynamical process of a CAS system where an evolution is an essential feature as

of the cases relevant to most real-world complex network systems. We also study the Boolean

dynamics in growing networks which are generated by adding agents successively in a way that

each new agent has K directed links to the existing ones, meaning the junior agents are influenced

by the senior ones. We find that the dynamics in such growing networks is generally stable and

independent of K, radically different from that of Kauffman NK random network. To better model

the real-world networks where there usually exists a small probability that an older node may be

influenced by a younger one, we allow a small percentage of link reversal in the growing directed

network. We show that the dynamics of the scale-free network with a small fraction of link reversal

seems to be operated on \the edge of chaos". Thus in a simple evolutionary Minority Game the

global coordination emerges in a very high order that its performance approaches the theoretical

bound most of times. We identify three key components for a high level coordination to emerge in a

directed network dynamics: growing and linear preferential attachment network, small percentage

of link-reversal, and evolutionary dynamics. We suggest these components be used as some essential

elements in studying dynamical and evolutionary behavior of many real-world complex network

systems.

Prof. X.S. Chen       (Institute of Theoretical Physics, Chinese Academy of Sciences,
                               P. O. Box 2735, Beijing 100080, China)

e-mail: chenxs@itp.ac.cn

trip: Hangzhou, 4 Nov 13:45 MU 5172, to 9 (?) Nov

Nonuniversal finite-size scaling in anisotropic systems

We study the bulk and finite-size critical behavior of the O(n)
symmetric \varphi^4 theory with spatially anisotropic
interactions of non-cubic symmetry in d<4 dimensions. In such
systems of a given (d,n) universality class, two-scale factor
universality is absent in bulk correlation functions, and
finite-size scaling functions including the Privman-Fisher scaling
form of the free energy, the Binder cumulant ratio and the Casimir
amplitude are shown to be nonuniversal. In particular it is shown
that, for anisotropic confined systems, isotropy cannot be
restored by an anisotropic scale transformation.
　

Prof. K. Chen       (Department of Computational Science, National University of Singapore,
                              SOC1 #07-21, 3 Science Drive 2, Singapore 117543)

e-mail: cscchenk@nus.edu.sg

trip: Hangzhou, 4 Nov 20:50, KA 622 (from Hongkong), 9 Nov 8:30, KA 623
      Lily hotel, a double room

        A Conditional Probability Measure of Volatility Clustering
                               in Financial Time Series.

We propose a measure of volatility clustering in financial time series.
This measure is based on the probability distribution of asset return,
conditioned upon the return in the previous time interval. We show that,
the conditional probability distribution, when properly scaled, collapses
to a universal curve with a power-law fat tail. The scale factor depends on
the return in the previous interval, and the dependence is almost linear at
large price fluctuations. We construct a stochastic dynamical model of
volatility clustering, which reproduces all essential features of
volatility clustering, as exhibited in the conditional probability
distribution of empirical data. The model illustrates a universal
stochastic process for both the formation of volatility clustering and the
emergence of fat tails in return distribution. The model also has a
potential application to option pricing, which is traditionally based on
the quite inadequate assumption that the price fluctuations follow a random
Gaussion process.
　

Prof. Z.R. Di        (Beijing Normal University, School of Management,
                              Department of Systems Science, Beijing, 100875, China)

e-mail: zdi@bnu.edu.cn

trip: 4 Nov 16:00   share a double room with DH Wang

PHOTO

Weighted Networks of Scientific Communication:

Empirical Study and Evolutionary Model

Menghui Li, Ying Fan, Peng Zhang, Jiawei Chen, Liang Gao, Zengru Di,

Department of Systems Science, School of Management, Beijing Normal University, Beijing, 100875, P.R. of China

Jinshan Wu

Department of Physics, Simon Fraser University, Burnaby, B.C. Canada, V5A 1S6

In order to take the weight of connection into consideration and to find a natural measurement of weight, we have collected papers in Econophysics and constructed a network of scientific communication to integrate idea transportation among econophysicists by collaboration, citation and personal discussion. Some basic statistics such as weight per degree are reported. Clustering coefficient of weighted network is introduced and empirically studied in this network. We also compare the typical statistics on this network under different weight measurements, including random and inverse weight. The conclusion from weight-randomized network is helpful to the investigation of the geometrical role of weight. Inspired by scientists’ collaboration network, an evolving model for weighted network is presented. The internal links are allowed and the repeating times of links is converted to the weight of the link. For a growing network and under the mechanism of preferential attachment, the network gives rise to power law distribution on degree and weight. An extended model and dissimilarity weight is introduced to give more detailed description for collaboration network. The generalized model also shows the scale-free phenomena in degree and weight distribution. The short term numerical results are consistent well with the empirical data in qualitative properties.

Prof. D.R. He    (Yangzhou University, College of Physics Science and Technology,
                                  Yangzhou 225002, China)

e-mail: drhe@yzu.edu.cn

trip: Hangzhou, 4 - 9 November 2004

PHOTO

Simulation on some complex networks with community structures

Da-Ren He

College of Physics Science and Technology
Yangzhou University, Yangzhou 225002, China

We briefly introduce our network description and computer simulation on some practical systems, such as traditional Chinese medicine, Chinese Huai-Yang recipe, Chinese railway transportation, bus transportation in Beijing and Yangzhou, Chinese power grid, and the economic development including trade activities, competitions, and evolutions of the open ports along Yangtze River. We want to show there are community structures in all these complex networks. It means that a node can only select part of the other nodes to   perform interaction according some rules. Some statistical results obtained by the simulation show very good agreement with practical statistical results.
　

Prof. Xiao Hu   (Computational Materials Science Center, National Institute for Materials Science,
                                Sengen 1-2-1, Tsukuba 305-0047, Japan)

e-mail: HU.Xiao@nims.go.jp
tel: +81-29-859-2661
fax: +81-29-859-2601

trip:  Hangzhou, 4 Nov. CA1710 21:50, Shanghai, 7 Nov. CA919 16:25
       Lily hotel

Physics of superconductivity vortex matter

Superconductivity is described by U(1) symmetry.  When a type II superconductor is immersed in a strong enough magnetic field, quantized magnetic inductions penetrate the system accompanied by circular super-currents, as predicted by Abrikosov in 1957 who got the Novel Prize last year. Since vortex lines align themselves into the triangular lattice, a second symmetry associated with the vortex lattice is involved.  Since the discovery of high-Tc superconductivity in oxide cuprates which is known extremely type II, the new field known as physics of vortex matter becomes very active and fruitful. One of the most notions is that the vortex lattice melts via a thermodynamic first-order phase transition, in sharp contrast to the mean-field theory by Abrikosov. This phenomenon raises fundamental issues in statistical physics which are still not yet settled completely.

A simple nevertheless good model for this system is given by the frustrated XY Hamiltonian, equivalent to the lattice London model.  It describes the phase variables of superconductivity order parameter and includes effect of magnetic field in terms of frustrations felt by the phases, which brings about singularities in phase distribution, i.e. vortices.  By performing large scale MC simulations we have clarified the properties of the first-order vortex lattice melting [1].  We also predict that there is a KT-like phase when a strong magnetic field is applied parallel to the superconducting layers, where vortex lines interact each other very anisotropically and feel strong layer pinning [2,3]. In the presence of point-like defects, the phase diagram of vortex state becomes very rich, in additional to the Bragg glass one probably sees vortex glass, vortex slush [4].  The dynamics of superconductivity vortices is also very interesting, which will be addressed in by Prof. Q. H. Chen [5] in this workshop.

References:

[1] X.Hu, S.Miyashita and M.Tachiki, Phys. Rev. Lett., 79, p.3498 (1997).
[2] X. Hu and M. Tachiki, Phys. Rev. Lett. 85, p.2577 (2000).
[3] X. Hu and M. Tachiki, Phys. Rev. B, 70, 064506 (2004).
[4] Y. Nonomura and X. Hu, Phys. Rev. Lett. 86, p.5140 (2001).
[5] Q. H. Chen and X. Hu, Phys. Rev. Lett., 90, 117005 (2003).

Prof. P.M. Hui    (Department of Physics, The Chinese University of Hong Kong
                         Shatin, New Territories,Hong Kong)

e-mail:   pmhui@sun1.phy.cuhk.edu.hk
http://www.phy.cuhk.edu.hk/people/teach/pmhui/pmhui.html

trip: Hangzhou, 4 Nov. to 8 Nov,   Lily hotel

Plateaux Formation and Abrupt Transitions
in a Competitive Population with Limited Resources

We study, both numerically and analytically, a Binary-Agent-Resource (B-A-R) model [1-3] consisting of N agents who compete for a limited resource 1=2 _ L=N < 1, where L is the maximum available resource per turn for all N agents [4]. The agents may or may not be networked [5] for information sharing. Detailed numerical simulations reveal that the system exhibits well-de_ned plateaux regions in the success rate which are separated from each other by abrupt transitions. As L increases, the maximum and the mean success rates of the agents over each plateau take on a constant value, e.g., the maximum success rate forms a well de_ned sequence of simple fractions. To understand the features analytically, it is important to study the outcome time series, the dynamics of the strategies' performance ranking pattern, and the dynamics of the system in the history space [6]. For example, the simple fraction representing the maximum success rate achievable by the agents for a given L corresponds to the fraction of 1's in the pattern of the outcome time series speci_ed by L. While the system tends to explore the whole history space due to its competitive nature, an increasing L has the e_ect of driving the system to a restricted portion of the history space. Thus the underlying cause of the observed features is an interesting self-organized phenomena in which the system, in response to the global resource level, e_ectively avoids particular patterns of history outcomes. We will also compare results in networked population [5] with those in non-networked population.

Work supported in part by the Research Grants Council of the Hong Kong SAR Government under CUHK4241/01P. Work done in collaborations [4] with H.Y. Chan and T.S. Lo of the Chinese University of Hong Kong and Neil F. Johnson of Oxford University.

[1] N.F. Johnson, P. Je_eries, and P.M. Hui, Financial Market Complexity (Oxford Univ. Press 2003).

[2] N.F. Johnson and P.M. Hui, e-print cond-mat/0306516.

[3] N.F. Johnson, P.M. Hui, D. Zheng, and C.W. Tai, Physica A 269, 493 (1999).

[4] H.Y. Chan, T.S. Lo, P.M. Hui, and N.F. Johsnon, e-print cond-mat/0408557.

[5] S. Gourley, S.C. Choe, P.M. Hui, and N.F. Johnson, Europhys. Lett. 67, 867 (2004).

[6] T.S. Lo, H.Y. Chan, P.M. Hui, and N.F. Johnson, Phys. Rev. E (Oct. 2004); e-print cond-mat/0406391.

Prof. N. Ito     (Department of Applied Physics, School of Engineering, The University of Tokyo,
                       Tokyo 113-8656,  JAPAN)

e-mail: ito@ap.t.u-tokyo.ac.jp

trip: Hangzhou at 12:05 of Nov. 4 on flight JAL635,
               depart at 13:15 of Nov. 8 on JAL636,   Lily hotel

PHOTO

       Computer Emulation of Nonequilibrium Phenomena

                   Nobuyasu Ito

              Department of Applied Physics
                 School of Engineering
                The University of Tokyo
                 Tokyo 113-8656, JAPAN

Major purpose of the statistical physics is to elucidate
nonequilibrium phenomena. Despite the success on equilibrium
state and some response theories in linear nonequilibrium
state, theoretical studies have been confronted with difficulties in
rather basic problems like transport coefficients. For example,
statistical mechanical studies of the heat conductivity has
been continued for more than half a century, but nanoscopic
origin is still not clear. On the other hand, the computer
simulations of nanoscale to macroscale phenomena are becoming feasible.
In this talk, results of statistical physics simulation on
nonequilibrium phenomena are to be given. Starting from basic
transport problems, flows and other dynamic phenomena are presented.
From application side, these are intended to bridge the
nanoscopic dynamics and macroscopic nonequilibrium phenomena
by reconstructing mesoscopic behavior. It turns out that the
description using continuum mechanics with local equilibrium works
surprisingly well even in the nanoscale, and that such statistical
physics simulations provides useful analyzing tools for complex
phenomena, like multiphase thermal flow. Now our computer can not
only just simulate real phenomena, but also emulate them.
　

Prof. Kimmo Kaski   (Helsinki University of Technology, Laboratory of Computational Engineering, 
                             P.O.Box 9203, FI-02015 HUT, FINLAND)

email: kimmo.kaski@hut.fi
tel.  +358 - 9 - 451 4825
fax.  +358 - 9 - 451 4833
http://www.lce.hut.fi/~kaski

trip: Hangzhou, 4 Nov. 13:45 MU 5172, to 8 Nov 13:15 JAL 636

Turing Systems as Models for Pattern Formation in Nature

In 1952 one of the key scientists of the 20th century, Alan Turing proposed a system of reaction-diffusion equations describing chemical reactions and diffusion to account for morphogenesis, i.e., the  development of patterns, shapes and structures found in nature.  These complex systems have been used in explaining,  e.g. patterns on animal skins and the segmentation in embryos. Here we will discuss our study of such pattern formations and their structures obtained through numerical simulation of the Turing  mechanism in two and three dimensions. We have investigated the  dependence of resulting structures on the system parameters,  transitions between these structures, and connectivity of  chemicals in the system. In addition, we have studied the effect  of random noise on developing these structures because noise plays a crucial role in the behaviour of various systems in  nature. For example we have observed melting or stabilization  of well order patterns due to noise. In summary Turing type -  relatively simple - systems can result in a rich complexity of  morphological patterns, which look remarkably similar to various  patterns appearing in chemical or biological systems.  

 Refs.
[1]   A.M. Turing: The chemical basis of morhogenesis,  Phil. Trans R. Soc. Lond.B237, 37-72 (1952).  [2]  T. Leppänen, M. Karttunen, K. Kaski, R.A. Barrio, and L. Zhang:  A new dimension to Turing    
        patterns, Physica D168-169, 35-44 (2002).
[3]   T. Leppänen, M. Karttunen, K. Kaski, and R. A. Barrio: The effect of noise on Turing patterns,
        Prog. Theor. Phys. (Suppl.) 150, 367-370 (2003).
 [4]  T. Leppänen, M. Karttunen, K. Kaski, and R. A. Barrio: Dimensionality effects in Turing pattern
        formation,  Int. J. Mod. Phys. B17, 5541-5553 (2003).
 [5]  T. Leppänen, M. Karttunen, K. Kaski, and R. A. Barrio: Turing systems as models of complex
        pattern formation, Braz. J. Phys.34, 368-372 (2004).
[6]  T. Leppänen, M. Karttunen, R.A. Barrio, and K. Kaski: Morphological transitions and bistability in
       Turing systems, to appear in Phys. Rev. E, 2004.

Prof. B.J. Kim         (Department of Molecular Science & Technology,
                                   Ajou University, Suwon 442-749, Korea)

e-mail: beomjun@ajou.ac.kr
Tel: (+82) 31-219-2571
Fax: (+82) 31-219-1615

trip: Hangzhou, 4 Nov. OZ 0359 13:50,  8 Nov. OZ 0360 15:00,  Lily hotel

Coarsegraining of Complex Networks
　

We perform the renormalization-group-like numerical analysis
of geographically embedded complex networks on the two-dimensional
square lattice. At each step of coarsegraining procedure, the four vertices
on each 2*2 square box are merged to a single vertex,
resulting in the coarsegrained system of the smaller sizes.
Repetition of the process leads to the observation that the
coarsegraining procedure does not alter the qualitative characteristics
of the original scale-free network, which opens the possibility of
subtracting a smaller network from the original network
without destroying the important structural properties.
The implication of the result is also suggested in the context
of the recent study of the human brain functional network.
　

Prof. Bongsoo Kim        (Changwon National University, Korea)

e-mail: bskim@changwon.ac.kr
fax: +82-55-267-0264.
　

Nonequilibrium critical dynamics of
                           the triangular antiferro-Ising model  

Bongsoo Kim

Department of Physics

Changwon National University

Changwon, 641-773, Korea

We present the nonequilibrium dynamics of AF Ising model on a triangular lattice subjected to the zero (critical) temperature quench via spin-flip Monte Carlo kinetics. Macroscopic degeneracy of the ground state fundamentally affects the dynamics. In particular, the defects and loose spins (whose flip costs no energy) play key roles in the dynamics. The long time evolution is characterized by a critical dynamic scaling with a growing length scale. With random initial conditions, this length exhibits a subdiffusive growth in time, while it grows diffusively for the relaxation within the dominant sector of ground state manifold. Persistence and the two-time temporal properties are also discussed.

Prof. B.W. Li   (Department of Physics, National University of Singapore, 117542,  Singapore)

E-mail: phylibw@nus.edu.sg
Phone: (65) 6874 2962
Fax: (65) 6777 6126

trip: Hangzhou, 5 Nov. 18:30, Hangzhou, 8 Nov.,   Lily hotel

PHOTO

Thermal diode and thermal transistor:

Controlling heat flow through nonlinear dynamics

Baowen Li 

The invention of semiconductor transistor and its relevant devices that control the charge flow has revolutionized our daily life in every aspect. However, over half century has been past, similar devices for controlling heat flow are still lacking.

In this talk, I will give a detailed discussion about our recent invention of thermal diode and thermal transistor, the two devices for controlling heat flow.  Emphasis will be given on the physical principle/mechanism of these two devices.

The thermal diode is a one way road for heat flow [1]. It allows the heat flow from one direction, while it prohibit heat flow for the another one. Like the electronic counterpart, the thermal transistor [2] is a three-terminal device with the important feature that the current through the two terminals can be controlled by small changes we make in the current or temperature at the third terminal. This control feature allows us to amplify the small current or to switch the device from an “off” (insulating) state to an “on” (conducting) state.  

The work is supported by FRG of NUS and the Temasek Young Investigator Award of DSTA Singapore and NUS.

Refs.

[1] B Li, L Wang and G Casati, “Thermal diode: A one way road for heat current”, cond-mat/0407093, to appear in Phys. Rev. Lett..

[2] B Li, L Wang and G Casati, “The thermal transistor: A switch and a control for heat current”.

Prof. H.R. Ma   (Institute of Theoretical Physics, Shanghai Jiao Tong University, Shanghai, China)

e-mail: hrma@sjtu.edu.cn
http://hrma.physics.sjtu.edu.cn, http://www.softmatter.net
Phone & Fax: 021-54743241

Monte Carlo simulations on the effective interactions of colloids

Hongru Ma and Weihua Li
Institute of Theoretical Physics, Shanghai Jiao Tong University, Shanghai 200240, China

The effective interactions between colloid particles in two component colloid systems are important in understanding the phase properties of colloids. By means of the Acceptance Ratio Method (ARM) developed long ago by Bennett, we simulated the effective interactions of different colloidal systems and
obtained excellent results. Our calculations show that ARM is a very powerful method in the studies of effective interactions of colloids. In this talk we will first introduce the general ARM method, followed by the implementation of  the method in the studies of colloid systems and the preliminary results. Then we
give some examples of the effective interactions obtained by the ARM as well as other methods, and finally we will give comparisons of the results and calculation efficiency of different methods.
　

Prof. Y.Q. Ma        (Nanjing University,  National Laboratory of Solid State  
                          Microstructures, Nanjing 210093, China)

e-mail: myqiang@nju.edu.cn

trip: Hangzhou, 4 Nov., by bus

　                  Phase Ordering in Soft Matter

                                             Yu-qiang Ma

National Laboratory of Solid State Microstructures,
Nanjing University, Nanjing 210093, China

In this talk, I first present a review of a wide variety of self-organizing behaviors in soft materials, and then discuss how to design and control self-assembled ordering structures from our recent works. In particular, we will examine in detail the formation of self-organizing ordering structures in  phase-separating systems due to competing interactions, hydrodynamic interactions, and entropy effects, and find that a moderate amount of fluctuations may assist the formation of highly ordered structures by removing topological defects or reorganizing structures. The results clearly indicate a possibility for the production of highly ordered and defect-free multi-scale materials by introducing “disordering” factors, and provide an interesting and universal picture to account for “fluctuation-induced ordering” phenomena in multicomponent materials.

Dr. Y. Nakayama       (Surface Control Engineering, Department of  Chemical Engineering,
                                      Kyoto University, Kyoto 615-8510,  Japan）

e-mail: nakayama@cheme.kyoto-u.ac.jp   
http://stat.scphys.kyoto-u.ac.jp/~nakayama
tel : +81-75-753-2672
fax : +81-75-753-2692

trip: Hangzhou, 4 Nov NH951 12:20,  to 8 Nov NH158 14:05

A Simulation Method to Resolve

Hydrodynamic Interactions in Colloidal Dispersions

 Y. Nakayama and R. Yamamoto

The flow of a colloidal dispersions is difficult to quantify
experimentally or to predict theoretically.  The direct inter-particle
interactions affects the flow. Moreover, internal microstructures of
solvent induce effective inter-particle interactions and change the
flow dramatically.

The interaction mediated by solvent is a key factor in understanding
dynamical behavior of colloids.  A fundamental understanding of
dynamical behavior of colloids is limited to simple systems. Numerical
simulations can be useful tools to extend the understanding such a
complex problems in colloidal dispersions.

We developed a numerical scheme to simulate colloidal dispersions in
complex fluids, named ``Smoothed Profile method''(SP method).  In
Smoothed Profile method, material transport and flow of solvent is
solved by hydrodynamic equations and motions of colloids are solved in
the manner of classical molecular dynamics simulation.  In
conventional methods, moving solid-fluid boundary conditions and
evaluation of force acting on colloids demands complex algorithms
and/or huge computational resources.  In SP method, the solid-fluid
interface is modeled as diffuse interface.  SP method is a way to deal
with moving boundary conditions and to evaluate forces on colloids
both efficiently and accurately.  By using SP method, many-colloid
systems can be simulated without neglecting many-body interaction
mediated by solvent, such as hydrodynamic interactions.

Ref:

[1] cond-mat/0403014
　

Prof. T. Nakamura      (Laboratory of Natural Science, Faculty of Engineering,
                                         Shibaura Institute of Technology, 307 Fukasaku, Minuma,
                                         Saitama 337-8570, Japan)

e-mail: tota@sic.shibaura-it.ac.jp
Tel/Fax:  (+81)-48-687-5028

trip: Hangzhou at 12:05 of Nov. 4 on flight JAL635,
               depart at 13:15 of Nov. 8 on JAL636,   Lily hotel

PHOTO

Nonequilibrium relaxation analysis of the spin-glass transition

Existence or nonexistence of a spin-glass phase transition in continuous
spin-glass model is now under debate. We study this problem with the nonequilibrium relaxation(NR) method. It takes a different approach to the thermodynamic limit. Most simulations take the equilibrium limit first, and then take the infinite-size limit by the finite-size-scaling analysis. The NR method takes the infinite-size limit first, and takes the equilibrium limit by the finite-time-scaling analysis. The exchange of the limits are done by the dynamic scaling hypothesis t=L^z. It solves many difficulties of simulations of so-called 'slow-dynamic systems', where the systems mainly order in the time direction not in the space direction. In this talk we clarify the phase transition in the Heisenberg and the XY spin-glass models in three dimensions. Relaxation functions of the spin-glass susceptibility, the Binder parameter, and the spin-glass correlation length are presented. Our estimates for the transition temperature suggest that the spin-glass and the chiral-glass transitions occur simultaneously, and that the phase transition is mainly driven by the spin degrees of freedom. If possible, we also present results
on a random-bond quantum spin chain, which yields the impurity-inducing AF long-range order phenomenon.

Prof. Y. Ozeki         (University of Electro-Communications, Japan)

e-mail: yozeki@pc.uec.ac.jp

trip: Hangzhou at 12:05 of Nov. 4 on flight JAL635,
               depart at 13:15 of Nov. 8 on JAL636,   Lily hotel

Nonequilibrium relaxation analyses on universality classes for
complex phase transitions
　

It has been shown that the nonequilibrium relaxation (NER) process is
useful to study phase transitions and critical phenomena. Observing
relaxation of physical quantities, one can locate the critical point
finely and estimate critical exponents accurately. The advantage of
the method originates in the facts that the behavior in the
thermodynamic limit is observed more easily. Since the equilibration
is not necessary, it is appropriate much to study systems with slow
dynamics such as frustrated and/or random systems.

The applications are expanding to wide variety of statistical models.
It has applied to disordered systems including the SG transition. For
the KT transition, the finite-time scaling analysis is useful to
estimate the transition point and critical exponents. The 2D fully
frustrated $XY$ models and the gauge glass models are analyzed and
some conclusive results are presented for controversial problems. In
this talk, I show the results on universality classes for the
following complex transitions; (1) the FM transition for the Ising SG
model in 2D and 3D, (2) the chiral transition for the fully frustrated
XY models in 2D, (3) the KT transition for the XY models in 2D. 

Dr. Simon Trebst  (Theoretische Physik, ETH Zurich, CH-8093 Zurich,
                              Switzerland Computational Laboratory, ETH Zurich, CH-8092 Zurich, Switzerland)

e-mail: trebst@itp.phys.ethz.ch
http://www.itp.phys.ethz.ch/staff/trebst/

trip: Shanghai, 4 Nov 10:45h, KL 0895,  9 Nov 12:15h, KL 0896

Optimizing the ensemble for equilibration
in broad-histogram Monte Carlo simulations

Simon Trebst, David A. Huse, Matthias Troyer

　

Flat-histogram or multicanonical sampling methods such as the
Wang-Landau algorithm have been employed in a variety of fields in
computational physics over the last three years. Only recently it was
realized that flat-histogram methods perform sub-optimally: For 2D
Ising models it was shown that the typical round-trip times needed to
sample the entire bandwidth of energies do not scale with the number of
spins N as the minimal N^2 of an unbiased random walk in energy
space, but exhibit an additional power law scaling O(N^{z}).

In this talk we present an adaptive algorithm which overcomes this
slowdown by optimizing the statistical-mechanical ensemble in a
generalized broad-histogram Monte Carlo simulation to maximize the
system's rate of round trips in total energy.

We use a feedback loop that reweights the ensemble based on preceding
measurements of the local diffusivity of the total energy. This detects
the ``bottlenecks'' in the simulation as minima in the diffusivity,
e.g. at critical points, and reallocates resources to those energies in
order to minimize the slowdown.

The scaling of the mean round-trip time from the ground state to the
maximum entropy state for this local-update method is found to exhibit
only logarithmic corrections O([\log N]^2) for both the ferromagnetic
and the fully frustrated 2D Ising model with N spins compared to the
minimal N^2 scaling. Our new algorithm thereby substantially
outperforms flat-histogram methods which are slowed down by an
additional power law scaling O(N^{z}). We demonstrate a speedup of
equilibration times by a factor of 50 for the fully frustrated 2D Ising
model.

We find that the resulting statistical errors in the density of states
as estimated by this new algorithm are nearly uniform in energy, in
strong contrast to flat-histogram simulations where the errors are much
larger at low energy than at high energy.

We will further outline applications to continuous systems, such as
high density Lennard-Jones fluids, and strongly correlated quantum
systems. The feedback idea can also be applied to optimize the
temperature set in the parallel tempering algorithm.

Contributing talks

Prof. Q.H. Chen         (Zhejiang University)

e-mail: qhchen@zju.edu.cn

           Dynamics of Josephson vortices in 3D Layered Superconductors

      Simulations on in-plane current driven Josephson vortex systems in 3D Layered Superconductors have been extensively performed. For high magnetic field (or high anisotropy), we find a Lorentz-force
independence of resistivity, due to the quasi-long-range order inside the block layers and interlayer short-range order. The power-law current-voltage characteristic at low temperatures and an evident
jump of the IV exponent from 3 to 1 at a critical temperature may be attributed to a possible novel Kosterlitz-Thouless like phase transition. For low magnetic field (or low anisotropy), a Lorentz-force independence of resistivity is observed, owing to both intralayer and interlayer long-range-order. To clarify the mechanism of the above observation, the behavior of Josephson vortex  and pancake vortices  have been be described in a microscopic level. Our numerical results are consistent
with the previous experimental observations.

Longjiu Cheng *  （Department of chemistry, University of science and technology of China,
                                     Hefei  230026, Anhui, P. R. China）

e-mail: clj@ustc.edu
tel: 86-551-360-6160 handy: 13965133383
http://mail.ustc.edu.cn/~clj/

trip: Hangzhou, 4 - 7 Nov

Effective algorithms for the optimization of
the Lennard-Jones clusters and  (C60)N clusters

We proposed a novel and effective cluster similarity checking method using the connectivity table (CT). CT can contain the topological information of the structure, and easily to be realized. An adaptive immune optimization algorithm (AIOA) is utilized for optimization of Lennard-Jones (LJ) clusters up to LJ(110) using the CT for similarity checking. It is proved that the method is very efficient, and the method is also capable for optimizations of larger clusters, e.g., LJ(200).

A high efficient unbiased global optimization method named as dynamic lattice searching (DLS) was also proposed. Because the DL can greatly reduce the searching space and the number of the time-consuming local minimization procedures, the proposed DLS method runs in a very high efficiency, especially for the clusters of large size. The performance of the DLS is investigated in the optimization of Lennard-Jones (LJ) clusters up to 309 atoms, and the structure of the LJ₅₀₀ is also predicted. Furthermore, the idea of dynamic lattice can be easily adopted in optimization of other molecular or atomic clusters. It may be a promising approach to be universally used for structural optimizations in chemistry field. By comparing the results of DLS and the annealing experiments of (C60)N clusters, there are well correspondence between our theoretical study and the experiments.

Ref.

[1] Chemical Physics Letters 389 (2004) 309-314

[2] Journal of Computational Chemistry 25 (2004) 1693-1698

Dr. P. Crompton       (University of Leipzig)

e-mail: crompton@itp.uni-leipzig.de

trip: Hangzhou, 1 - 14 Nov.

A Fisher Zeroes Analysis of the Continuous-Time
Quantum Monte Carlo Method

We report on a new measurement technique to evaluate partition function zeroes, the zeroes of a semi-analytic polynomial expansion in coupling, evaluated directly from the generic quantum Monte Carlo transfer matrix. We comment on, and compare the finite-size-scaling arguments of these
zeroes with those of conventional analyses and thermodynamic observables used to extract the critical exponents of two mixed-spin chain models of recent focus. These two spin-chains undergo a 2nd order phase transition as a function of coupling anisotropy, in agreement with nonlinear sigma
model treatments, but in addition we quantify corrections to scaling associated with the appearance of a finite-volume dependent topological charge associated with the lack of exact ergodicity of the continuous time Quantum Monte Carlo algorithm on finite lattices. We further relate this topological scaling effect to the phenomenological O(1/N_c) scaling of quantum corrections to operator mixing under renormalisation in light-cone QCD, as a new effective approach for nonperturbative input.

Ref:  P.R. Crompton, W. Janke, Z.X. Xu, H.P. Ying
Presented at 22nd International Symposium on Lattice Field Theory (Lattice 2004),
Batavia, Illinois, 21-26 Jun 2004.
e-Print Archive: hep-lat/0408041

Prof. J.Q. Fang     (China Institute of Atomic Energy, P. O. Box 275-81, Beijing 102413)

e-mail: fjq96@iris.ciae.ac.cn
tel: 010-60386433

trip: Hangzhou, 5 Nov. 16:30, railway

PHOTO

Dynamical Simulation of High Current Beam Transport Network

An intense beam propagating through a periodic focusing beam transport network (PFBTN) should be described by the nonlinear Vlasov-Maxwell-Poisson equations (VMPE).  It is important  to be to study the VMPE, the equilibrium and stability properties of general distribution functions for the periodically-focused beams. Despite its limited practical interest, the K-V beam equilibrium and its generation to a rotating beam in a PFBTN, has been the only known periodically-focused equilibrium solution to the describe an intense beam propagating through a PF-BTN field. The difficulty of solving the VPME in general lies in the fact that the Hamiltonian for the motion of an individual beam particle is time dependent. A third-order Hamiltonian averaging technique using a canonical transformation to average over the fast time scale associated with the betatron oscillations has been developed. More challenge is to develop effective multiple-particles in cell simulation approach and technique in various PFBTN for different initial particle distributions . In this article, we will give a general review of recent progress and main simulation on studies of time evolution of charged-particle bunches in PFBTN, the self-consistent treatment of  space-charge forces in beam macro-particles are carried out. Several kinds of methods for Control and synchronization of beam halo-chaos in the PFBTN are developed by us in recent years. Potential of application prospective will be pointed out finally. 
　

   
Prof. Wenan GUO          (Beijing Normal University)

e-mail: waguo@bnu.edu.cn
http://202.112.94.136/include/people/waguo.htm
　

trip: Hangzhou, 4 Nov.16:00,  share a double room.

Monte Carlo renormalization: test on the triangular Ising model

Wenan Guo, Henk W.J. Bloete and Zhiming Ren

Physics Department, Beijing Normal University,
Beijing 100875, P. R. China
and
         Faculty of Applied Sciences, Delft University of Technology,
  P.O. Box 5046, 2600 GA Delft, The Netherlands

We test the performance of the Monte Carlo renormalization method using the Ising model on the triangular lattice. We apply block-spin transformations which allow for adjustable parameters so that the
transformation can be optimized. This optimization takes into account the relation between corrections to scaling and the location of the fixed point. To this purpose we determine corrections to scaling of the triangular Ising model with nearest- and next-nearest-neighbor interactions, by means of transfer matrix calculations and finite-size scaling. We find that the leading correction to scaling just vanishes for the nearest-neighbor model. However, the fixed point of the commonly used majority-rule block-spin transformation lies far away from the nearest-neighbour critical point. This raises the question whether the the majority rule is suitable as a renormalization transformation, because corrections to scaling are supposed to be absent at the fixed point. We define a modified block-spin transformation which shifts the fixed point back to the vicinity of the nearest-neighbour critical Hamiltonian.  This modified transformation leads to results for the Ising critical exponents that converge faster, and are more accurate
than those obtained with the majority rule.

Dr. Hwee Kuan Lee     (Department of Physics, Tokyo Metropolitan University,
                                       1-1 Minami-osawa, Hachioji, Tokyo, Japan 192-0397)

e-mail: hweekuan@phys.metro-u.ac.jp

trip: Hangzhou, 05 Nov 12:50 hours - 08 Nov 13:50 hours
　

Reweighting for Nonequilibrium Markov Processes

Hwee Kuan Lee and Yutaka Okabe

Department of Physics, Tokyo Metropolitan University

We present a generic reweighting method for nonequilibrium Markov
processes. With nonequilibrium simulations at a single temperature, one
calculates the time evolution of physical quantities for a range of
temperatures. Using the dynamical finite-size scaling analysis for the
nonequilibrium relaxation, one can study the dynamical properties of
phase transitions as well as the equilibrium ones. We demonstrate the
procedure for the Ising model with the Metropolis algorithm, but the
present formalism is general and can be applied to a variety of systems
and with different Monte Carlo update schemes.

Jia-Lin Lo *       (Taiwan National Central University
                                       15F.-2, No.26, Wenhua St., Pingzhen City, Taoyuan County 324, Taiwan)

e-mail: jl.lo@msa.hinet.net
Fax:  886-3-42653

trip: Hangzhou, MF 898Q Nov 4, 15:00, to Nov 9, 1600 

Parallel-tempering simulation in protein folding & distributive computing

Parallel tempering was applied to molecular dynamics simulations
in explicit solvent in a study of the folding thermodynamics of the
designed 20-residue peptide Trp-Cage. The GROMOS96 43a1 force field
was used for the molecular dynamics. The actual computer simulations
were performed on the distributive computing facility "Protein@CBL".
Initial conformations of peptide were extended random coils and the
simulation temperature range was 250 to 680 degree K. After an
accumulated simulation time of 800 ns, one peptide sampled a
conformation in which the rms positional deviation of its C_alpha's
relative to the the native structure of the peptide was 0.173 nm.

T. Otobe *  (Department of Physics, Waseda University, Shinjuku-ku, Tokyo 169-8555, JAPAN)

E-mail: otobe@hep.phys.waseda.ac.jp
tel: +81-3-5286-3444,   fax: +81-3-3204-1567

trip:  Shanghai, 4 Nov MU524 16:00, - 9 Nov (?)
　

            Short-time dynamics in lattice gauge theory

Short-time dynamic scaling is utilized to investigate the finite temperature
phase transition in SU(2) lattice gauge theory. A method of fixing the
critical point solely through the non-equilibrium dynamic relaxation
procedure is discussed in detail. Using the critical point so obtained, the
short-time critical relaxation starting either from cold or hot initial
configuration is investigated to fix the values for various critical
exponents.

Prof. Gang Sun           (Institute of Physics, Chinese Academy of Science, Beijing 100080)

e-mail: gsun@aphy.iphy.ac.cn

A New Type of Segregation

in Vertically Vibrated Binary Granular Mixtures

Granular binary mixtures shaken vertically in vacuum are studied by molecular dynamic simulations based on soft sphere approach. The particle size is always set to be the same, but the density can be different. We find that when the density ratio of the two kinds of particles is high enough a new type of segregation occurs. In this segregation, lighter particles tend to go up and form a pure layer on the top of the system, while the heavier and the remained lighter ones stay at the bottom and form a mixed layer. The thickness of the pure top layer is increased as the density ratio increases. The detailed boundary of the occurrence of the segregation also depends on the particle size, shaken frequency and amplitude. A scaling relation between particle size and shaken frequency is suggested for this system. The observed segregation can be illustrated by the competition between the impact of momentum of the heavier particles and the stiffness of the layer composed by the mixed particles. Similar experiments are also carried out, and the results are consistent with that of simulations.

Prof. Yougui Wang       (Department of Systems Science, School of Management,
                                           Beijing Normal University, Beijing, 100875, P. R. China)


Email: ygwang@bnu.edu.cn
Tel: 010-58807876
　

The applications of transfer model on monetary distribution and dynamics

 Yougui Wang, Ning Ding, Ning Xi

Department of Systems Science, School of Management

Beijing Normal University, Beijing 100875, P.R. China

Recently, statistical mechanics and simulation method have been widely applied to the economic and financial issues. To explore the universe mechanism behind income or wealth distribution, a series of money transfer models have been developed by some econophysists[1-5]. These models are so powerful that they could be applied to many aspects of monetary issue. In this presentation, we review recent studies on wealth distribution based on money transfer models, and present our research works on this subject, including circulation and mobility[5-8]. The simulation results on these models show that the shape of the holding time distribution is exponential when money is randomly exchenged, implying money circulation is a Poisson process, but changes to a power type when preferential behavior is introduced[6,7]. To compare the mobilities in four money transfer models, we recorded the agents' ranks all through the time, and found that even though different models have the same shape of distribution, their mobilities may be quite different[8]. We believe that these investigations can help us to comprehend the dynamic mechanism behind the distribution.

References

[1] A. Drăgulescu and V. M. Yakovenko, Eur. Phys. J. B 17, 723 (2000).

[2] A. Chakraborti and B. K. Chakrabarti, Eur. Phys. J. B 17, 167 (2000).

[3] B. Hayes, Am. Scientist 90, 400 (2002).

[4] A. Chatterjee, B.K. Chakrabarti and S.S. Manna, Physica A 335, 155 (2004).

[5] N. Ding, Y. Wang, J. Xu and N. Xi. Int.J.Mod.Phys. B18, 17-19, 2725 (2004)

[6] Y. Wang, N. Ding, and L. Zhang, Physica A 324, 665 (2003).

[7] N. Ding, N. Xi, and Y. Wang, Eur. Phys. J. B 36, 149 (2003).

[8] N. Ding, N. Xi, and Y. Wang, submitted to Physica A (2004)