FisEs 2011

XVII Congreso de Física Estadística

Libro de resúmenes

Barcelona, 2-4 Junio de 2011

Facultad de Física, Universitat de Barcelona
© FisEs11, 2011

Cubierta:
bi composición
i i con cuatro fi figuras seleccionadas
l i d entre llas contribuciones
ib i all congreso. En sentido
id horario,
h i y
empezando por el panel superior izquierdo, J. C. R. E. Oliveira et al. (P-117); G. R. Lázaro et al. (P-82); R. Ledesma-
Aguilar et al. (O-14); J. Sala et al. (P-147).
Banner: skyline de Barcelona adaptado de imagen © vectorine, [Link].

Maquetación y compilación: comité organizador FisEs11.

Comité Cientı́fico
• Carlos Cabrillo (Instituto de Estructura de la Materia, CSIC)
• Rodolfo Cuerno (Universidad Carlos III, Madrid)
• Enrique de Miguel (Universidad de Huelva)
• Elvira Guàrdia (Universitat Politècnica de Catalunya)
• Diego Maza (Universidad Navarra)

• Miguel Ángel Muñoz (Universidad de Granada)

• Ignacio Pagonabarraga (Universitat de Barcelona)

• Miguel Ángel Rodrı́guez (Universidad de Cantabria)

• Marı́a José Ruiz (Universidad de Sevilla)
• Pedro Tarazona (Universidad Autónoma de Madrid)
• Raúl Toral (Universitat de les Illes Balears - CSIC)
• Juan A. White (Universidad de Salamanca)

Comité Local
• Giancarlo Franzese (Universitat de Barcelona)
• Elvira Guàrdia (Universitat Politècnica de Catalunya)
• Marta Ibañes (Universitat de Barcelona)
• Jordi Ignés (Universitat de Barcelona)
• Jordi Martı́ Rabassa (Universitat Politècnica de Catalunya)
• Ignacio Pagonabarraga (Universitat de Barcelona)
 Índice General

I Conferencias invitadas 17
1. Thermodynamics of a small system in a μ,T reservoir 19
Dick Bedeaux

2. La teorı́a del coarse-graining, aplicaciones y problemas abiertos 20

Pep Español

3. Frustrated nematic order in spherical geometries 21

Alberto Fernandez-Nieves

4. Graphene and its unique properties 22

Francisco Guinea

5. Elastic Instabilities Lead To Novel Material Properties 23

Tom Mullin

6. On the short range order of liquid phases 24

L. C. Pardo, M. Rovira-Esteva, S. Pothoczki, A. Ottochian, and J. Ll. Tamarit

7. Multiscale mobility networks and the large scale spreading of infectious diseases 25
José J. Ramasco

8. First-principles simulation of biomolecular processes 26

Carme Rovira

9. ¿ Cuan de normales son las anomalı́as en el transporte y difusión de partı́culas brownianas ? 27

José M. Sancho

10. The emergent dynamics of collective cell migration 28

Xavier Trepat

11. Zaragoza Scientific Center for Advanced Modeling (ZCAM) 29

Michel Mareschal

II Orales 30
1. Delayed Coupling Theory of Vertebrate Segmentation 31
Saúl Ares, Luis G. Morelli, Leah Herrgen, Christian Schröter, Frank Jülicher, and Andrew C. Oates

2. Deterministic optical rogue waves 32

Cristian Bonatto, Cristina Masoller, José R. Rios Leite, Michael Feyereisen, Stéphane Barland, Massimo
Giudici, and Jorge R. Tredicce

3. Instabilities of conducting fluid layers in cylindrical cells under external forcing 33

Javier Burguete and Montserrat Miranda

4. Mecánica Estadı́stica del Sistema Inmune 34

J. Currie, M. Castro, Grant Lythe y C. Molina-Parı́s

5. Nonuniversal results induced by diversity distribution in coupled excitable systems 35

L. F. Lafuerza, Pere Colet, and R. Toral

6. Leyes de Potencia y Leyes de Escala en la Distribución de Energı́a de los Huracanes 36

Álvaro Corral, A. Ossó y J. E. Llebot

7. Lı́mite a la descripción hidrodinámica en el modelo de Maxwell para gases granulares 37

J. J. Brey, M. I. Garcı́a de Soria y P. Maynar

8. Hidden Symmetries in Nonequilibrium Systems 38

P. I. Hurtado, C. Pérez-Espigares, J. del Pozo, and P. L. Garrido
9. Prisoner’s Dilemma on a sizeable network: An experiment with human subjects 39
Jelena Grujić, Constanza Fosco, Lourdes Araujo, José A. Cuesta, and Angel Sánchez

10. Statistical mechanics approaches to complex network inference and reconstruction 40

R. Guimerà and M. Sales-Pardo

11. Caracterización y Modelización de Sistemas Dinámicos no Lineales. Medidas de Desorden

Dinámico y ”Self-Correlation” 41
Juan Antonio Hernández Álvarez, Rosa Marı́a Benito y Juan Carlos Losada

12. Universalidad en el flujo de un medio granular a través de un orificio 42

A. Janda, I. Zuriguel, A. Garcimartı́n y D. Maza

13. Natural Correlations in Networks: Origin and Effects 43

Samuel Johnson, Joaquı́n J. Torres, J. Marro, and Miguel A. Muñoz

14. Wetting-controlled drop emission in forced microfluidic filaments 44

R. Ledesma-Aguilar, R. Nistal, A. Hernández-Machado, and I. Pagonabarraga

15. Self-assembly of viral capsids 45

Antoni Luque and David Reguera

16. Complex networks and glassy dynamics: walks in the energy landscape 46
Paolo Moretti, Andrea Baronchelli, Alain Barrat, and Romualdo Pastor-Satorras

17. Pattern Formation in Reactive-Fluid Systems 47

Alberto P. Muñuzuri, Jacobo Guiu-Souto, Dario M. Escala, Alexandra Von Kameke, and Jorge Carballido-
Landeira

18. Relaxation Dynamics in the Columnar Liquid Crystal Phase of Oblate Hard Spherocylinders:
Effect of Transient Cages and Permanent Barriers 48
Alessandro Patti, Simone Belli, René van Roij, and Marjolein Dijkstra

19. The helical crack-front instability 49

A. J. Pons

20. Estudio mediante geometrı́a intrı́nseca de procesos de crecimiento cinéticos con rugosidad 50
Javier Rodriguez-Laguna, Silvia N. Santalla y Rodolfo Cuerno

21. Thermodynamic tradeoffs in sensory adaptation: the energy-speed-accuracy relation 51

G. Lan, Pablo Sartori, S. Neuman, V. Sourjik, and Y. Tu

22. Flujo de Couette para una impureza inmersa en un gas granular 52

Francisco Vega Reyes, Vicente Garzó y Andrés Santos

III Paneles 53
1. Impact of subdiffusion on the formation and properties of morphogen gradients: a reaction-
diffusion approach 55
Santos Bravo Yuste, Enrique Abad, and K. Lindenberg

2. Propiedades del agua en condiciones extremas: región subenfriada y presiones negativas 56

J.L.F. Abascal, M.A. González y C. Vega

3. Depósitos y Flujo de Partı́culas con Caras Planas en un Silo 57

M. Acevedo, I. Zuriguel y D. Maza

4. Universal critical avalanches in the Coulomb glass 58

Martin Goethe and Matteo Palassini

5. Stabilization of self-propelling particle clusters 59

F. Alarcón and I. Pagonabarraga
6. Molecular dynamics study of polarizable ion models for molten copper halides 60
Olga Alcaraz, Vicente Bitrián, and Joaquim Trullàs

7. Application of an effective medium theory to heterogeneous reaction-diffusion systems 61

Sergio Alonso and Markus Bär

8. Estudio Monte Carlo de la fase de spin-glass de sistemas diluidos de dipolos de Ising 62

Juan J. Alonso y Julio F. Fernández

9. Vidrios de espı́n en cuatro dimensiones en presencia de un campo magnético 63

R. Álvarez-Baños, A. Cruz, L. A. Fernández, J. M. Gil-Narvión, A. Gordillo-Guerrero, D. Iñiguez, A.
Maiorano, F. Mantovani, E. Marinari, V. Martı́n-Mayor, J. Monforte-Garcı́a, A. Muñoz-Sudupe, D.
Navarro, G. Parisi, S. Pérez-Gaviro, J. J. Ruiz-Lorenzo, S. F. Schifano, B. Seoane, A. Tarancón, R.
Tripiccione y D. Yllanes

10. Caos y arritmogénesis en la dinámica del calcio intracelular 64

E. Alvarez-Lacalle, I. R. Cantalapiedra, B. Echebarria y A. Peñaranda

11. Aggregation of superparamagnetic colloids in magnetic fields: the quest for the equilibrium
state 65
J. S. Andreu, J. Camacho, and J. Faraudo

12. Experimental study of different LFF regimes in semiconductor lasers with an external cavity 66
Andrés Aragoneses, Jordi Zamora-Munt, Jordi Tiana-Alsina, Nicolas Rubido, Cristina Masoller, and
Maria Carme Torrent

13. Evolución temporal en el flujo longitudinal uniforme de un gas granular. Estudio mediante
simulaciones en computación grid 67
Antonio Astillero and Andrés Santos

14. Effects on the histeresis in martensitic transformations 68

V. Torra, A. Isalgue, C. Auguet, and G. Carreras

15. Buckling transition in quasi-spherical viral capsids 69

Marı́a Aznar, Antoni Luque, and David Reguera

16. Voter models on weighted networks and limits of the mean-field approach 70
Andrea Baronchelli, Claudio Castellano, and Romualdo Pastor-Satorras

17. Percolative analysis of nanoconfined supercooled water 71

Valentino Bianco and Giancarlo Franzese

18. Turbulent bubble dispersions in microgravity. Drop Tower experiments 72

Pau Bitlloch, Xavier Ruiz, Laureano Ramı́rez-Piscina, and Jaume Casademunt

19. A nonlinear mechanism of cell motility in lamellar actomyosin fragments 73

Carles Blanch-Mercader and Jaume Casademunt

20. Camino aleatorio, subdifusión, memoria, series divergentes y funciones de Bessel de primera
especie 74
Santos Bravo Yuste y Enrique Abad

21. Puzzling Out the Mechanical Stability and the Dynamics of the DV Organizer in the Wing
Imaginal Disc of Drosophila 75
Javier Buceta

22. Grand Canonical Monte Carlo simulations of adsorption using functionalized amorphous
silica 76
Santiago Builes and Lourdes F. Vega

23. Demographic growth and competition shape the size-area relationship for human languages 77
Susanna C. Manrubia, Jacob B. Axelsen, and Damián H. Zanette
24. A First Passage Time Analysis of Atomic-Resolution Simulations of the Ionic Transport
through a Bacterial Porin 78
Carles Calero, Jordi Faraudo, and Marcel Aguilella-Arzo

25. Solvation properties of mono- and di-valent cations in water from DFT molecular dynamics
simulations 79
Ausias-March Calvo, Elvira Guàrdia, and Marco Masia

26. Estudio del flujo sanguı́neo en una arteria estenótica 80

Marta Lage, Beatriz Rastrollo, Alexis Cantizano y Mario Castro

27. Thermodynamics of feedback controlled systems 81

Francisco J. Cao

28. Scale invariance in marine population dynamics 82

José A. Capitán and Gustav W. Delius

29. Harmonic vibration modifies the Turing pattern morphology from White-Spots to Black-
Spots 83
Jorge Carballido-Landeira, Jacobo Guiu-Souto, and Alberto P. Muñuzuri.

30. Detección automática de palabras clave en textos cortos 84

C. Carretero Campos, P. Bernaola-Galván, A. V. Coronado y P. Carpena.

31. Electrocinética DC de suspensiones de partı́culas concentradas en medios electrolı́ticos gen-

erales 85
Félix Carrique y Emilio Ruiz-Reina

32. Self-similarity and chaos in coupled ratchets. Unfolding the complexity

of cooperative Brownian motors 86
Carles Blanch-Mercader, Javier G. Orlandi, and Jaume Casademunt

33. Laminar oscillatory flow of Maxwell and Oldroyd–B fluids: Theoretical analysis and experi-
ments 87
L. Casanellas and J. Ortı́n

34. Control of spiral wave dynamics by low energy stimuli 88

Jorge Castro, Matias Rafti, Flavio H. Fenton, Jorge Carballido-Landeira, and Alberto P. Muñuzuri

35. First steps in the study of unconventional magnetic fluids 89

Juan J. Cerdà, Pedro A. Sánchez, and Tomás Sintes

36. Nature of Waterlike Liquid-Liquid Criticality 90

Claudio A. Cerdeiriña and Pablo G. Debenedetti

37. Coordination of mitotic axes guides morphogenesis of epithelial cysts 91

Benedetta Cerruti, Alberto Puliafito, Annette M. Shewan, Keith E. Mostov, Guido Serini, Antonio Celani,
and Andrea Gamba

38. Capillary rise between parallel plates: Velocity dependence of the capillary pressure 92
Xavier Clotet and Jordi Ortı́n

39. Dynamics of passively phased fiber laser arrays 93

A. Jacobo, P. Colet, E.J. Bochove, Y. Braiman, A.B. Aceves, R. Deiterding, C.A. Miller, C. Rhodes, and
S.A. Shakir

40. Time delay identification using permutation information theory quantifiers 94

Miguel C. Soriano, Luciano Zunino, Ingo Fischer, and Claudio R. Mirasso

41. Quasi-stationary analysis of the contact process on scale-free networks 95

Silvio C. Ferreira, Ronan S. Ferreira, and Romualdo Pastor-Satorras

42. Simple rules govern finite-size effects in scale-free networks 96

Sara Cuenda and Juan A. Crespo
43. Replicator dynamics for the n-player Prisoner’s Dilemma with “moody” conditional cooper-
ators 97
Jelena Grujić, José A. Cuesta, and Angel Sánchez

44. Coarse graining en dinámica browniana: modelos de difusión discreta 98

J. A. de la Torre y Pep Español

45. Phase diagrams of binary mixtures of patchy colloids with distinct numbers of patches 99
Daniel de las Heras, José Maria Tavares, and Margarida M. Telo da Gama

46. Universality of rain event size distributions 100

Anna Deluca, O. Peters, A. Corral, J. D. Neelin, and C. E. Holloway

47. Teorema de fluctuación-disipación para un oscilador activo ruidoso, el haz ciliar del oı́do
interno. 101
L. Dinis, P. Martin, J. Barral, J. Prost y J.F. Joanny

48. Information routing driven by background chatter in signaling networks 102

Núria Domedel-Puig, Pau Rué, Antonio J. Pons, and Jordi Garcı́a-Ojalvo

49. Del colapso gravitatorio a la descomposición espinodal 103

Alvaro Domı́nguez, Johannes Bleibel, Martin Oettel y Siegfried Dietrich

50. Supernormal conduction in cardiac tissue promotes concordant alternans and action potential
bunching 104
Blas Echebarria, Georg Röder, Harald Engel, Jörn Davidsen, and Markus Bär

51. Local routing of the Internet based on a linear projection of complex networks 105
Pau Erola, Sergio Gómez, and Alex Arenas

52. Changes in buoyancy-driven instabilities by using a reaction-diffusion system 106

Darı́o Martı́n Escala and Alberto Pérez Muñuzuri

53. A coarse-grained model for water in hydrophobic confinement 107

Cristian Esplugas and Giancarlo Franzese

54. Noise-tolerant signal detection in genetic circuits 108

Javier Estrada and Raúl Guantes

55. Modelo mesoscópico de interacción DNA-Proteı́na 109

R. Tapia-Rojo, D. Prada-Gracia, J. J. Mazo y F. Falo

56. Coupled Ising models and interdependent discrete choices under social influence in homoge-
neous populations 110
Ana Fernández del Rı́o, Elka Korutcheva, and Javier de la Rubia

57. Update rules and interevent time distributions: Slow ordering vs. no ordering in the Voter
Model 111
J. Fernández-Gracia, V. M. Eguı́luz, and M. San Miguel

58. Fluctuaciones de Pelı́culas Adsorbidas 112

Eva M. Fernández, Enrique Chacón y Pedro Tarazona

59. Una perspectiva termodinámica al estudio de comunidades microbianas: dinámica y estruc-

tura de población en cultivos in vitro de Plasmodium falciparum 113
Jordi Ferrer, Clara Prats y Daniel López

60. Tuning water transport in carbon nanotubes with a strong perpendicular electric field 114
Luis Figueras and Jordi Faraudo

61. Regulation of neuronal differentiation at the neurogenic wave front 115

Pau Formosa-Jordan, Marta Ibañes, Saúl Ares, and José Marı́a Frade

62. Large decrease of fluctuations for supercooled water in hydrophobic nanoconfinement 116
Elena G. Strekalova, Marco G. Mazza, H. Eugene Stanley, and Giancarlo Franzese
63. Fluctuations and Stochastic Cell State Switching 117
D. Frigola, J.M. Sancho, and M. Ibañes

64. Dos escenarios de extinción en redes de interacción planta-polinizador. Aplicación en redes

alpinas de alta montaña 118
Javier Galeano, Juan Manuel Pastor, Silvia Santamarı́a y Marcos Méndez

65. Estudio de la Solvatación de una Molécula de Amoniaco en Nanogotas de Helio Utilizando

Técnicas de Difusión de Monte Carlo Cuánticas 119
Antonio Gamboa, Jordan A. Ramilowski y Rosa M. Benito, and David Farrelly

66. Modelos primitivos para sistemas con autoensamblado 120

Noé G. Almarza

67. Dynamics of encapsulated water 121

Miquel Garcia-Ratés, Pere Miró, Carles Bo, and Josep Bonet Avalos

68. Un obstáculo a la salida disminuye la probabilidad de atasco 122

I. Zuriguel, A. Garcimartı́n, A. Janda y C. Lozano

69. Influencia de la asociación en las funciones respuesta de metanol: efecto de la temperatura y

la presión en condiciones supercrı́ticas 123
P. Gómez-Álvarez, A. Dopazo-Paz, J. M. Mı́guez, L. Romanı́ y D. González-Salgado

70. La introducción de las interacciones efectivas en la mecánica estadı́stica de las disoluciones:

el trabajo de Einstein del año 1905 sobre el movimiento browniano 124
Juan Luis Gómez Estévez

71. Self-localized states in one and two dimensions in lasers with external feedback 125
P.V. Paulau, Damià Gomila, Pere Colet, B. Malomed, and W.J. Firth

72. Autoorganización de filamentos de proteı́na FtsZ sobre un sustrato 126

P. González de Prado y P. Tarazona

73. Mesoscopic perturbation on a reaction-diffusion system modifies the Turing instability 127
Jacobo Guiu-Souto, Jorge Carballido-Landeira, and Alberto P. Muñuzuri.

74. An Age-Dependent Branching Model for Macroevolution 128

Emilio Hernández-Garcı́a, Stephanie Keller-Schmidt, Murat Tuğrul, Vı́ctor M. Eguı́luz, and Konstantin
Klemm

75. Control of nematic liquid crystals orientation using paramagnetic asymmetric microparticles 129
Sergi Hernàndez-Navarro, Pietro Tierno, Josep Claret, Jordi Ignés-Mullol, and Francesc Sagués

76. Amplitude equation description of vertebrate segmentation 130

Adrián Jacobo, Damià Gomila, Manuel Matı́as, Saul Ares, Luis Morelli, Andrew Oates, and Frank
Jülicher

77. Equilibrio y ruptura de simetrı́a en gases granulares vibrados 131

Nagi Khalil y J. Javier Brey

78. Coupled Ising models and interdependent discrete choices under social influence in homoge-
neous populations 132
Ana Fernández del Rı́o, Elka Korutcheva, and Javier de la Rubia

79. Analytical solution of a stochastic birth and death process including delay 133
Luis F. Lafuerza and Raul Toral

80. Transient dynamics and geometrical properties in an spatial predator-prey model 134
Karina Laneri, Alejandro Kolton, Manojit Roy, and Mercedes Pascual

81. Fluctuaciones grandes y función de grandes desviaciones en un sistema disipativo sencillo 135
A. Lasanta, A. Prados y Pablo I. Hurtado
82. Stable shapes in the discocyte-equinocyte transition of the Red Blood Cell 136
Guillermo R. Lázaro, Kathryn Melzak, Ignacio Pagonabarraga, and Aurora Hernández-Machado

83. Quantum annealing of a hard combinatorial problem 137

Daniel Lecina Casas and Matteo Palassini

84. Aggregation dynamics and fluctuations in in vitro cultures of Mycobacterium tuberculosis 138
I. Llopis, N. Caceres, C. Prats, C. Vilaplana, P-J. Cardona, and D. Lopez

85. Renormalization-group corrections of White’s approach for the prediction of the property
fluctuations in the critical region applied to SAFT equations 139
F. Llovell, E. Forte, A. Galindo, and L.F. Vega

86. Slow dynamics in a highly turbulent von Kármán swirling flow 140
M. López and J. Burguete

87. La Compatibilidad como mecanismo de Complejidad en las Redes 141

J. P. Cárdenas, M. L. Mouronte, R. M. Benito y J. C. Losada

88. Análisis experimental de la estabilidad de los arcos en un medio granular 142

C. Lozano, I. Zuriguel, A. Garcimartı́n y G. Lumay

89. Arritmias cardiacas producidas por heterogeneidad en la dinámica del calcio. 143
C. Lugo, E. Alvarez-Lacalle, I. R. Cantalapiedra, B. Echebarria y A. Peñaranda

90. Feigenbaum graphs: a complex network perspective of chaos 144

Bartolo Luque, Lucas Lacasa, Fernando J. Ballesteros, and Alberto Robledo

91. Dinámica estocástica y células T 145

Grant Lythe, Carmen Molina-Parı́s y Mark Day

92. Ion distribution around charged spherical colloids from molecular dynamics calculations with
explicit water molecules 146
Miroslava Nedyalkova, Stoyan Pisov, Francesc Mas, Eudald Vilaseca, and Sergio Madurga

93. Virial coefficients and divergence of pressure for hard disks and spheres 147
M. A. G. Maestre, A. Santos, M. Robles, and M. López de Haro

94. Energy landscape of constraint satisfaction problems via exhaustive enumeration 148
S. Mandrà and M. Palassini

95. Intrinsic structure, elastic properties and stability of the Newton Black Film 149
Héctor Martı́nez, Pedro Tarazona, Enrique Chacón, and Fernando Bresme

96. Temporal Griffiths Phases 150

Ricardo Martı́nez, Federico Vázquez, Juan A. Bonachela, Cristóbal López, and Miguel A. Muñoz

97. Synchronization in delayed mutually coupled optoelectronic oscillators 151

Jade Martı́nez Llinàs and Pere Colet Rafecas

98. Determinación de las propiedades interfaciales de la mezcla binaria agua-metano mediante

simulación molecular 152
J. M. Mı́guez, D. González-Salgado, G. Pérez-Sánchez, J. L. Legido y M. M. Piñeiro

99. Fases nemáticas biaxiales en fluidos de partı́culas duras con geometrı́a de tipo losa 153
Yuri Martı́nez-Ratón, Szabolcs Varga y Enrique Velasco

100. Emergencia de estrategias en juegos iterados con reciprocidad directa 154

Luis A. Martı́nez-Vaquero, José A. Cuesta y Angel Sánchez

101. DNA compaction induced by the anticancer peptide Kahalalide F: a single-molecule study 155
J. Camunas-Soler, C. V. Bizarro, S. de Lorenzo, X. Sisquella, A. Mata, R. Eritja, F. Albericio, S. B. Dev,
and F. Ritort
102. Emergent structures in bacterial suspensions under gravity 156
Ricard Matas Navarro and Ignacio Pagonabarraga

103. Nonlocality-induced front interaction enhancement 157

Manuel A. Matı́as, Damià Gomila, Pere Colet, Lendert Gelens, and Guy Van der Sande

104. Entropy of continuous mixtures and the measure problem 158

Pablo Maynar and Emmanuel Trizac

105. A vortex crystal that flows like a liquid: Grain-boundary scars in flat geometry 159
M. Carmen Miguel, Adil Mughal, and Stefano Zapperi

106. El papel de las correcciones de largo alcance en la determinación de las propiedades inter-
faciales del agua mediante simulación molecular 160
J.M.Mı́guez, D. González-Salgado, P. Gómez-Álvarez, J.L. Legido y M.M. Piñeiro

107. The dynamical strength of social ties in information spreading 161

Giovanna Miritello, Esteban Moro, and Rubén Lara

108. Nonequilibrium spin glass dynamics with magnetic field 162

R. Alvarez Baños, A. Cruz, L. A. Fernández, J. M. Gil-Narvión, A. Gordillo-Guerrero, M. Guidetti,
A. Maiorano, F. Mantovani, E. Marinari, V. Martı́n-Mayor, J. Monforte-Garcı́a, A. Muñoz Sudupe, D.
Navarro, G. Parisi, S. Pérez-Gaviro, J. J. Ruiz-Lorenzo, S. F. Schifano, B. Seoane, A. Tarancón, R.
Tripiccione, and D. Yllanes

109. Sincronización macrocópica de osciladores de Stuart-Landau 163

Ernest Montbrió y Diego Pazó

110. Estudio Computacional de un medio granular húmedo forzado 164

Silvio René Morales Suárez e Ignacio Pagonabarraga Mora

111. Análisis numérico del stress inducido por bombardeo iónico 165
Ana Moreno Barrado y Mario Castro

112. Integrating multiple signals into cell decisions by networks of protein modification cycles 166
Luca Cerone, Javier Muñoz-Garcı́a, and Zoltán Neufeld

113. Macroscopic effects of internal noise on Fisher fronts 167

Svetozar Nesic, Esteban Moro, and Rodolfo Cuerno

114. Digital key for chaos communication performing time delay concealment 168
Romain Modeste Nguimdo, Pere Colet, Laurent Larger, and Luı́s Pesquera

115. How do cells break their symmetry? A simple reaction-diffusion mechanism for cell polar-
ization during asymmetric cell division 169
Ernesto M. Nicola, Philipp Khuc Trong, Nathan W. Goehring, and Stephan W. Grill

116. Dynamical study of the three dimensional Saffman-Taylor problem 170

Matteo Nicoli, Hervé Henry, and Mathis Plapp

117. Unified paradigm for interface dynamics 171

J. C. R. E. Oliveira, P. P. Avelino, and R. Menezes

118. Cooperative tube extraction by single-headed kinesin motors 172

David Oriola and Jaume Casademunt

119. Unraveling spontaneous activity in neuronal cultures 173

Javier G. Orlandi, Enric Álvarez-Lacalle, Sara Teller, Jordi Soriano, and Jaume Casademunt

120. Translational and rotational dynamics in a supercooled molecular liquid 174

Jordi Ortiz de Urbina, Gemma Sesé, and Ricardo Palomar

121. Fluctuaciones de la velocidad en fluidos en flujo 175

José Marı́a Ortiz de Zárate
122. Propiedades interfaciales de fluidos moleculares 176
Jose M. G. Palanco, Jorge Benet y Luis G. MacDowell

123. Collective and individual cell memory in multistable biochemical switches 177
David Palau-Ortı́n and Marta Ibañes

124. Simulación numérica de fluidos Newtonianos y no-newtonianos en cilindros cerrados bajo

forzamientos periódicos 178
Carles Panadès Guinart y Francico Marques

125. Microbes answer to Santa Rosalia: reconsidering the importance of aggregation in ecological
networks 179
Alberto Pascual-Garcı́a, Javier Tamames, and Ugo Bastolla

126. La estructura de redes mutualistas pesadas. Medida del clustering en redes bipartitas 180
Juan Manuel Pastor, Luis J. Gilarranz y Javier Galeano

127. Thresholds for epidemic spreading in networks 181

Romualdo Pastor-Satorras

128. Vectores de Lyapunov caracterı́sticos en sistemas con retraso 182

Diego Pazó y Juan M. López

129. Additivity of Current Fluctuations in Two Dimensions and its Breakdown 183
Carlos Pérez-Espigares, Pedro L. Garrido, and Pablo I. Hurtado

130. Modelling the F1 -ATPase 184

R. Perez-Carrasco and J. M. Sancho

131. La lı́nea de coexistencia hielo seco-CO2 fluido: un análisis a través de simulación por el
método de Monte Carlo 185
G. Pérez-Sánchez, J. M. Mı́guez, D. González-Salgado y M. M. Piñeiro

132. Chiral Selection by Interfacial Shearing of Self-Assembled Achiral Molecules 186

Núria Petit-Garrido, Jordi Ignés-Mullol, Josep Claret, and Francesc Sagués

133. Blood flow in microvessels 187

Clara B. Picallo, Tomás Alarcón, and Aurora Hernández-Machado

134. A microcanonical multifractal approach to the characterization of heartbeat dynamics 188

Oriol Pont, Michel Haı̈ssaguerre, Hussein Yahia, Nicolas Derval, and Méléze Hocini

135. Computación distribuı́da sobre redes complejas. Aplicación a la optimización de tiempos de

ejecución de tareas en entornos GRID 189
Francisco Prieto Castrillo, A. Astillero y Elena Rey Espada

136. Phase Transition of a Meniscus in a Capillary under the Influence of Gravity 190
C. Rascón, A. O. Parry, S. Ivell, A. Thorneywork, and D. Aarts

137. Coexistencia de cooperadores y no cooperadores en poblaciones bien mezcladas: Un escape

del dilema del prisionero en la competición por recursos 191
Rubén J. Requejo y Juan Camacho

138. Towards an understanding of nucleic acids stucture in gas phase 192

Annalisa Arcella and Modesto Orozco

139. Efectos de tamaño iónico finito en suspensiones coloidales concentradas salt-free 193
Rafael Roa, Félix Carrique y Emilio Ruiz-Reina

140. Formulación dinámica del Efecto Casimir 194

Pablo Rodrı́guez-López, Ricardo Brito y Rodrigo Soto

141. Irreversibilidad y disipación en procesos estocásticos 195

Édgar Roldán y Juan M. R. Parrondo
142. Escalamiento de tamaños finitos en la transición de rellenado de cuña del modelo de Ising
tridimensional 196
J. M. Romero-Enrique, L. F. Rull y A. O. Parry

143. Large Scale Modeling of the Self-Assembly of Nano-Objects using a Grid Interpolation
technique 197
Thomas Roussel and Lourdes F. Vega

144. Propagación de Daño y orden ferromagnético en vidrios de spin tridimensionales 198

M. L. Rubio Puzzo, F. Romá, S. Bustingorry y P. M. Gleiser

145. Resonant cycle length polymodality and coherence in a noise-induced genetic oscillator 199
Pau Rué, Jordi Garcia-Ojalvo, and Gürol Süel

146. Dynamical simulations of virus wrapping and budding 200

Teresa Ruiz-Herrero, Michael F. Hagan, and Enrique Velasco

147. Confined aqueous electrolytes within cylindrical nanocavities 201

J. Sala, P. Videla, D. Laria, J. Martı́, and E. Guàrdia

148. Numerical Study of Natural Convection Inside a Horizontal Cylinder 202

O. Sánchez Casals, I. Mercader, and O. Batiste

149. Frequency detuning and communication between brain areas 203

Belén Sancristóbal, Raul Vicente, Gordon Pipa, and Jordi Garcia-Ojalvo

150. Translational and rotational velocity correlations and cumulants in a granular gas 204
Andrés Santos and Gilberto M. Kremer

151. Teorı́a y simulación de Histéresis Angular en gotas sésiles 205

M. J. Santos y J. A. White

152. Crystallization Mechanism of Hard Sphere Glasses 206

Eduardo Sanz, Chantal Valeriani, Emanuela Zaccarelli, Wilson Poon, Peter Pusey, and Mike Cates

153. Network-based scoring system for genome-scale metabolic reconstructions 207

M. Ángeles Serrano and Francesc Sagués

154. Drift of a spiral wave in a heterogeneous heart tissue 208

Ana Simic and Jean Bragard

155. Magneto-rheological properties of stiff magnetic filaments near an adsorbing surface 209
Pedro A. Sánchez, Joan J. Cerdá, Tomás Sintes, V. Ballenegger, and Christian Holm

156. Investigating the folding kinetics in DNA hairpins using molecular constructs with short
and long handles 210
N. Forns, S. de Lorenzo, M. Manosas, K. Hayashi, J. M. Huguet, and F. Ritort

157. Experiments on Patterned Neuronal Cultures 211

Sara Teller, Javier G. Orlandi, Enric Álvarez-Lacalle, Jaume Casademunt, Elisha Moses, and Jordi Sori-
ano

158. Inter-spike correlations induced by dichotomous noise modulation in an excitable laser 212
Jordi Tiana-Alsina, Tilo Schwalger, M. Carme Torrent, Benjamin Lindner, and Jordi Garcia-Ojalvo

159. Transversal dynamics of paramagnetic colloids in a longitudinal magnetic ratchet 213

Pietro Tierno

160. On the Gaussian approximation for master equations 214

R. Toral and L. F. Lafuerza

161. Delivering nutrients to a tissue: blood flow and capillary growth 215
Rui D. M. Travasso, Tobias Schwartz, Eugenia Corvera Poiré, Mario Castro, Juan Carlos Rodrı́guez-
Manzaneque, Orlando Oliveira, Joana Oliveira, and A. Hernández-Machado
162. The role of asymmetric interactions on the effect of habitat destruction in mutualistic
networks 216
Guillermo Abramson, Claudia Trejo Soto, and Leonardo Oña

163. Análisis de la variabilidad multiescala en series temporales de precios de materias primas 217
Antonio Turiel y Conrad Pérez-Vicente

164. Colloids in a bacterial bath: simulations and experiments 218

Chantal Valeriani, Martin Li, John Novosel, Jochen Arlt, and Davide Marenduzzo

165. Divide and Conquer 219

Teresa Vaz Martins and Raúl Toral

166. Effect of polydispersity and soft interactions on the nematic vs. smectic phase stability in
platelet suspensions 220
Y. Martı́nez-Ratón and E. Velasco

167. Malware como depredador en ecosistemas ip 221

Rafael Vida y Elka Korutcheva

168. Crystal polymorphism of a water monolayer under hydrophobic nanoconfinement 222

Oriol Vilanova and Giancarlo Franzese

169. Diffusion in macromolecular crowded media: Monte Carlo simulation of obstructed diffusion
vs. FRAP experiments 223
Eudald Vilaseca, Isabel Pastor, Adriana Isvoran, Sergio Madurga, Josep-Lluı́s Garcés, and Francesc Mas

170. Critical, interfacial and surface properties of ionic liquids by a molecular-based equation of
state 224
O. Vilaseca, F. Llovell, and L.F. Vega

171. Chaos and unpredictability in evolutionary dynamics in discrete time 225

Daniele Vilone, Alberto Robledo, and Angel Sánchez

172. La ecuación de Kuramoto-Sivashinsky estocástica pertenece a la clase de universalidad

Kardar-Parisi-Zhang en dos dimensiones 226
Matteo Nicoli, Edoardo Vivo y Rodolfo Cuerno

173. Variational formulation for the KPZ equation: consistency, Galilean-invariance, and other
issues in real-space discretization 227
H.S. Wio, J.A. Revelli, C. Escudero, R.R. Deza, and M. S. de La Lama

174. Crowd Synchrony and Quorum Sensing Transition in Star-coupled Non-identical Semicon-
ductor Lasers with Time Delay 228
Jordi Zamora-Munt, C. Masoller, Jordi Garcia-Ojalvo, and Rajarshi Roy

175. Tapping vertical en una capa granular 229

José Damas, Iker Zuriguel y Diego Maza

176. Thermodynamics of RNA hybridization inferred from out of equilibrium unzipping experi-
ments 230
C. V. Bizarro, J. M. Huguet, and F. Ritort

177. How proteins fold?: Unfolding/folding of the single protein barnase induced by mechanical
forces 231
B. Rey Serra, A. Alemany, S. Frutos, C. Larroy, C. Ceconni, and F. Ritort

178. Force spectroscopy of individual molecular motors 232

L. Bongini, M. Manosa, C. V. Bizarro, and F. Ritort

179. Entropy production and coarse graining in Markov processes 233

Simone Pigolotti, Andrea Puglisi, Lamberto Rondoni, and Angelo Vulpiani
IV Índice de Autores 235

V Asistentes al Congreso 243

 Parte I
Conferencias invitadas
XVII Congreso de Fı́sica Estadı́stica FisEs11 19

Thermodynamics of a small system in a µ,T reservoir

Dick Bedeaux∗
Department of Chemistry, Norwegian University of Science and Technology,
Trondheim 7491, Norway

Due to advances in experimental techniques operating these properties can be computed from fluctuations at
at the nanoscale, it is possible to compute properties from the nanoscale. We find a 1/L finite size effect for all
density fluctuations by studying “snapshots” of particle thermodynamic quantities for a small system in contact
configurations. Thermodynamics on a small scale is dif- with a reservoir, where L is the length of the system in
ferent from thermodynamics in bulk systems. We show a single dimension.
how the molar enthalpy h and the inverse thermodynam-
ic correction factor Γ−1 depend on system size and how
∗
[Link]@[Link]

Junio de 2011, Barcelona Conferencia Invitada I–1

20 XVII Congreso de Fı́sica Estadı́stica FisEs11

La teorı́a del coarse-graining, aplicaciones y problemas abiertos

Pep Español∗
Departamento de Fı́sica Fundamental
Universidad Nacional de Educación a Distancia
28040 Madrid

Con el propósito de simular de manera eficiente sis- • Transitividad. El resultado de pasar de un nivel
temas moleculares de muchos átomos durante tiempos microscópico 0 a un nivel mesoscópico 1 y de éste
largos se suelen introducir modelos de grano grueso de a otro nivel macroscópico 2, debe producir idéntico
los cuales se espera que capturen las caracterı́sticas más resultado que si obtenemos el nivel macroscópico 2
relevantes del sistema en cuestión a costa de detalles que directamente a partir del microscópico 0.
se espera sean poco importantes. De esta forma, por
ejemplo, se suelen introducir potenciales efectivos que se
obtienen a partir de la función de distribución radial de
centros de masa. Sin embargo, es bien conocido que es-
tos potenciales efectivos reproducen bien las propiedades
de equilibrio (por construcción) pero suelen dar resulta-
dos muy pobres para las propiedades dinámicas (difusión,
correlaciones, etc.)
En este trabajo partimos de la teorı́a de Zwanzig de
operadores de proyección1. Esta teorı́a constituye una
teorı́a completa y práctica2 para la formulación de de-
scripciones de grano grueso y es, de hecho, una formu-
lación muy elegante de la mecánica estadı́stica fuera de
equilibrio. Mostraremos cómo se pueden construir con es-
ta teorı́a modelos de grano grueso a partir de la dinámica
microscópica del sistema en dos ejemplos en particu-
lar: polı́meros estrella descritos en términos de sus cen-
tros de masa (ver figura)2 y modelos discretos para la
hidrodinámica2 ,3 . Figura 1. Descripción de grano grueso de un fundido de
Sin embargo, existen varios problemas abiertos que polı́meros estrella
necesitan ser abordados para que la construcción genérica
de modelos de grano grueso:
• Cómo seleccionar las variables de grano grueso.
La teorı́a nos dice, dadas las variables de grano ∗
grueso, cómo se mueven, pero no ofrece ningún pep@fi[Link]
1
método para “detectar” cuales pueden ser las vari- Memory Effects in Irreversible Thermodynamics,
[Link], Phys. Rev. 124, 983 (1961)
ables apropiadas. 2
Mori-Zwanzig formalism as a practical computational tool,
• Efectos no-Markovianos, Una de las hipótesis Carmen Hijón, Pep Español, Eric vanden-Eijnden, Rafael
básicas en el formalismo es la separación de es- Delgado-Buscalioni, Faraday Discussions 144 301-322
calas temporales entre las variables seleccionadas (2010).
3
y el resto. Esto no siempre se cumple. Microscopic derivation of discrete hydrodynamic vari-
ables, P. Español Jesús G. Anero, e I. Zúñiga, J. Chem.
• La maldición de la dimensionalidad: Los términos Phys.131, 244117 (2009)
de arrastre y difusión de la ecuación de Fokker- 4
On the definition of discrete hydrodynamic variables, P.
Planck obtenida en la teorı́a de Zwanzig dependen Español e I. Zúñiga, J. Chem. Phys. 131, 164106 (2009)
de todas las variables relevantes y son funciones 5
Obtaining fully dynamic coarse-grained models from MD,
en un espacio de muchas dimensiones, difı́cil de P. Español e I. Zúñiga, aceptado en Physical Chemistry
muestrear5 . Chemical Physics DOI:10.1039/C0CP02826F.

Conferencia Invitada I–2 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 21

Frustrated nematic order in spherical geometries

Alberto Fernandez-Nieves∗
School of Physics
Georgia Institute of Technology

When nematic liquid crystals are confined between function of thickness inhomogeneity. In addition, we will
spheres to form shells, complex defect structures emerge. also briefly discuss our recent approach to the generation
These structures are characterized by a varying number of non-spherical surfaces, such as a torus2 , which we hope
of point defects and disclination lines, all complying with to use in the near future to address frustration in these
the topological constraints imposed by the spherical ge- closed surfaces.
ometry. Interestingly, even if the shell thickness per se is
what brings about shells with different number and type ∗
of defects, it is the thickness inhomogeneity what deter- [Link]@[Link]
1
mines the actual defect arrangement1. We will present T. Lopez-Leon, V. Koning, S. Devaiah, V. Vitelli, A.
the rich phenomenology we observe in our shells, where Fernandez-Nieves, Nature Physics (accepted).
2
E. Pairam, A. Fernandez-Nieves, Phys. Rev. Lett. 102,
defects move continuously and/or discontinuously as a
234501 (2009)

Junio de 2011, Barcelona Conferencia Invitada I–3

22 XVII Congreso de Fı́sica Estadı́stica FisEs11

Graphene and its unique properties

Francisco Guinea∗
Instituto de Ciencia de Materiales de Madrid. CSIC

Graphene, two dimensional membrane one atom thick etc. Some of these properties, along with the research to
is a novel material which shows features not found pre- elucidate their origin and consequences, will be reviewed.
viously in other systems. It is a two dimensional metal
whose properties can be tuned, chemically inert, extreme-
∗
ly stiff, charge carriers behave as relativistic particles, [Link]@[Link]

Conferencia Invitada I–4 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 23

Elastic Instabilities Lead To Novel Material Properties

Tom Mullin
Manchester Centre for Nonlinear Dynamics
University of Manchester
Oxford Road, Manchester M13 9PL, UK
tom@[Link]

Periodic elastomeric cellular solids with holes arranged ingly, have recently been shown to also operate at the
on a square lattice give rise a novel uniform transforma- nanoscale. The mechanism has proved to be useful for
tion of the structure when subjected to uniaxial compres- the imprinting of complex patterns in photonic crystals
sion. The original pattern is transformed into a more so that new band gap structures can be induced in a con-
complex one above a critical load. The results of a nu- trolled way. Controllable, negative Poisson ratio effects
merical investigation reveal that the pattern switch is have also been revealed and these also have potential over
triggered by a reversible elastic instability. The phenom- a wide range of scales
ena were discovered at the mm length-scale and, surpris-

Junio de 2011, Barcelona Conferencia Invitada I–5

24 XVII Congreso de Fı́sica Estadı́stica FisEs11

On the short range order of liquid phases

L. C. Pardo, M. Rovira-Esteva, S. Pothoczki, A. Ottochian, J. Ll. Tamarit

1
Grup de Caracterització de Materials, Departament de Fı́sica i Enginyeria Nuclear, ETSEIB, Universitat Politècnica de
Catalunya, Diagonal 647, 08028 Barcelona, Catalonia, Spain

The structure of disordered systems and its implication number of molecules. Finally we account for the crucial
in their dynamics is far to be understood. In the case of role of conformational disorder on the interplay between
the liquid phase, for example, it was believed that the inter- and intramolecular structure. To highlight such a
structure had no effect on the dramatic slowing down of case, we will show recent results on a Freon derivative
molecules giving rise to a glass. Actually the opinion with two conformers3, trans, devoid of dipolar moment,
that liquids can not be described only by their density, and gauche, with a dipolar moment of 0.26 D .
but also by an order parameter characterizing their short
range order is growing. In this work we present a way to
quantify the short range order of molecules in disordered
systems by means of angular bivariate analysis.
The positional order is determined defining an axis sys-
tem having into account the molecular symmetry and
studying the probability distribution function of the az-
imuthal and equatorial angles for the centers of mass of
the neighboring molecules(an example can be seen in the
figure). On what concerns the orientational order it de-
pends on molecular position. To have this dependence in- Figura 1. Bivariate probability P(cosθ,φ) of finding a first
to account we study separately the relative orientation of neighbour molecule around a reference one
two molecules as a function of their position. We present
three examples where the aforementioned method has
been successfully used.
The first example is devoted to the effect of dipolar ∗
[Link]@[Link]
moment in the molecular ordering of quasitetrahedral 1
L. C. Pardo, J. Ll. Tamarit, N. Veglio, F. J. Bermejo, and
molecules. Choosing CCl4 as a reference1 molecule de- G. J. cuello, Phys. Rev. B 76, 134203 (2007)
void of dipolar moment. We disentangle the effect of the 2
M. Rovira-Esteva, A. Murugan, L. C. Pardo, S. Busch,
electrostatic interaction on CCl3 Br and CCl2 Br2 . The M. D. Ruiz-Martı́n, M.-S. Appavou, J. Ll. Tamarit, C.
second example concerns the structure of the high and Smuda, T. Unruh, F. J. Bermejo, G. J. Cuello, and S. J.
low temperature liquids for trans-dichloroethylene2. In Rzoska, Phys. Rev. B 81, 092202 (2010)
this case, in order to study the differences in the molec- 3
L.C. Pardo, F.J. Bermejo, [Link]. Tamarit, G.J. Cuello, P.
ular ordering it was necessary to study separately the ef- Lunkenheimer, and A. Loidl, J. Non-Cryst. Solids 353,
fect of two configurations in the molecular coordination 999 (2007)

Conferencia Invitada I–6 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 25

Multiscale mobility networks and the large scale spreading of infectious diseases

José J. Ramasco∗
Instituto de Fı́sica Interdisciplinar y Sistemas Complejos IFISC (CSIC-UIB), 07122 Palma de Mallorca, Spain
Complex Networks Lagrange and Epidemiology Laboratories, ISI Foundation, Viale S. Severo 65, 10133, Torino, Italy

Epidemiology is one of fields where the application of provide a global description of commuting patterns up
the theory of complex networks is crucial. Infectious dis- to 300 kms; (ii) we integrate in a worldwide structured
eases such as the influenza or the HIV transmit from metapopulation epidemic model population level data,
person to person using mainly connections of social net- airline mobility data and the commuting information as
works. Whether a disease propagates across the world or illustrated in Figure 1. The different time-scales of the
is restricted to a localized area depends on the structure mobility processes are integrated by using a time-scale
and properties of these networks. Contact networks are separation approach for evaluating the force of infection
characterized by a multiscale structure with people in- due to multiscale mobility processes in the disease dy-
teracting within local areas and displacing by car, bike, namics. This is the origin of the GLobal Epidemic and
train, etc, to nearby communities where they can carry Mobility (GLEaM) modeler for the spreading of infec-
or contract a disease. This propagation mode reminds tious diseases1–3 . Commuting flows are found, on av-
thus of a spatial diffusion process. However, a new ele- erage, to be one order of magnitude larger than airline
ment has been introduced during the last century. This flows. However, their introduction into the worldwide
new factor is airplane traveling that allows for fast con- model shows that the large scale pattern of the simulat-
nection of very far apart geographical areas. The recent ed epidemic exhibits only small variations with respect to
propagation of the H1N1 influenza epidemic until becom- the baseline case where only airline traffic is considered.
ing a flu pandemic offers a good example of the effect The presence of short range mobility increases however
that long-range traveling has in the global spreading of the synchronization of subpopulations in close proximity
an infectious disease. The outbreak was first detected in and affects the epidemic behavior at the periphery of the
Mexico City in mid March as an anomalous increase in airline transportation infrastructure.
the number of cases of influenza, the Mexican authori- Apart from these results, the GLEaM modeler has
ties recognized the circulation of a new strain on April been also used in several other applications(which will
23, just after the analysis performed by the American be briefly described) such as the estimation of the disease
Center of Disease Control (CDC), and for then it had parameters of the the H1N1 pandemic4 , the inference of
already extended to the US, and would arrive at Spain the level of use of antibiotics during the pandemic5 , a
a few days later on April 28. An extraordinary propaga- comparison between metapopulation models and agent-
tion velocity crossing thousands of kilometers in a matter based models6 or the assessment of the efficacy of travel
of weeks. restrictions to delay the propagation of a pandemic dis-
ease7 .

∗
jramasco@ifi[Link]
1
D. Balcan, V. Colizza, B. Gonçalves, H. Hu, J.J. Ramas-
co and A. Vespignani, Proc. Natl. Acad. Sci. (USA) 106,
21484 (2009).
2
D. Balcan, B. Gonçalves, H. Hu, J.J. Ramasco, V. Coliz-
za, and A. Vespignani, Journal of Computational Science
Figura 1. Sketch showing the different layers forming the 1, 132 (2010).
3
GLEaM modeler. The application is available online at
[Link]
4
D. Balcan et al., BMC Medicine 7, 45 (2009).
In order to study the interplay between small-scale hu- 5
P. Bajardi et al., Emerging Health Threats Journal 2, e11
man mobility (commuting flows) and long-range airline (2009).
traffic in shaping the spatio-temporal pattern of a glob- 6
M. Ajelli et al., BMC Infectious Diseases 10, 190 (2010).
7
al epidemic we (i) analyze mobility data from 30 coun- P. Bajardi, C. Poletto, J.J. Ramasco, M. Tizzoni, V. Col-
tries around the world and find a gravity model able to izza and A. Vespignani, PLoS ONE 6, e16591 (2011).

Junio de 2011, Barcelona Conferencia Invitada I–7

26 XVII Congreso de Fı́sica Estadı́stica FisEs11

First-principles simulation of biomolecular processes

Carme Rovira∗
†
Computer Simulation & Modeling Laboratory (CoSMoLAB), Parc Cientı́fic de Barcelona, Baldiri Reixac 10-12, 02028
Barcelona, Spain and Institut de Quı́mica Teòrica i Computacional (IQTCUB).
‡
Institució Catalana de Recerca i Estudis Avançats (ICREA), Passeig Lluı́s Companys 23, 02010 Barcelona

Molecular (atomistic) simulation (MS) is a rapidly composition of hydrogen peroxide by catalase enzymes4
growing area of computer science that is having an enor- and the study of conformational free energy landscapes
mous impact on science and technology, because of the of carbohydrates.5–7
increasing need to understand the microscopic details of
matter and because of the phenomenal growth in comput-
∗
ing power in recent years.1 Molecular dynamics (MD) is crovira@[Link]
1
one of the most used MS techniques. In MD, some form W. F. van Gunsteren, D. Bakowies, R. Baron, I. Chan-
for the interaction among atoms needs to be assumed. drasekhar, M. Christen, X. Daura. P. Gee, D. P. Geerke,
This is at the same time a strength and a limitation, A. Glättli, P. H. Hünenberger, M. A. Kastenholz, C. Oost-
especially for the study of chemically complex processes enbrink, M. Schenk, D. Trzesniak, N. F. A. van der Vegt
such as when covalent bonds break up and new ones form and H. B. Yu, Angew. Chem. Int. Ed. 45, 4064 (2006).
2
(e.g. chemical reactions) and in general for those process- R. Car and M. Parrinello, Phys. Rev. Lett. 55, 2471 (1985)
3
C. Rovira, in Methods in Molecular Biology, 305, Protein-
es in which significant electronic reorganizations or po-
Ligand Interactions: Methods and Applications, ed.: G. U.
larization play a role. One of the most powerful methods
Nienhaus, Humana Press, Totowa NJ, pp. 527-566 (2005).
currently available to describe such complex processes is 4
M. Alfonso-Prieto, X. Biarnés, P. Vidossich and C. Rovira,
first-principles or ab initio molecular dynamics (AIMD), J. Am. Chem. Soc. 131, 11751 (2009).
which is based on Density Functional Theory. Since the 5
A. Ardèvol, X. Biarnés, A. Planas and C. Rovira, J. Am.
pioneering work of Car and Parrinello in 1985,2 AIMD Chem. Soc. 132, 16058 (2010).
experienced a rapid development, being now commonly 6
X. Biarnés, J. Nieto, A. Planas and C. Rovira, J. Biol.
used in many areas of science. In this talk I will show Chem. 281, 1432 (2006).
a few recent applications performed in our group in the 7
L. Petersen, A. Ardèvol and C. Rovira, P. Reilly, J. Am.
fields of biochemistry and biophysics,3 such as the de- Chem. Soc. 132, 8291(2010).

Conferencia Invitada I–8 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 27

¿ Cuan de normales son las anomalı́as en el transporte y difusión de partı́culas

brownianas ?

José M. Sancho
Departament d’Estructura i Constituents de la Matèria, Facultad de Fı́sica, Universidad de Barcelona

Desde el trabajo seminal de Einstein1 se sabe que el - Subdifusión: D(τ ) decae en el tiempo.
transporte (o bien la velocidad de arrastre de partı́culas - Superdifusión: D(τ ) aumenta en el tiempo.
en un fluido) no es independiente de la dispersión o di- - Subtransporte: v(τ ) decae en el tiempo.
fusión de las mismas. Sabemos que esta velocidad es Las anomalı́as fuertes pueden verse en la figura adjun-
debida a una fuerza y por la tanto es una propiedad ta.
determinista que deberı́a explicarse simplemente con la
mecánica newtoniana. En cambio la difusión es una con- 0.5
τ
secuencia de las fluctuaciones térmicas y para partı́culas 10
1
10
1

libres bajo una fuerza constante F es, τ

0.4

kB T dv 10
0
10
0

D= = kB T , (1)
λ dF

D(τ)
v(τ)
-1 -1
10 10
donde kB T es la energı́a térmica disponible para la −0.3
τ
partı́cula, y γ es el coeficiente de fricción de Stokes. La
-2 -2
segunda igualdad es una generalización que nos dice que 10 10

la difusión es proporcional al coeficiente de transporte

obtenido usando mecánica clásica: v = F/γ. La expre- 10
-3
−0.7
−0.4
τ 10
-3

τ
sión anterior es muy llamativa pero como veremos solo
1 2 3 4 5 6 1 2 3 4 5 6
se cumple en el caso particular resuelto por Einstein. 10 10 10 10 10 10 10 10 10 10 10 10
τ τ
Sistema normales: Si considerarı́amos como sis- Figura 1. Izquierda: transporte normal y subtransporte.
temas normales los que obedecen la expresión de Einstein Derecha: difusión normal, super y subdifusión5 . Cada lı́nea
entonces pocos sistemas normales existen. Sobre la base corresponde a una longitud de correlación del desorden que sin
de que las partı́culas a una temperatura finita experimen- embargo es débil: σ = 0.05. A menor longitud de correlación
tan fluctuaciones térmicas y que bajo una fuerza F fija mayor es la anomalı́a.
se mueven, definimos el transporte y la difusión como los
siguiente promedios estadı́sticos,
< x(τ ) > < Δ2 x(τ ) > El origen de todas las anomalı́as es la presencia de un
v = lim , D = lim . (2) potencial no lineal que puede ser periódico, débilmente
τ →∞ τ τ →∞ 2τ
desordenado o totalmente aleatorio. Sin embargo basta
Si estos lı́mites existen sin problemas se considera que una dinámica simple como la ecuación de Langevin sobre
el sistema es normal. Hay que aclarar aquı́ que en un amortiguada,
experimento o simulación no alcanzamos tal lı́mite pero
esperamos obtener un valor asintótico estable. Si por γ ẋ = −U  (x) + F + ξ(τ ), (3)
el contrario observamos algún comportamiento diferente
entonces podemos pensar en una anomalı́a que clasifi- para ver una gran variedad de anomalı́as. ξ(τ ) es el cono-
camos en, débiles o fuertes. cido ruido térmico. El potencial no lineal está separado
Anomalı́as débiles en dos partes, U (x) = (1 − σ)Vp (x) + σVr (x): la primera
Estas son de varios tipos: es un potencial periódico y la segunda es un potencial
- Difusión gigante o fuerte aumento de D para ciertos aleatorio (desorden). El parámetro σ controla el peso
valores de F. relativo de cada parte. La simulación numérica de la
- Difusión nula a pesar de que la temperatura es finita. ecuación de Langevin depara muchas más sorpresas.
- Aparición de una componente ortogonal de la veloci-
Todo este escenario será explicado en la charla ası́ co-
dad respecto de la fuerza F . Este efecto es el responsable mo los fundamentos teóricos de cada anomalı́a2–5 .
de la separación de partı́culas diferentes para una misma
fuerza (“sorting”).
Anomalı́as fuertes 1
A. Einstein, Ann. Phys. 17, 549 (1905)
Estas se caracterizan por el hecho de que la velocidad 2
J.M. Sancho, et al. PRL92,250601(2004)
o la difusión no son constantes en el tiempo. Aparecen 3
A.M. Lacasta, et al. PRL94,160601 (2005)
4
cuando el sistema presenta desorden. Tenemos los si– K. Lindenberg, et al. PRL98,020602(2007)
5
guientes casos: M. Khoury, et al. PRL106, 090602(2011)

Junio de 2011, Barcelona Conferencia Invitada I–9

28 XVII Congreso de Fı́sica Estadı́stica FisEs11

The emergent dynamics of collective cell migration

Xavier Trepat1,2,3∗
1
Institut de Bioenginyeria de Catalunya
2
Facultat de Medicina, Universitat de Barcelona
3
Institució Catalana de Recerca i Estudis Avançats

In essential physiological functions including morpho- heterogeneities, both in space and in time, that emerge
genesis, wound healing, and tissue regeneration, the spontaneously, propagate over great distances, and co-
prevalent mode of cellular migration is innately collec- operate over the span of many cell bodies. To explain
tive. Collective cellular migration is also recognized as the severe ruggedness of this force landscape and its role
being a ubiquitous mechanism of invasion in epithelial in collective cell guidance, the well know mechanisms of
cancers. In order for a cell collective to migrate cohe- chemotaxis, durotaxis, haptotaxis are clearly insufficient.
sively, it has long been suspected that each constituent In a broad range of epithelial and endothelial cell sheets,
cell must exert physical forces not only upon its extra- collective cell migration is governed instead by a newly
cellular matrix but also upon its neighboring cells. I will discovered emergent mechanism of innately collective cell
present the first comprehensive maps of these distinct guidance − phlithotaxis.
force components. Further, I will provide evidence that
physical forces within the monolayer develop dramatic
∗
xtrepat@[Link]

Conferencia Invitada I–10 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 29

ZCAM, Spanish node of the CECAM network

Michel Mareschal∗
Zaragoza Scientific Center for Advanced Modeling (ZCAM)

CECAM’s workshops and tutorials have been given a discussed.

reputation of scientific excellence and have been associ-
ated with the leading position of European scientists in
computational condensed matter. The growing impor-
tance of this field had led to creating Cecam nodes in
Europe, among which Zcam, the Spanish node located
in Zaragoza. In this short presentation, I will list present
opportunities Zcam is offering the community, both in
∗
terms of instruments and in terms of topics currently mmaresch@[Link]

Junio de 2011, Barcelona Conferencia Invitada I–11

Parte II
Orales
XVII Congreso de Fı́sica Estadı́stica FisEs11 31

Delayed Coupling Theory of Vertebrate Segmentation

Saúl Ares1∗ , Luis G. Morelli1,2 , Leah Herrgen,2 Christian Schröter,2 Frank Jülicher,1 Andrew C. Oates2
1
Max Planck Institut for the Physics of Complex Systems, Dresden, Germany
2
Max Planck Institute for Molecular Cell Biology and Genetics, Dresden, Germany

The body plan of all vertebrate animals has a segment- end of the system. The posterior boundary is extended
ed organization that is reflected in the repeated arrange- towards one side by the addition of new oscillators at a
ment of vertebra and ribs. This structure forms during rate v/a, where v is an extension velocity and a the dis-
the development of the organism by a process termed tance between neighboring cells. We consider a frequency
segmentation. The segments —called somites— form se- profile which is moving together with the expanding end.
quentially along a linear axis, one by one, with a precise-
ly controlled timing, see Fig. 1. This timing is set by a A B
genetic clock, is realized by oscillations of the levels of

period 2π/Ω (min)

28
certain proteins in individual cells1 . The genetic oscilla-
1.2

T/T(0)
wt
tions of cells in the tissue are coordinated by a molecular 24
des
aei 1.1
signaling system that introduce a coupling of neighboring sat. DAPT
mib
cellular oscillators. This gives rise to a collective spatio- wt +Mib 1.0
20
temporal pattern which consists of waves that travel and 15 20 25 30 0 20 40 60 80 100
delay τ (min) DAPT concentration (μM)
eventually stop and arrest in a periodic arrangement of Figura 2. (A) Collective period Ω as a function of time de-
somites. Signaling gradients ranging over larger distances lay τ for different coupling strengths (solid lines). The sym-
control the slow down and arrest of the cellular oscillators bols indicate operating points for wild-type and different mu-
and guide spatio-temporal patterns during segmentation. tants as indicated. (B) Experimentally determined collective
We have developed a theoretical description of somitoge- period T = Ω/2π (symbols) as a function of DAPT concentra-
nesis based on a coarse grained representation of cellular tion, which is a drug that influences coupling strength. The
oscillators as phase oscillators. Slow intercellular commu- theoretical prediction of the delayed coupling theory is shown
nication introduces a time delay in the coupling between as a solid line.
oscillators.

A LATERAL B DORSAL
Our theory makes key predictions regarding the effects
VIEW VIEW
head le of coupling and coupling delays on the collective oscilla-
nc y profi
freque tor patterns. After an initial transient dynamics, the
system settles in a spatio-temporal limit cycle with col-
segmen

lective frequency Ω which obeys the relation

v
Ω = ωA − ε sin(Ωτ ).
ts

v anterior posterior (2)

tail PSM
PSM
Figura 1. (A) Schematic lateral view of a zebrafish em- This frequency is governed by the autonomous frequency
bryo showing formed segments (purple) and waves of gene ωA of the fastest oscillators at the posterior side, modified
expression (blue) in the unsegmented tissue, the presomitic by effects of coupling described by the coupling strength
mesoderm (PSM). The tail grows with velocity v. (B) From ε. This implies that changes in coupling strength would
a dorsal view the PSM is a U-shaped tissue. A frequency lead to changes in oscillation period and thus in varia-
profile along the PSM causes faster genetic oscillations in the tions of the wavelength of cyclic gene expression patterns
posterior PSM. as well as the resulting segment length. Furthermore,
the theory predicts the existence of a dynamic instabil-
ity when the delay is reduced. We have experimentally
The spatio-temporal patterns of genetic oscillations are verified all these predictions3 . In this way we have: a)
described by coupled sets of phase oscillators which are discovered the first period mutants of the segmentation
arranged in space. The state of a single oscillator is char- clock (see Fig. 2); b) made the first characterization, to
acterized by the phase θi (t), where i labels the oscillator. our knowledge, of a natural system of coupled oscillators
The dynamic equations for the phases are given by2 where coupling delay plays a significant role.
εi (t) 
θ̇i (t) = ωi (t) + sin [θj (t − τ ) − θi (t)] (1)
ni j ∗
saul@[Link]
1
O. Pourquié, Science 301, 328 (2003).
where the sum is over all neighbors j of cell i. Here, 2
L.G. Morelli, S. Ares, L. Herrgen, C. Schröter, F. Jülicher,
ε denotes the coupling strength and τ is the time delay A.C. Oates, HFSP J. 3, 55 (2009).
3
involved in coupling. We solve these equations in one L. Herrgen, S. Ares, L.G. Morelli, C. Schröter, F. Jülicher,
or two-dimensional space with moving boundary at one A.C. Oates, Curr. Biol. 20, 1244 (2010).

Junio de 2011, Barcelona Contribución Oral O–1

32 XVII Congreso de Fı́sica Estadı́stica FisEs11

Deterministic optical rogue waves

Cristian Bonatto, Cristina Masoller

Departament de Fı́sica i Eng. Nuclear, Universitat Politècnica de Catalunya,
Edif. GAIA, Rambla de Sant Nebridi s/n, 08222 Terrassa, Barcelona, Spain

José R. Rios Leite

Departamento de Fı́sica, Universidade Federal de Pernambuco, 50670-901, Recife, Brazil

Michael Feyereisen, Stéphane Barland, Massimo Giudici and Jorge R. Tredicce

Université de Nice Sophia Antipolis, Institut Non-Lineáire de Nice, UMR 6618, 06560 Valbonne, France

Rare extreme events can be observed in different sys-

tems in nature. A typical example are rogue waves ob-
served in the oceans, where waves higher than 30 me-
ters are more or less common phenomena. This fact is in
contradiction with the Gaussian models often used to de-
scribe fluctuations of the wave height in the sea1,2 . Scien-
tific interest on extremely high waves increased substan-
tially during the last decade not only in oceanographic
studies but also in other systems such as capillary waves3
and optical waves4–7 . Both, from the theoretical and
from the experimental points of view there are several
questions still remaining unsolved. The physical mech- Figura 1. Time trace for the laser intensity showing the
anisms that originate them, the way they develop, the occurrence of a large rare event obtained by (a) measuring
probability for them to occur , the type of system able experimentally the laser intensity and (b) simulating the the-
to generate such extreme events, and the connections be- oretical model.
tween extreme events in systems which are apparently
completely different, are being the subjects of intensive
research.
In this work we investigate, both experimentally and
theoretically, the appearance of rare giant pulses or rogue
waves in a semiconductor laser subject to optical injec-
tion. We perform a detailed experimental characteriza-
tion of the parameter region where rogue waves appear,
and compare the experimental observations with numer-
ical results from the simplest rate-equation model.
A typical experimental time series, where a sporadic
large intensity pulse is observed, is shown in Fig. 1(a). Figura 2. Experimental histograms obtained for the laser
Figure 1(b) shows a numerical time series obtained for intensity showing the cases (a) without rogue waves and (b)
the laser model, in a good qualitative agreement with the with rogue waves.
experiment. To investigate the rarity of the large pulse
events, histograms for the laser intensity were measured,
as shown in Fig. 2. Figure 2(a) shows a typical histogram
for a time series without rogue waves and Fig. 2(b) shows
an histogram with rogue waves. In this case, the prob- 1
M. S. Longuet-Higgins, Phil. Trans. Roy. Soc. A 249, 321
ability distribution function of the pulse amplitude dis- (1957).
2
plays clear non-Gaussian features, with a long tail and an S. Aberg and G. Lindgren, Prob. Eng. Mech. 23, 359
abnormality index that confirm the rogue wave character (2008).
3
of the intensity pulsations. M. Shats, H. Punzmann and H. Xia, Phys. Rev. Lett. 104,
The analysis of the theoretical model allows to discuss 104503 (2010).
4
the main mechanisms associated with the appearance of D. R. Solli, C. Ropers, P. Koonath, B. Jalali, Nature 450,
rogue waves. Parameter regions where rogue waves occur 1054 (2007).
5
are identified and is shown that the rogue waves can be J. M. Dudley, G. Genty and B. J. Eggleton, Opt. Express
16, 3644 (2008).
understood as a result of a deterministic nonlinear pro- 6
A. Mussot, A. Kudlinski, M. Kolobov, E. Louvergneaux,
cess. The role of noise in the system is investigated and
M. Douay and M. Taki, Opt. Express 17, 17010 (2009).
its influence to induce or inhibit rogue waves is discussed. 7
A. Montina, U. Bortolozzo, S. Residori, and F.T. Arecchi,
Phys. Rev. Lett. 103, 173901 (2009).

Contribución Oral O–2 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 33

Instabilities of conducting fluid layers in cylindrical cells under external forcing

Javier Burguete∗ , Montserrat Miranda

Departamento de Fı́sica y Matemática Aplicada, Universidad de Navarra
Edificio Los Castaños, Irunlarrea 1, E-31008 Pamplona

Flows created in neutral conducting fluids remain one magnetic field generates an azimuthal current, that pro-
of the less studied topics of fluid dynamics, despite their duces a radial force. This force produces the destabiliza-
importance both in fundamental research (dynamo ac- tion of the static fluid layer, and a flow is created.
tion, turbulence suppression) and applications (continu- The geometry of the experimental cell is a disc lay-
ous casting, aluminum production, biophysics). er with external diameter smaller than 94 mm, with or
without internal hole. The layer is up to 20mm depth,
and we use an In-Ga-Sn alloy as conducting fluid. There
are no external currents applied on the problem, only an
external magnetic field. This field evolves harmonical-
ly with a frequency up to 10Hz, small enough to not
to observe skin depth effects. The magnitude ranges
from 0 to 0.1 T. With a threshold of 0.01T a dynami-
cal behaviour is observed, and the main characteristics of
this flow have been determined: different temporal res-
onances and spatial patterns with different symmetries
Figura 1. Inner ear model. The equilibrium (spinning) is (azimuthal wavenumbers m=3,4,5,8,...).
controlled by the three circular channels placed on the right To our knowledge there are very few experimental
side and oriented in the three spatial axes. The left part cor- works on this field. Only the group of Y. Fautrelles
responds to the cochlea, where the sound waves are converted in Grenoble has performed some measurements but in
in electrical signals. a strongly non-linear regime1 . On the other hand, our
One of these challenging topics is the interaction be- system allows a much more precise threshold determina-
tween electrolytes and high magnetic fields. This point is tion and spatial symmetry description, so we can com-
relevant in biophysics, because of the secondary effects re- pare with the theory that predicts the instability without
ported by patients in MRI devices. Between these effects threshold2 .
are vertigoes, i.e. the sensation of motion or spinning
when the body is at rest. The equilibrium in the human
body is controlled by the inner ear, in the vestibular cav-
ity, by three toroidal channels filled with endo- and per-
ilymph (K+ or Na+ electrolytes). When an alternating
magnetic field is applied to one of these channels it be-
haves as a coil, an induced current is produced and a
radial Lorentz force acts on the fluid.

Parabolic Mirror

Beam Splitter Camera

Light Source Figura 3. Top view of the fluid layer. The surface is de-
flected presenting an azimuthal m = 5 mode.
Hydrochloric Coils
Acid 5% InGaSn Alloy
Coils
∗
Figura 2. Lateral view of the experimental setup. Very javier@[Link] -
small surface deflections can be easily registered near the [Link]
1
threshold. F. Debray and Y. Fautrelle, Adv. in Turb. IV, in: Appl.
Sci. Res., vol. 51, Kluwer (1993) p 31. F. Ingwiller, F. Bon-
Having in mind this application, but with a simpler nel, Y. Fautrelle, S. Daugan, and J. Etay Actes du 9eme
geometry, we present the effect of a time-dependent mag- FLUVISU (2001) p 18
2
netic field parallel to the axis of circular cavities. As ex- Y. Fautrelle, A.D. Sneyd, European Journal of Mechanics
plained above, due to the Lenz s law, the time-dependent B/Fluids 24 (2005) 91-112

Junio de 2011, Barcelona Contribución Oral O–3

34 XVII Congreso de Fı́sica Estadı́stica FisEs11

Mecánica Estadı́stica del Sistema Inmune

J. Currie1 , M. Castro2 , Grant Lythe1 , C. Molina-Parı́s1

1
Department of Applied Mathematics, School of Mathematics, University of Leeds, Leeds, LS2 9JT, UK
2
Grupo de Dinámica No-Lineal and Grupo Interdisciplinar de Sistemas Complejos (GISC), Escuela Técnica Superior de
Ingenierı́a (ICAI), Universidad Pontificia Comillas, E28015, Madrid, Spain

Los vertebrados superiores tienen complejos sistemas receptor-ligando o el tiempo de vida media de uno de es-
inmunológicos que les permiten sobrevivir en un ambi- os enlaces a nivel molecular las que determinan esta re-
ente lleno de agentes infecciosos y también para luchar spuesta inmune. Nosotros mostramos que dicho proceso
contra comportamientos celulares anómalos como los tu- es puramente estocástico y que por tanto dichas canti-
mores. Las células inmunes permanecen inactivas en un dades de equilibrio no dan una respuesta completamente
individuo sano y entran en acción sólo cuando el cuerpo satisfactoria4.
está en peligro, por ejemplo, durante las infecciones1 .
El sistema inmune es un sistema complejo de células
y moléculas (contiene más de 107 clones distintos) que 6
10
se comunican mediante un mecanismo estándar de con-
tacto célula-célula y mediante la secrección de ciertas
moléculas.
Sin duda, la realización de tareas complejas en 4

D (μm /min)
10
este proceso de respuesta ante la amenaza de agentes
patógenes (aprendizaje, reconocimiento, memoria, . . . )
2
requiere una aproximación propia de la Fı́sica Es-
tadı́stica, máxime teniendo en cuenta que involucra es- 2
10
calas de tiempo tan dispares como los segundos (tiempo
caracterı́stico de sensing) y los dı́as (tiempo caracterı́stico
en el que el cuerpo humano padece una infección)2 .
En esta comunicación se describen distintos problemas 0
10 0 4 8 12
fundamentales desde el punto de vista inmunológico y 10 10 10 10
algunas respuestas desde la Mecánica Estadı́stica. Mass (g/mole)
En particular, se abordan dos problemas, uno a la Figura 1. Diagrama log-log de la difusividad en función de
escala de la célula (y su motilidad) y otro respecto a la masa (tomada de la Ref.5 ). La lı́nea recta discontinua es
la dinámica de los receptores de las llamadas células T un ajuste a la ley fenomenológica D ∼ M −1/3 .
(células del sistema inmune que son producidas en el
timo, de ahı́ su nombre).
En el primer problema, hay un debate interesante
tratando de entender si las células difunden aleatoria- ∗
mente o responden quimiotácticamente a estı́mulos pro- marioc@[Link]
1
C.A. Janeway, P. Travers, M. Walport, and J.D. Capra,
ducidos por otras células.
Immunobiology: the immune system in health and disease,
Como se muestra en la figura 1, una hipótesis de bal-
(Current Biology London, 1996).
ance entre fluctuaciones térmicas y disipación viscosa 2
A. Perelson y G. Weisbuch, Immunology for physicists,
permite estimar la difusividad como función de la masa Rev. Mod. Phys. 69, 1219 (1997).
de la célula y, por tanto, discriminar entre un movimiento 3
M. Castro, A physicist approach to immunology, capı́tulo
puramente difusivo y otro dirigido3 . del libro Current Mathematical Models in T cell Biology,
En el segundo problema, se hace uso de ecuaciones es- editado por G. Lythe y C. Molina-Parı́s (Springer, 2011).
tocásticas (fuera del equilibrio) para responder a la sigu- 4
J. Currie, M. Castro, G. Lythe, C. Molina-Parı́s y E.
iente pregunta: ¿cuál es la cantidad que determina la Palmer, Time is precious, so what does a T cell mea-
decisión de una célula T de responder a un estı́mulo? sure (sense): equilibrium quantities or stochastic events?,
En este contexto, se han postulado diferentes hipótesis preprint (2011).
5
siendo la visión más aceptada la que afirma que son can- R.A. Freitas, Nanomedicine, Volume I: Basic Capabilities,
tidades de equilibrio como las afinidades de la reacción (Austin, TX: Landes Bioscience, 1999).

Contribución Oral O–4 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 35

Nonuniversal results induced by diversity distribution in coupled excitable systems

L. F. Lafuerza, Pere Colet, R. Toral

IFISC, Instituto de Fı́sica Interdisciplinar y Sistemas Complejos, CSIC-UIB, Campus UIB, 07122 Palma de Mallorca

Synchronization phenomena play a prominent role in positive slope, so for ω < 1 increasing disorder one never
many branches of science. They have been analyzed in encounters Regime (ii).
terms of phase models which successfully describe sys- We show that, quite generally, the Ott-Antonsen
tems of weakly coupled limit cycle oscillators. In par- method can be successfully applied to any non-singular
ticular, the Kuramoto model1 has become paradigmatic g(ω) provided one solves numerically an integrodifferen-
to describe the emergence of synchronization in an en- tial equation6 . For the Gaussian distribution the SNIC
semble of oscillators with diversity among the individual starts at ω = 1 with a negative slope (Fig.1(left)). There-
units. Diversity in the oscillators is usually introduced by fore for ω < 1, as disorder increases one finds first regime
taking their natural frequencies from a probability dis- (i), then crosses the SNIC lower boundary entering in
tribution. Although, on general grounds (central limit regime (ii) and finally crosses the reentrant upper bound-
theorem), this distribution should be well approximated ary entering in regime (iii) where a stable steady state is
by a Gaussian form, theoretical studies usually consider present again. Region (ii) moves upwards and broad-
a Lorentzian form since it allows for an easier analyt- ens increasing the coupling K. The same result is ob-
ical treatment. It is generally believed that the main tained using the Ott-Antonsen method solved numeri-
results concerning the global synchronization properties cally (circles) and using an alternative approach we have
are qualitatively independent of the precise form of the developed6 (solid line). Besides the Gaussian, we find
distribution as long as it is symmetric and unimodal. that this reentrance is generic for all distributions with
We consider a variant of the Kuramoto model for an well defined moments6 .
ensemble of globally coupled active rotators2: The diversity induced transition to collective firing is
a genuine transition. Singularly enough it is not present
K
N
φ̇j = ωj − sin φj + sin(φl − φj ) j = 1, ..., N. (1) for a Lorentzian distribution, for which the first moment
N integral is only defined as a principal value. The non-
l=1
generic behavior of the system with a Lorentzian distri-
A natural frequency ωj < 1 (respectively, ωj > 1) corre- bution of natural frequencies warns about its indiscrim-
sponds to an excitable (respectively, oscillatory) behavior inate use in order to understand generic properties of
of the rotator j when it is uncoupled. K is the coupling coupled oscillators.
intensity. Diversity is introduced by considering that the
ωj ’s are distributed according to a probability density
function g(ω), with mean value ω and variance σ 2 .
For ω <∼ 1 the system displays three different regimes:
(i) for small diversity, almost all units are at rest at sim-
ilar fixed points; (ii) increasing diversity one enters a dy-
namical state in which a macroscopic fraction of units fire
at (roughly) the same time; (iii) for even larger diversity,
the system reenters a desynchronized state.
We had developed an approximate theory to describe
this diversity-induced collective firing3 . The theory was Figura 1. (Partial) Bifurcation diagram for Gaussian (left)
independent of the form of the natural frequencies dis- and Lorentzian (right) distributions.
tribution and was also applicable to identical units sub-
ject to noise. A recent method developed by Ott and
Antonsen4 allows to solve exactly this model (and a large
family of related ones) in the infinite number of oscillators
limit and in a number of cases that include the Lorentzian 1
Y. Kuramoto, Chemical Oscillations, Waves and Turbu-
g(ω). Childs and Strogatz5 used this method to obtain lence (Springer-Verlag, New York, 1984).
the full bifurcation diagram of the complex variable r(t) 2
Y. Kuramoto, in International Symposium on Mathemat-
for the Lorentzian distribution. Contrarily to our results, ical Problems in Theoretical Physics, 1975, p. 39.
their exact solution implies that there is no transition to 3
C. J. Tessone, A. Scire, R. Toral, and P. Colet, Phys. Rev.
collective firing increasing the diversity for ω < 1 as il- E 75, 016203 (2007).
lustrated in Fig.1 (right). Regime (ii) takes place for the 4
E. Ott and T.M. Antonsen, Chaos 18, 037113 (2008).
parameter region located to the right of the solid line 5
L.M. Childs and H.S. Strogatz, Chaos 18, 043128 (2008).
6
signaling the SNIC (saddle node on the invariant circle) L.F. Lafuerza, P. Colet and R. Toral, Phys. Rev. Lett.
bifurcation. The SNIC line always starts at ω = 1 with 105, 084101 (2010).

Junio de 2011, Barcelona Contribución Oral O–5

36 XVII Congreso de Fı́sica Estadı́stica FisEs11

Leyes de Potencia y Leyes de Escala en la Distribución de Energı́a de los Huracanes

Álvaro Corral∗ , A. Ossó, J.E. Llebot

Centre de Recerca Matemàtica
Edifici Cc, Campus Bellaterra
E-08193 Barcelona, Spain

The influence of climate variability and global warming ½¼

on the occurrence of tropical cyclones is a controversial


½¼  
issue. Existing historical databases on the subject are  
not fully reliable, but a more fundamental hindrance is ½¼  


 

      
the lack of basic understanding regarding the intrinsic


½¼
nature of tropical-cyclone genesis and evolution. It is
known that tropical cyclones involve more than a passive ½¼  

 
response to changing external forcing, but it is not clear ½¼
which dynamic behaviour best describes them. Here we  
½


present an approach based on the application of the pow-
er dissipation index (P DI), which constitutes an estima- ½¼ 
 
tion of released energy1 , to individual tropical cyclones.

A robust law emerges for the statistics of power dissipa- ½¼

tion index, valid in four different ocean basins and over  

½¼
long time periods. In addition to suggesting a description

of the physics of tropical cyclones in terms of critical phe- ½¼
¼ ¼¼¼½ ¼ ¼¼½ ¼ ¼½ ¼½
nomena, the scaling law enables us to quantify their re-    

 

 

sponse to changing climatic conditions, with an increase Figura 2. Scaling of P DI distributions conditioned
in the largest power dissipation index values with sea sur- to sea surface temperature (SST ) and El Niño. P DI
face temperature or the presence of El Niño phenomenon, probability densities calculated separately for years with high
depending on the basin under consideration. or low SST and for years with M EI > 0 (El Niño) or
M EI < 0 (La Niña). Tropical depressions (storms whose
½¼  maximum vt is below 34 knots, 1 knot = 1.85 km/h) are ex-
cluded from the Northwestern Pacific dataset, in order to give

½¼ all basins the same treatment. Time periods and vertical off-
sets are as in Fig. 1. In all cases the data can be fit by a
½¼
power law, being the worst one that of the North Atlantic
½¼ 
with low SST , which yields α = 1.26 ± 0.08 with a p−value
    


equal to 9%.
½¼


½¼


In this way, we demonstrate that the recent upswing in
½¼
North Atlantic hurricane activity2 does not involve trop-

½¼ ical cyclones that are quantitatively different from those
½¼ in other sustained high-activity periods before 19703.
  

  


½¼   

  
 ∗

acorral@[Link]
½¼ 1
½¼ ½¼ 
½¼ ½¼ 
Emanuel, K. Increasing destructiveness of tropical cy-
  
clones over the past 30 years. Nature 436, 686–688 (2005).
2
Figura 1. Power-law distributions of tropical-cy- Goldenberg, S. B., Landsea, C. W., Mestas-Nuñez, A. M.
clone P DI values. P DI probability densities for tropical & Gray, W. M. The Recent Increase in Atlantic Hurricane
cyclones in the North Atlantic, Northeastern Pacific, North- Activity: Causes and Implications. Science 293, 474–479
western Pacific, and Southern Hemisphere basins. The period (2001).
3
considered is either 1966-2007 or 1986-2007, depending on the Corral, A., Ossó, A. & Llebot, J. E. Scaling of Tropical-
reliability of the records.√ The values in the
√ vertical axis are Cyclone Dissipation. Nature Phys. 6, 693–696 (2010).
divided by the factors 1, 1000, 1000 and 10003 , to separate
the curves for clarity. The distributions are consistent with
a power law (straight lines) over some portion of their range,
with exponents α = 1.19 ± 0.06, 1.175 ± 0.05, 0.96 ± 0.02 and
1.11 ± 0.04, from top to bottom and the Kolmogorov-Smirnov
(KS) test yields p−values larger than 20 % in all basins. De-
viations from the power law at large P DI values reflect the
finite size effect.

Contribución Oral O–6 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 37

Lı́mite a la descripción hidrodinámica en el modelo de Maxwell para gases granulares

J. J. Brey, M. I. Garcı́a de Soria∗ , P. Maynar

Fı́sica Teórica. Facultad de Fı́sica
Universidad de Sevilla
Aptdo. de Correos 1065. 41080 Sevilla

Las ecuaciones hidrodinámicas para gases granulares or del parámetro que caracteriza la inelasticidad, uno de
han sido derivadas, en el lı́mite de baja densidad, a par- los modos cinéticos decáe más lentamente que uno de los
tir de la ecuación de Boltzmann para esferas o discos modos hidrodinámicos. Como consecuencia de esto, no
duros inelásticos1 . Sin embargo, los métodos empleados existe, en este régimen de parámetros, una descripción
son formales y no permiten establecer la existencia de la hidrodinámica cerrada. También se discuten algunas im-
descripción hidrodinámica en sı́ misma. En el caso de plicaciones de este comportamiento en los coeficientes de
esferas o discos duros elásticos, el conocimiento del es- Navier Stokes.
pectro del operador de Boltzmann linealizado2 permite
resolver el problema de forma satisfactoria. En el caso ∗
inelástico, los modos hidrodinámicos han sido identifica- gsoria@[Link]
1
dos en el lı́mite de longitud de onda grande3,4 , pero casi J.J. Brey, J.W. Dufty, C.S. Kim, and A. Santos, Phys.
nada se conoce sobre la parte cinética del espectro. Rev. E, 58, 4638 (1998).
2
J.A. McLennan, Introduction to Nonequilibrium Statistical
En este trabajo, analizamos la validez de la descrip-
Mechanics (Prentice-Hall, Engle-wood Cliffs, NJ, 1989).
ción hidrodinámica usando una ecuación de Boltzmann 3
J.J. Brey, J.W. Dufty, and M.J. Ruiz-Montero, in Gran-
inelástica simplificada, la del modelo de Maxwell in- ular Gas Dynamics, edited by T. Poeschel and N. Bril-
elástico5–7 . Esta ecuación cinética se obtiene a partir de liantov (Springer-Verlag, Berlin, 2003).
la ecuación de Boltzmann inelástica para esferas o discos 4
J.W. Dufty and J.J. Brey, Phys. Rev. E 68, 030302(R)
duros, sustituyendo la frecuencia de colisión dependien- (2003).
te de la velocidad relativa por un valor medio efectivo 5
A.V. Bobylev, J.A. Carrillo, and I.M. Gamba, J. Stat.
proporcional a la velocidad térmica. Phys. 98, 743 (2000).
Las autofunciones del operador de Boltzmann li- 6
E. Ben-Naim and P.L. Krapivsky, Phys. Rev. E 66 011309
nealizado homogéneo para moléculas de Maxwell co- (2002).
7
rrespondientes a los modos hidrodinámicos pueden ser M.H. Ernst and R. Brito, J. Stat. Phys. 109, 407 (2002).
8
calculadas, identificándose también algunos modos no J.J. Brey, M.I. Garcı́a de Soria, and P. Maynar, Phys. Rev.
hidrodinámicos8 . Se muestra que por debajo de un val- E 82, 021303 (2010).

Junio de 2011, Barcelona Contribución Oral O–7

38 XVII Congreso de Fı́sica Estadı́stica FisEs11

Hidden Symmetries in Nonequilibrium Systems

P.I. Hurtado, C. Pérez-Espigares, J. del Pozo, P.L. Garrido

Instituto Carlos I de Fı́sica Teórica y Computacional. Universidad de Granada. Spain.

Fluctuations arise universally in Nature as a reflection time-reversibility on the statistics of nonequilibrium fluc-
of the discrete microscopic world at the macroscopic lev- tuations. The new symmetry implies remarkable hierar-
el. Despite their apparent noisy origin, fluctuations en- chies of equations for the current cumulants and the non-
code fundamental aspects of the physics of the system at linear response coefficients, going far beyond Onsager’s
hand, crucial to understand irreversibility and nonequi- reciprocity relations and Green-Kubo formulae. We con-
librium behavior [1,2,3,4]. In this work we focus on the firm the validity of the new symmetry relation in exten-
large deviation properties of the currents present on sys- sive numerical simulations. This opens an unexplored
tems at nonequilibrium steady states and, in particular, route toward a deeper understanding of nonequilibrium
on the joint statistics of the current and any system con- physics by bringing symmetry principles to the realm of
figurational properties (say the density or energy spatial fluctuations.
distribution) both averaged over a long but fixed time References
interval. When this time interval is set to infinity all [1] Bertini L, De Sole A, Gabrielli D, Jona-Lasinio G,
system observables tend to their corresponding steady Landim C (2001) Fluctuations in stationary nonequilib-
state value. However for long but finite times it is known rium states of irreversible processes. Phys. Rev. Lett.
that the joint statistics exhibits complex structure which 87:040601.
couples a given current fluctuation with a precise value
[2] Bodineau T, Derrida B (2004) Current fluctuations
of the complementary observable studied. The hidden
in nonequilibrium diffusive systems: an additivity prin-
symmetries appear at this level of description. We find
ciple. Phys. Rev. Lett. 92:180601.
that all the isometric values of current fluctuations (e.g.
currents related by rotations) have the same values of [3] Bertini L, De Sole A, Gabrielli D, Jona-Lasinio G,
some associated configurational property. This allows us Landim C (2007) Stochastic interacting particle systems
to derive an isometric fluctuation relation which links in out of equilibrium. J. Stat. Mech. P07014.
a strikingly simple manner the probabilities of any pair [4] Hurtado PI, Garrido PL (2009) Test of the additiv-
of isometric current fluctuations. This relation, which re- ity principle for current fluctuations in a model of heat
sults from the time-reversibility of the dynamics, includes conduction. Phys. Rev. Lett. 102:250601.
as a particular instance the Gallavotti-Cohen fluctuation [5] Gallavotti G, Cohen EGD (1995) Dynamical ensem-
theorem [5] in this context but adds a completely new bles in nonequilibrium statistical mechanics. Phys. Rev.
perspective on the high level of symmetry imposed by Lett. 74:2694-2697.

Contribución Oral O–8 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 39

Prisoner’s Dilemma on a sizeable network: An experiment with human subjects

Jelena Grujić1 , Constanza Fosco1 , Lourdes Araujo2 , José A. Cuesta1 , Angel Sánchez1,3
1
GISC, Dpto. de Matemáticas, Universidad Carlos III de Madrid, Leganés
2
NLP and IR, Dpto. de Lenguajes y Sistemas, UNED, Madrid
3
BIFI, Universidad de Zaragoza, Zaragoza

The mechanisms underlying the emergence of cooper- both heterogeneity and a “moody” conditional coopera-
ation among unrelated individuals are as yet an unsolved tion strategy, in which the probability of cooperating also
puzzle. Many theoretical models have shown that the ex- depends on the player’s previous action, are required to
istence of structure in a population can help understand- understand the outcome of the experiment. These results
ing the widespread emergence of cooperation, particular- could impact the way game theory on graphs is used to
ly in the framework of the Prisoner’s Dilemma (PD), but model human interactions in structured groups.
the results of these models largely depend on details such
as the type of spatial structure or the evolutionary dy-
namics. Therefore, experimental work suitably designed
to address this question is needed to probe these issues. experiment 1
0.5 control
We have designed an experiment to test the emergence experiment 2

fraction of cooperators
of cooperation when humans play PD on a network whose 0.4
size is comparable to that of simulations1 . In our exper-
iment, volunteers played a PD game with each of their 0.3

eight neighbors taking only one action, either to cooper-

0.2
ate (C) or to defect (D), the action being the same against
all the opponents. The resulting payoff was calculated
0.1
by adding all eight interaction payoffs. Payoffs of the PD
game were set to be 7 cents for mutual cooperation, 10 0
0 10 20 30 40 50 60
cents for a defector facing a cooperator, and 0 cents for round
any player facing a defector (i.e., the same weak PD set-
up of Nowak and May2 ). The full experiment consisted
Figura 1. Fraction of cooperators in every round of the
of three parts: experiment 1, control, and experiment 2.
three parts of the experiment. The cooperation level declines
In experiment 1 players remained at the same positions
to a low but non-zero level.
in the lattice with the same neighbors throughout the
experiment. In the control part we removed the effect of
the lattice by shuffling players every round. Finally, in 1.0
experiment 2 players were again fixed on a lattice, albeit after C
different from that of experiment 1. after D
probability of playing C

We find that the cooperation level declines to an

asymptotic state with low but nonzero cooperation (Fig.
1). Regarding players’ behavior, we observe that the pop- 0.5
ulation is heterogeneous, consisting of a high percentage
of defectors, a smaller one of cooperators, and a large
group that shares features of the conditional cooperators
of public goods games, except that the players are not
influenced only by how much cooperation they observe
0.0
but also by how they themselves behaved in the previous 0 2 4 6 8
number of Cs in the neighbourhood
round (Fig. 2). We propose an agent-based model based
on the coexistence of these different strategies that is in
good agreement with all the experimental observations. Figura 2. “Moody” conditional cooperators. Probabilities
The large size of our experimental setup and the da- of cooperating after playing C or D, conditioned to the con-
ta analysis allow us to answer two important questions. text (number of cooperators in the previous round).
First, we have observed that the existence of a lattice
giving structure to a population playing a PD does not
lead to an increase of the cooperation level; the residual 1
Grujić J, Fosco C, Araujo L, Cuesta JA, Sánchez A, PLoS
level is around 20% which is typical in public goods ex- ONE 5, e13749 (2010)
periments and has also been observed on 4×4 lattices3 . 2
Nowak MA, May RM, Nature 359: 826–829 (1992)
Second, regarding the manner in which people update 3
Traulsen A, Semmann D, Sommerfeld RD, Krambeck HJ,
their strategies, we have not found evidence in favor of Milinski M, Proc Natl Acad Sci USA 107: 2962–2966.
imitate-the-best behavior. Our findings also indicate that (2010)

Junio de 2011, Barcelona Contribución Oral O–9

40 XVII Congreso de Fı́sica Estadı́stica FisEs11

Statistical mechanics approaches to complex network inference and reconstruction

R. Guimerà∗ , M. Sales-Pardo∗∗
ICREA and Departament d’Enginyeria Quı́mica
Universitat Rovira i Virgili, Av. dels Paı̈sos Catalans 26
43007 Tarragona

A Real Network analysis is currently used in a myriad of con-

texts, from identifying potential drug targets to predict-
ing the spread of epidemics and designing vaccination
strategies and from finding friends to uncovering criminal
activity. Despite the promise of the network approach,
the reliability of network data is a source of great con-
cern in all fields where complex networks are studied.
We present a general mathematical and computational
framework to deal with the problem of data reliability
in complex networks1. In particular, we are able to re-
B Observed
liably identify both missing and spurious interactions in
≥1.0
Missing link
noisy network observations. Remarkably, our approach
Degree rel. error

0.5 Spureous link

0.0 Correct links not shown also enables us to obtain, from those noisy observations,
-0.5
network reconstructions that yield estimates of the true
-1.0
network properties that are more accurate than those
provided by the observations themselves. We will illus-
trate the methods with examples from systems biology
(metabolome and proteome) and from the social sciences
(the voting patterns of US Supreme Court justices).

∗
C Reconstructed [Link]@[Link]
≥1.0 ∗∗
Missing link
[Link]@[Link]
Degree rel. error

0.5 Spureous link

1
0.0 Correct links not shown
Guimera, R, Sales-Pardo, M., Missing and spurious inter-
-0.5

-1.0
actions and the reconstruction of complex networks, Proc.
Natl. Acad. Sci. U. S. A. 106, 22073-22078 (2009)

Figura 1. Reconstruction of the air transportation network

of Eastern Europe. A, The true air transportation network.
The area of each node is proportional to its betweenness cen-
trality, with Moscow being the most central node in the net-
work. B, The observed air transportation network, which we
build by randomly removing 20% of the real links and replac-
ing them by random links. C, The reconstructed air trans-
portation network that we obtain, from the observed network,
applying the heuristic reconstruction method described in the
text and methods. For clarity, in B (respectively, C) we do
not depict the correct links, but only: (i) missing links in
orange, which exist in the true network but not in the obser-
vation (reconstruction), and (ii) spurious links in blue, which
do not exist in the true network but do exist in the obser-
vation (reconstruction). As in A, the area of each node is
proportional to its betweenness centrality, with the black cir-
cle representing the true betweenness centrality of each node.
The color of each node represents the relative error in the
degree of the node, with respect to the true degree.

Contribución Oral O–10 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 41

Caracterización y Modelización de Sistemas Dinámicos no Lineales. Medidas de

Desorden Dinámico y ”Self-Correlation”.

Juan Antonio Hernández Álvarez∗ , Rosa Marı́a Benito, Juan Carlos Losada
Grupo de Sistemas Complejos, ETSI, Agrónomos
Universidad Politécnica de Madrid
28040 Madrid

A lo largo de los años se han desarrollado multi- provenientes de un sistema de ecuaciones, como en una
tud de técnicas e inventado indicadores destinados a serie temporal obtenida experimentalmente. La técnica
caracterizar sistemas no lineales en general y caóticos básica es la misma. La metodologı́a está definida para el
en particular. Unos pocos, pero importantes ejemplos estudio de sistemas no lineales en general, tanto teóricos
son: el ”Exponente de Lyapunov”1, la ”Entropı́a de como experimentales, independientemente de cualquier
Kolmogorov”2, ”La sección de Poincaré”3 y distintas clasificación a priori del objeto a estudiar.
definiciones de dimensión como puedan ser la ”Fractal”4
y la de Hausdorff5 . Estos indicadores, diseñados para El objetivo perseguido es doble, por un lado caracteri-
hacer mediciones sobre un sistema, reflejan el gran in- zar el sistema y por otro definir modelos fiables constru-
terés existente en la caracterización, que persigue en no idos sobre dicha caracterización. Dos aplicaciones bien
pocas ocasiones la obtención de información sustancial distintas son presentadas para ilustrar la teorı́a general:
que pueda ser útil en la construcción de modelos fiables, la caracterización de un sistema caótico de dos dimen-
siempre con la esperanza de que presenten las mismas siones y la obtención de modelos para series temporales
caracterı́sticas que el sistema original. telefónicas obtenidas de una operadora nacional de tele-
comunicaciones.
Especial mención merece el estudio de aquellos sis-
temas que presentan un comportamiento caótico. El Son relevantes varios aspectos que afloran en la apli-
”Caos Determinista”es una disciplina completamente cación del método expuesto, y que están relacionados
desarrollada como teorı́a en las últimas décadas, aunque con indicadores bien conocidos como son el ”Exponente
ha sido en años recientes cuando ha llegado la verdadera de Lyapunov”y la ”Información mutua”. El ”Producto
avalancha de hallazgos de comportamiento caótico en Escalar”promedio puede ser utilizado como complemen-
multitud de sistemas fı́sicos y sociales. Algunos de estos to del primero y el ”Producto Perpendicular”promedio
descubrimientos han tenido lugar en campos tan dis- puede suponer una mejora del segundo en determinados
pares como puedan ser las series económicas6, el viento7 , sistemas.
los procesadores de computador8 o las poblaciones de
células9.
∗
Aunque la teorı́a de Sistemas Dinámicos no Lineales [Link]@[Link]
1
está firmemente desarrollada desde hace tiempo, nuevos A. Wolf, J. B. Swift, H. L. Swinney and J. A. Vastano,
enfoques han aparecido en fechas recientes. Algunos de Physica D, 16, 285 (1985).
2
ellos son: el estudio de redes complejas derivadas de una A. N. Kolmogorov, Dokl. Akad. Nauk SSSR, 124, 754
serie temporal10 , la combinación de medidas de comple- 3
(1959).
jidad con la entropı́a de Shannon11 y la detección de 4
H. Poincaré, Gauthier-Villars et fils (1892).
patterns prohibidos en series temporales12 . 5
B. B. Mandelbrot, W. H. Freeman & Co. (1983).
F. P. Hausdorff, Math. Annalen, 79, 157 (1919) .
6
D. Guegan, Ann. Rev. Control, 33, 89 (2009).
En este trabajo se pretende explicar una nueva 7
T., E. Karakasidis and A. Charakopoulos, Chaos Solit.
metodologı́a para caracterizar sistemas no lineales, basa-
Fract. 41, 1723 (2009) .
da en la medida de ciertas caracterı́sticas de una serie 8
Z. Halbiniak and I. J. Jozwiak, Chaos Solit. Fract. 31, 409
temporal que hemos denominado ”Dynamical Order”y
(2007).
”Self-Correlation”13. La primera se corresponde con lo 9
M. Laurent, J. Deschatrette and C. M. Wolfrom, PLoS
desordenado que es el movimiento de la serie temporal ONE, 5, 9346 (2010) .
en un espacio de estados bidimensional, mientras que 10
J. Zhang and M. Small, Phys. Rev. Lett. 96, 238701
la segunda es una medida de auto-correlación no lineal. (2006).
Asimismo, se presentan los productos ”Escalar”y ”Per- 11
O. A. Rosso, H. A. Larrondo, M. T. Martin, A. Plastino
pendicular”convenientemente promediados a lo largo de and M. A. Fuentes, Phys. Rev. Lett. 99, 154102 (2007).
la serie temporal, como indicadores para la medición de 12
J. M. Amigo, S. Elizalde and M. B. Kennel, J. Comb.
las mencionadas caracterı́sticas. Theory A, 115, 485 (2008) .
13
J. A. Hernández, R. M. Benito and J. C. Losada, Int. J.
La misma técnica puede ser empleada, tanto en órbitas Bifurcation and Chaos, accepted for publication (2010).

Junio de 2011, Barcelona Contribución Oral O–11

42 XVII Congreso de Fı́sica Estadı́stica FisEs11

Universalidad en el flujo de un medio granular a través de un orificio

A. Janda∗ , I. Zuriguel, A. Garcimartı́n, D. Maza

Grupo de Medios Granulares†
Departamento de Fı́sica y Matemática Aplicada
Facultad de Ciencias, Universidad de Navarra
31080 Pamplona

El flujo de un medio granular a través de un orificio

es un claro ejemplo de las peculiaridades de este tipo
de materiales. Si bien una descripción hidrodinámica de
este proceso pudiera parecer factible, la realidad es que
mediante ella no es posible reproducir rigurosamente el
comportamiento de los granos. En la literatura está am-
pliamente aceptado que el flujo de descarga de un silo
viene determinado por la dinámica de las partı́culas en
la cercanı́a del orificio1 . Aunque se ha especulado so-
bre las escalas caracterı́sticas de este proceso, no existen
pruebas experimentales sobre la forma de los perfiles de
velocidad y fracción de compactación a la salida. Por
otro lado, existe un debate sobre la influencia en el flujo
de la aparición de atascos cuando la salida es de tamaño
comprable a los granos, ası́ como sobre la posible existen-
cia de un tamaño crı́tico del orificio por encima del cual
no se producirı́an2–4 .

Figura 1. (a) Perfiles de la componente vertical de la ve-

locidad a lo largo del orificio de salida. Los distintos sı́mbolos
corresponden a diferentes tamaños de la abertura. (b) Se han
representado los mismos datos que en (a) pero reescalados de
manera apropiada para mostrar su comportamiento universal.

Con el fin de arrojar luz sobre estas cuestiones, se han

estudiado sistemáticamente las propiedades del flujo en
un amplio rango de tamaños de la salida. Durante la
∗
descarga se han determinado experimentalmente con una ajandaga@[Link]
†
gran resolución espacial y temporal los campos de veloci- [Link]
1
dad y fracción de compactación en el orificio. Ası́, se han R. M. Nedderman, “Statics and Kinematics of Granu-
encontrado las escalas caracterı́sticas del flujo y estudia- lar Materials”, Cambridge University Press, Cambridge,
do las propiedades de los perfiles de dichas magnitudes en (1992).
2
función del tamaño de la abertura. Sorprendentemente I. Zuriguel, A. Garcimartı́n, D. Maza, L.A. Pugnaloni and
se obtiene un comportamiento universal del flujo5 y sus J.M. Pastor, “Jamming during the discharge of granular
fluctuaciones, tanto para aberturas pequeñas en las que matter from a silo”, Phys. Rev. E 71 (2005), 030301.
3
A. Janda, I. Zuriguel, A. Garcimartı́n, L.A. Pugnaloni
los atascos son frecuentes, ası́ como para orificios grandes
and D. Maza, “Jamming and critical outlet size in the
donde el flujo es prácticamente continuo. Hasta donde
discharge of a two-dimensional silo”, Europhys. Lett. 84
nosotros sabemos, este resultado es la primera prueba
(2008), 44002.
experimental que demuestra que el flujo es un estado bi- 4
K. To, “Jamming transition in two-dimensional hoppers
en definido del sistema e independiente de la aparición and silos”, Phys. Rev. E 71 (2005), 060301(R).
de atascos. Además, esta caracterı́stica es coherente con 5
Janda Galán, Alvaro. “Análisis experimental del proceso
la no existencia de un tamaño crı́tico de orificio. de descarga de medios granulares en silos.” Tesis de Doc-
torado, Universidad de Navarra (2010).

Contribución Oral O–12 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 43

Natural Correlations in Networks: Origin and Effects

Samuel Johnson∗ , Joaquı́n J. Torres, J. Marro, Miguel A. Muñoz

Dpto. de Electromagnetismo y Fı́sica de la Materia, and
Institute Carlos I for Theoretical and Computational Physics,
Universidad de Granada, 18071 Granada, Spain.

In recent years it has become apparent that a funda- maximizes this magnitude. At equilibrium – i.e., in the
mental aspect of many complex systems is the structure absence of correlating mechanisms – the system can be
of the underlying network of interactions between the el- expected to find itself in this largest volume of phase
ements. The origins and effects of statistical properties space. Contrasting our predictions against empirical da-
such as degree distributions, short paths, clustering or ta, it turns out that the correlations of many networks
community structure have been analysed in great detail. can indeed be explained in this way (for example, the
One such feature of empirical networks is the existence of metabolic, Web page and protein networks in Fig. 1).
correlations between the degrees of neighbouring nodes On the other hand, some networks with identifiable an-
– a node’s degree being the number of neighbours it has. ticorrelating mechanisms (such as the P2P network or
Networks with positive correlations are assortative, since the internet) are more disassortative than if they were
nodes of a kind tend to group together, or assort, whereas at equilibrium, while social networks (the actors graph)
negatively correlated ones are disassortative. This prop- are far more assortative – indicating that there are ho-
erty has been shown to be highly relevant for their be- mophilic processes at work.
haviour. For instance, assortative networks have lower
percolation thresholds and are more robust to targeted
attack, while disassortative ones make for more stable 0.2
ecosystems and are more “synchronizable”.

[Link]

internet
A striking characteristic of empirical networks is that

protein
they are seldom uncorrelated. Rather, social networks –
0

metabolic
in which the nodes are people and the edges represent
r*

actors
professional, sexual or any other form of interaction –
are usually assortative. Conversely, almost all other net-
P2P

works, whether biological (genetic, ecological, neural...), -0.2

information-related (linguistic, the Web...), or technolog-
ical (the internet, power-grids...), are significantly disas-
sortative. -0.4
Whence these non-trivial correlations? Do they, in 2 2.25 2.5
each case, serve some functional purpose? Or do they γ
share a common origin? The positive correlations of so- Figura 1. (Color online) Level of assortativity that maxi-
cial networks are perhaps to be expected, since humans mizes the entropy, r ∗ , for various real-world, scale-free net-
seem to group together deliberately according to all kinds works, as predicted theoretically against scale-free exponent
of features. In other words, the elements themselves put γ. Bar ends show the empirical values.
energy into the system, driving it from equilibrium. On
the other hand, negative correlations have been shown
to appear in highly heterogeneous networks when only We go on to show how our method can be used to
one edge per pair of nodes is allowed – but this effect study the effects of correlations in any kind of network.
alone does not account for the degree of disassortativity In particular, we apply it to a neural-network model and
observed. find that assortativity greatly enhances the robustness to
noise of such a system2 .
We show that there is a general reason for the “nat-
ural” (i.e., equilibrium) state of heterogeneous networks
not to be neutral but, in general, disassortative1. To ∗
samuel@[Link]
do this, we develop an analytical method to partition 1
S. Johnson, J.J. Torres, J. Marro, and M.A. Muñoz, “En-
the phase space of networks compatible with given con- tropic origin of disassortativity in complex networks,”
straints into equally correlated regions. We obtain the Phys. Rev. Lett. 104, 108702 (2010).
Shannon entropy (which coincides, asymptotically, with 2
S. de Franciscis, S. Johnson, and J.J. Torres, “Enhancing
the Gibbs entropy for intensive constraints) for each par- neural network performance via assortativity”, Phys. Rev.
tition, and determine the degree of assortativity that E, in press (2011).

Junio de 2011, Barcelona Contribución Oral O–13

44 XVII Congreso de Fı́sica Estadı́stica FisEs11

Wetting-controlled drop emission in forced microfluidic filaments

R. Ledesma-Aguilar∗,† , R. Nistal∗ , A. Hernández-Machado∗, I. Pagonabarraga‡

∗
Departament d’Estructura i Constituents de la Matèria
Facultad de Fı́sica
Universidad de Barcelona
08028 Barcelona
†
The Rudolf Peierls Centre for Theoretical Physics
University of Oxford
1 Keble Road, Oxford
OX1 3NP, United Kingdom
‡
Departament de Fı́sica Fonamental
Facultad de Fı́sica
Universidad de Barcelona
08028 Barcelona

The controlled formation of micron-sized drops is of

great importance in microfluidic technological applica-
tions. Here we present a novel, wetting-based, desta-
bilization mechanism of forced microfilaments on either
hydrophilic or hydrophobic dry stripes, that leads to the
periodic emission of droplets1 .

The drop emission mechanism is triggered above a

critical forcing, where the contact line no longer follows
the leading edge of the filament. We propose a dynamical
model which includes the effects of wetting, capillarity,
viscous friction and the driving force to determine the
interface cofiguration at the threshold. We compare our
theory to lattice-Boltzmann simulations and microfluidic
experiments, accounting for the emission threshold and
hence the size and emission period of droplets, which can
be controlled independently.

Our results show that the critical filament velocity

depends strongly on wetting, and exhibits a qualita-
Figura 1. Drop emission in microfilament forced on a hy-
tive different behaviour on hydrophilic and hydrophobic
drophilic substrate (Lattice-Boltzmann simulation).
stripes, which arises from the dependency of viscous dis-
sipation on the shape of the advancing interface. Our
results suggest that this new kind of instability in con-
tact lines is general to advancing fronts2 , and opens new
possibilities of exploiting wetting to handle interfaces at 1
R. Ledesma-Aguilar, R. Nistal, A. Hernandez-Machado
the microscale. and I. Pagonabarraga, Nature Materials, 2011 (accepted)
2
C. Duez, C. Ybert, C. Clanet and L. Bocquet, Nature
Physics, 3, 180 (2007)

Contribución Oral O–14 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 45

Self-assembly of viral capsids

Antoni Luque∗ , David Reguera

Facultad de Fı́sica
Universidad de Barcelona
08028 Barcelona

Viruses are biological entities that can infect a wide agreement with several experiments that show different
variety of organisms -from bacteria to mammals-, caus- aspects of virus assembly, e.g., assembly-disassembly hys-
ing diseases that have a huge ecological, medical, and teresis, capsid production lag time, or capsid formation
economical impact. In their simplest form, viruses are sigmoidal curves.
constituted by an infective genetic material (DNA/RNA)
and a protective protein shell, called capsid, which is
generally built in a spontaneous assembly process from
several copies of the same or similar proteins. In addi-
tion, viral shells have a well-defined size in the nanometer
range, are usually highly symmetric, and show relevant
mechanical properties. All these features have spread the
interest for viral capsids in different nanoscience fields,
where several technological and biomedical applications
have been developed.

Figura 2. Line energy. We show the simulation (points)

and theoretical (curves) results for two radii of a spherical
capsid made of 32 capsomers (protein tiles). The rim of an
intermediate structure (see Fig. 1) is an interface between the
Figura 1. Assembly of a viral capsid. We illustrate the free subunit and capsid phases, and has associated a line ener-
process of formation of a viral shell. The spontaneous aggre- gy responsible for the energy barrier represented. The points
gation of single subunits produces a sequence of intermediates in the last part of the assembly (right) show an interesting
until a closed capsid is formed. phenomena during capsid closure: the implosion effect.

Here we focus on the self-assembly of viral capsids (see Therefore, this represents an interesting example of
Fig. 1). In vivo and in vitro experiments have shown that how basic physical principles can explain and guide the
empty viral shells can be formed in different conditions, understanding of biological systems. Due to the gener-
for instance, of pH or salt concentration, leading to differ- ality of the concepts involved, the study could also be
ent kind of structures. Several studies have pointed out adapted to other systems in other scientific fields, such
that the free energy minimization principle governs the as soft condensed matter.
origin of these viral structures1,2 , and that the assembly
of viral capsids should be regarded as a thermodynamic
∗
process3 . toniluque@[Link]
1
In this contribution, we will show that the assem- R. Zandi, D. Reguera, R. F. Bruinsma, W. M. Gelbart,
bly and disassembly of viruses have important analogies and J. Rudnick. (2004). Origin of icosahedral symmetry
with the standard vapor-liquid phase transition. We will in viruses. Proc. Natl. Acad. Sci. USA 101:15556–15560.
2
A. Luque, R. Zandi, and D. Reguera. (2010). Optimal
also demonstrate that classical nucleation theory could
architectures of elongated viruses. Proc. Natl. Acad. Sci.
be adapted to study the self-assembly of viral capsids4 ,
USA 107:5323–5328.
which provides a solid thermodynamic and kinetic frame- 3
R. F. Bruinsma, W. M. Gelbart, D. Reguera, J. Rudnick,
work to understand viral shell assembly. and R. Zandi (2003) Viral self-assembly as a thermody-
In particular, we will investigate in detail the case for namic process. Phys. Rev. Lett. 90:248101.
spherical capsids, which are the most abundant type of 4
R. Zandi, P. van der Schoot, D. Reguera, W. Kegel, and H.
viral shells. We will propose a simple continuum ther- Reiss (2006) Classical nucleation theory of virus capsids.
modynamic model that captures the main ingredients Biophys. J. 90:1939–1948
of viral assembly, and is in agreement with simulation 5
A. Luque, D. Reguera, A. Morozov, J. Rudnick, and
studies5 (see Fig. 2). Then, we will develop the clas- R. Bruinsma (2011) The assembly of spherical shells:
sical nucleation theory of viral capsids for this partic- Line Energy, Implosion and Closure Catastrophe (Unpub-
ular model. We emphasize that this theory will be in lished).

Junio de 2011, Barcelona Contribución Oral O–15

46 XVII Congreso de Fı́sica Estadı́stica FisEs11

Complex networks and glassy dynamics: walks in the energy landscape

Paolo Moretti1∗ , Andrea Barochelli1 , Alain Barrat2,3, Romualdo Pastor-Satorras1

1
Departament de Fı́sica i Enginyeria Nuclear, Universitat Politècnica de Catalunya
Campus Nord B4, E-08034, Barcelona, Spain
2
Centre de Physique Théorique (CNRS UMR 6207), Luminy, 13288 Marseille Cedex 9, France
3
Complex Networks Lagrange Laboratory, Institute for Scientific Interchange (ISI), Torino, Italy

Modeling the mechanical properties of systems that code the heterogeneity of the energy landscape into sta-
exhibit glassy dynamics represents a major issue, both tistical properties of the network, such as degree distri-
from the computational and the analytical point of view. butions and correlations. We show how to use the tools
Using realistic inter-particle potentials, molecular dy- developed for the study of dynamical processes on com-
namics simulations suitably describe events over limited plex networks, in order to go beyond the mean-field sce-
time-spans. They provide great insight into microscopic nario and move towards a more realistic description of
dynamics, yet they may become inadequate in predict- the problem. We consider several non-local transition
ing long-time relaxation features for normal laboratory rates between minima and show that, under very general
scales. In this light, effective modeling techniques must hypotheses, the existence of a glassy phase depends on a
rely on realistic coarse graining procedures. delicate interplay between the network topology and the
A simple way of addressing glassy dynamics is that of relationship between energy and degree of a minimum3 .
considering the system as a random walker in the energy Interestingly, the network-degree correlations and the de-
landscape. In the regime where the dynamics is thermal- tails of the transition rates do not play any role in the
ly activated, the walker evolves through jumps between existence (nor in the value) of the glass-transition tem-
different energy minima or traps1 . Both experimen- perature.
tal and numerical studies have revealed that the emer- This approach helps contextualize previous studies
gence of glassy dynamics, for instance in super-cooled liq- that focused on the sampling of specific energy land-
uids, is accompanied by the observation of strong spatio- scapes obtained for small systems4 and shows how the
temporal heterogeneities, with different regions of the tools developed in complex network theory can be suc-
system presenting dynamics that vary from each other cessfully employed in this context.
even by orders of magnitude. The classical trap model
indeed predicts the emergence of glassy features, however ∗
[Link]@[Link]
it considers a mean-field scenario where heterogeneities 1
J. P. Bouchaud, J. Physique I (France) 2, 1705-1713
are not taken into account.
(1992).
Here, we propose to study the role of heterogeneities 2
A. Barochelli, A. Barrat, and R. Pastor-Satorras
in systems approaching the glass transition, by modeling 3
P. Moretti, A. Baronchelli, A. Barrat, and R. Pastor-
time evolution as a generalized random walk on a com- Satorras, submitted, preprint available arXiv:1101.3490
plex energy landscape, pictured as a network of minima2 . 4
S. Carmi, S. Havlin, C. Song, K. Wang, and H. A. Makse,
Complex network theory provides a powerful tool to en- J. Phys. A 42, 105101 (2009)

Contribución Oral O–16 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 47

Pattern Formation in Reactive-Fluid Systems

Alberto P. Muñuzuri, Jacobo Guiu-Souto, Dario M. Escala, Alexandra Von Kameke, Jorge Carballido-Landeira∗
Group of Non Linear Physics Fac. de Fisicas. Universidade de Santiago de Compostela Campus Sur. 15782 Santiago de
Compostela

Highly nonlinear mechanisms for pattern formation others will be considered in this context. A second part
have been extensively studied due to the important role of the contribution considers gravity and density forces
they play in Nature. The characteristic discrete nature of to be compatible with the propagation velocity of the
the biological systems (cellular compartment) has been patterns, so that interaction can be analyzed. Typical
recently analyzed. Nevertheless, many of these process- gravitational instabilities will be considered by changing
es occur in a fluidic medium where hydrodynamic prop- viscosities and densities of the fluids under study.
erties are especially relevant. Just no name a few ex-
Experimental demonstrations of the different cases dis-
amples; reactant pollutants dispersed in the atmosphere
cussed will be shown and the results complemented and
or in oceans; cells in living organisms are embedded in
understood by theoretical and numerical analysis.
fluids, etc. We will present an analysis of the different
properties characterizing the fluids on pattern formation.
We consider typical pattern-forming reactions (Belousov- ∗
uscfmapm@[Link]
Zhabotinsky reaction like) with and without cell com- 1
A. von Kameke, F. Huhn, G. Fernandez-Garcı́a, A.P.
partment but always in a liquid environment. Different Muũzuri and V. Perez-Muũzuri, ’Propagation of a Chem-
instabilities are considered and the effect on the patterns ical Wave Front in a quasi two-dimensional Superdiffusive
observed analyzed. A first part of the contribution is Flow.’ Phys. Rev. E 81, 066211 (2010)
devoted to the effect of external forcings on these sys- 2
Jacobo Guiu-Souto, Jorge Carballido-Landeira, Vicente
tems. A well controlled turbulent flow can be generated Perez-Villar, and Alberto P. Muũzuri, ’Manipulation of
that helps analyzing the effect of the different parame- diffusion coefficients via periodic vertical forcing controls
ters on the patterning. Gravity modulations (Faraday the mechanism of Turing pattern formation.’ Phys. Rev.
type) or centrifugal instabilities (Taylor problem) among E 82, 066209 (2010)

Junio de 2011, Barcelona Contribución Oral O–17

48 XVII Congreso de Fı́sica Estadı́stica FisEs11

Relaxation Dynamics in the Columnar Liquid Crystal Phase of Oblate Hard

Spherocylinders: Effect of Transient Cages and Permanent Barriers

Alessandro Patti∗ , Simone Belli, René van Roij, Marjolein Dijkstra

Debye Institute for NanoMaterials Science and Institute for Theoretical Physics
Utrecht University
Utrecht, The Netherlands

We perform Monte Carlo simulations to analyze the ids, with a relatively fast decay of the correlation func-
equilibrium dynamics and the long-time structural re- tions.
laxation decay of columnar liquid crystals (LCs) of disk-
like colloidal particles with diameter D and height L.
In the wake of recent studies on calamitic (rod-like)
colloidal particles in the smectic1–3 or in the columnar
mesophases4 , we now focus on the diffusion of their dis-
cotic counterparts, here modeled as oblate hard sphe-
rocylinders. In spite of the substantial differences in
the equilibrium phase behavior between LCs of calamitic
and discotic particles, we find interesting similarities in
their rattling-and-jumping diffusion and two-step relax-
ation dynamics. In all these studies, it was found that
to diffuse along the broken 1D (smectic) or 2D (colum-
nar) symmetry direction, a particle must overcome a
free-energy barrier of the order of a few kB T , depend-
ing mostly on the packing of the system, but also on
the particle anisotropy and on the rotational degrees of
freedom. Due to the combined action of transient cages
and periodic free-energy barriers, this diffusion presents
a non-Gaussian behavior.
More specifically, we find that at fixed packing frac-
tion the barrier height increases with decreasing particle
thickness, resulting into a more heterogeneous and non-
Gaussian dynamics for thinner platelets, and reducing
the inter-column diffusion coefficient. We observe the
characteristic two-step relaxation decay of the structure
in the plane perpendicular to the column axis. At very
short times, the discotic particles freely diffuse in the cage Figura 1. Mean square displacement (open symbols), in
formed by their nearest neighbors. At this stage, the sys- units of D2 , and non-Gaussian parameter (solid symbols) in
tem shows a Gaussian behavior with a linear mean square the plane perpendicular to the nematic director, for systems of
displacement and a fast exponential decay of the corre- oblate hard spherocylinders with L/D = 0.1 (a), L/D = 0.2
lation functions. As soon as the particles feel the pres- (b), and L/D = 0.3 (c). Circles, squares, and diamonds refer
to packing fractions η = 0.575, 0.600, and 0.630, respectively.
ence of their surrounding cage, the diffusion slows down
For comparison, also the results at L/D = 0.1 and η = 0.500
significantly, deviations from Gaussianity are observed,
are shown (triangles). Note that the vertical axes of MSDs
and the mean square displacement develops a plateau
and NGPs are on the left and right, respectively.
whose time extension increases with density and/or par-
ticle anisotropy. As time passes, an increasing number
of particles jumps from a column to another, hence con-
tributing to recover a homogeneous dynamics which re-
sults into a second diffusive regime. At the beginning ∗
[Link]@[Link]
of the long-time diffusion, which indicates the end of the 1
A. Patti, D. El Masri, R. van Roij, and M. Dijkstra, Phys.
cage regime, the deviations from Gaussian behavior start Rev. Lett., 2009, 103, 248304.
to decrease and go exponentially to zero. At high densi- 2
A. Patti, D. El Masri, R. van Roij, and M. Dijkstra, J.
ties, the plateau may extend beyond our simulation time Chem. Phys., 2010, 132, 224907.
and no significant diffusion over the trapping cages is ob- 3
R. Matena, M. Dijkstra, and A. Patti, Phys. Rev. E, 81,
served. 021704.
4
By contrast, the in-column dynamics is similar to the S. Belli, A. Patti, R. van Roij, and M. Dijkstra, J. Chem.
typical single-file diffusion of one-dimensional dense flu- Phys., 2010, 133, 154514.

Contribución Oral O–18 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 49

The helical crack-front instability

A. J. Pons
Edifici GAIA, despatx 1.08 Departament de Fı́sica i Enginyeria Nuclear Universitat Politècnica de Catalunya, Campus de
Terrassa Rambla de Sant Nebridi s/n, 08222 Terrassa, Spain

Planar crack propagation becomes universally unstable the destabilizing effect of far-field stresses and the stabi-
under tension loading (mode I) with the superposition of lizing effect of cohesive forces on the process zone scale,
a shear stress parallel to the crack front (mode III). Un- and we derive a theoretical estimate for this scale using
der this mixed-mode (I-III) loading configuration, an ini- a new propagation law for curved cracks in three dimen-
tially flat parent crack segments into an array of daugh- sions. The simulations reveal that planar crack propaga-
ter cracks that rotate towards a direction of maximum tion evolves nonlinearly into a segmented array of finger-
tensile stress. This segmentation produces stepped frac- shaped daughter cracks which gradually coarsen owing
ture surfaces with characteristic ‘lance-shaped’ markings to the growth competition of daughter cracks. The ro-
observed in a wide range of engineering and geological tation angles of coarsened facets are also compared to
materials. We perform large-scale simulations of mixed- theoretical predictions and available experimental data.
mode I-III brittle fracture using a continuum phase-field
method that describes the complete three-dimensional
1
crack-front evolution. We show that the dynamically pre- Pons, A. J.; Karma, A.,Helical crack-front instability in
ferred unstable wavelength is governed by the balance of mixed mode fracture, Nature, 464, 85-89, (2010).

Junio de 2011, Barcelona Contribución Oral O–19

50 XVII Congreso de Fı́sica Estadı́stica FisEs11

Estudio mediante geometrı́a intrı́nseca de procesos de crecimiento cinéticos con

rugosidad

Javier Rodriguez-Laguna†, Silvia N. Santalla‡∗ , Rodolfo Cuerno†

†
Dpt. Matemáticas y Grupo Interdisciplinar de Sistemas Complejos (GISC), Universidad Carlos III de Madrid, 28911
Leganés (España)
‡
Dpt. Fı́sica, Universidad Carlos III de Madrid. 28911 Leganés (España)

Los procesos cinéticos de crecimiento con rugosidad Proponemos3 una ecuación en derivadas parciales es-
fuera de equilibrio han sidio estudiados extensivamente tocástica definida en 2D, que no introduce ninguna difer-
tanto como un problema fundamental de mecánica es- encia entre las direcciones de crecimiento e interfacial.
tadı́stica como sistemas fı́sicos relevantes en sı́ mismos, Todos los operadores que aparecen son de naturaleza ge-
tales como el crecimiento de láminas delgadas, dinámica ométrica:
de fluidos, propagación de frentes de llamas o crecimiento
biológico indiferenciado1 .
El Ansatz de Family-Vicsek es un esquema que de- vn (r ) = A0 + A1 K(r) + A2 ∇2 K(r) + An η(r)
scribe con éxito la mayorı́a de los sistemas cinéticos con
rugosidad. Definamos la rugosidad del sistema W como donde vn es la velocidad a lo largo de la dirección normal
la anchura promedio de la interfaz. Comenzando con una en un punto de la interfaz, r es un punto tal, K es la cur-
configuración plana, W crece inicialmente en el tiempo vatura local, ∇2 es el operador de Laplace-Beltrami y η
con una ley de potencias, W ≈ tβ . Pero para largos es un ruido blanco, decorrelacionado en espacio y tiempo.
tiempos, la rugosidad satura en un valor que depende del Ası́, A0 , A1 , A2 y An son constantes libres.
tamaño del sistema L con otra ley de potencias: W ≈ Lα . Nuestra ecuación está complementada con un algorit-
Los exponentes α y β caracterizan el sistema. En la mo numérico que asegura la naturalidad geométrica del
práctica, muchos sistemas diferentes toman los mismos procedimiento, adaptativo con el fin de que la interfaz
valores para estos exponentes, mostrando un alto grado siempre sea simulada con un grado de resolución sufi-
de universalidad. ciente. Las medidas también se llevan a cabo mediante
procedimientos geométricamente naturales. Nuestros re-
Concretamenten, en 1+1D se han identificado varias
sultados serán referidos a geometrı́a banda y, asimismo,
clases de universalidad, que han sido asociadas a ecua-
a geometrı́a circular.
ciones en derivadas parciales estocásticas de evolución
temporal de cierta “función altura” h(x, t), tales co-
mo la ecuación de Edwards-Wilkinson o la renombrada ∗
[Link]@[Link]
ecuación de Kardar-Parisi-Zhang. Es interesante notar 1
A.-L. Barabási and H. E. Stanley, Fractal concepts in
que dichas ecuaciones, cuyas soluciones son conocidas surface growth (Cambridge University Press, Cambridge,
analı́ticamente2, son obtenidas dentro de la aproximación 1995).
de pendientes pequeñas, impidiendo la formación de mar- 2
T. Sasamoto and [Link], Phys. Rev. Lett. 104, 230602
quesinas. Una pregunta relevante, por tanto, es: ¿cambia (June 2010)
3
la clase de universalidad cuando estas restricciones no se J. Rodriguez-Laguna, S.N. Santalla, R. Cuerno, JSTAT,
aplican? submitted (2011).

Contribución Oral O–20 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 51

Thermodynamic tradeoffs in sensory adaptation: the energy-speed-accuracy relation∗

G. Lan1 , Pablo Sartori1,2 , S. Neuman3 , V. Sourjik3 , Y. Tu

IBM T.J. Watson Research Center
Yorktown Heights, 10598 NY

INTRODUCTION Adaptation is a fundamental

function of living systems. Importantly, adaptation al-
lows sensory systems to maintain high sensitivy over a
wide range of backgrounds. An everyday example is the
adaptation of the activity (a) of mammalian photore-
ceptors to prolonged light stimuli (s), by the action of
a chemical “memory” (m) through a negative feedback
(Fig. 1).

Figura 2. A low memory level balances a low background

signal to maintain the activity near the adaptaed state 0.5.
This is done at the expense of energy dissipation, as can be
readily seen by the pressence of phase-space fluxes.

By using probability flux conservation (Fig. 2), we an-

Figura 1. A Topology of a sensory system: a stimulus (s) alytically derive a universal relation between the Energy
enhances the activity (a) of a receptor, which through a mem- dissipation rate (Ẇ ), the adaptation Speed (ωad ), and
ory (m) mediated feedack recovers to a target value a0 . B the adaptation Accuracy (1/), the ESA relation:
Adaptive response of the activity a of a sensory system to
a step signal. After a transient peak at , the final activity Ẇ ∼ −ωad cσa2 log(/c ) (1)
af recovers to the target a0 with accuracy 1/
at a speed
ωad = 1/τad . Where σa2 is the variance of activity fluctuations, c de-
pends on the model details, and kB T = 1.
E. COLI CHEMORECEPTOR: MODEL AND
EXPERIMENTAL VERIFICATION We then fo-
cus on the E. Coli chemoreceptor, the best known exam-
ple of sensory adaptation. Since the reaction rates are
well known, we can perform extensive simulations of the
When leaving a dark room the immediate response of undergoing chemical kinetics.
the eye is strong. Right after, the eye becomes insen- Through simulations we recover the ESA relation in
sitive as it is overstimulated by light. However, a few Eq.1, where ωad is the rate of the reactions which violate
seconds later it adapts to its “ready-to-sense” state, and detailed balance. For  ∼ 1% as meaured, each transition
is then capable of resolving small relative differences in dissipates ΔW ≡ Ẇ /ωad ∼ 30kB T , roughly the energy
light intensity. of the S-adenosylmethionine (SAM) molecule hydrolized
in the chemotaxis pathway.
Our theory predicts that a reduction in the ammount
Although the benefits of adaptation are well known, its
of available SAM in a cell does not change the adaptation
costs remain poorly understood. In this work we argue
accuracy 1/, as ΔW ∼ 1SAM is the same. However Ẇ
that sensory adaptation is a dissipative phenomena, and
will decrease, and hence so will ωad . We measure in vivo
try to elucidate the energetic cost of sensory systems in
the response of deenergizing cells. While their adapta-
maintaining an adapted state.
tion accuracy remains constant, there is a clear decrease
in adaptation speed ωad . This confirms the theoretical
THE GENERAL CONTINUOUS MODEL We
prediction derived from the ESA reltion.
analyze the stochastic dynamics of the generic feedback
circuit in Fig. 1 responsible for sensory adaptation tho-
rugh Langevin and Fokker-Planck equations. We show ∗
Work currently under review in Science.
that adaptation processes are inherently dissipative, as 1
Both these authors contributed equally to this work.
they violate detailed balance. As a consequence, continu- 2
Part of this work was done in the facilities of the Statistical
ous energy consumption (entropy production) is required Physics group at the University of Granada.
to stabilize the adapted state. 3
S.N. & V.S. affilitation is the ZMBHU lab, in Heidelberg.

Junio de 2011, Barcelona Contribución Oral O–21

52 XVII Congreso de Fı́sica Estadı́stica FisEs11

Flujo de Couette para una impureza inmersa en un gas granular

Francisco Vega Reyes,∗ Vicente Garzó,† , Andrés Santos‡

Departamento de Fı́sica, Universidad de Exremadura, 06071 Badajoz

El estudio de las propiedades de transporte de los un modelo BGK adaptado a la inelasticidad de las col-
medios granulares tiene numerosas aplicaciones tec- isiones y ii) mediante el método de los 13 momentos de
nológicas e industriales, además de un evidente interés Grad. Además, hemos obtenido la solución numérica de
a nivel más fundamental.1 Por ejemplo, en el caso de ambas ecuaciones mediante el método de simulación di-
medios granulares termalizados y muy poco densos, es recta de Monte Carlo (DSMC). Comparamos en este tra-
posible una descripción de sus propiedades de transporte bajo los resultados obtenidos por estas tres vı́as, mostran-
en el contexto de la mecánica de fluidos. Ası́, es posi- do los valores de los coeficientes de transporte de la im-
ble deducir expresiones de los coeficientes de transporte pureza. Además, los resultados muestran claramente que
hidrodinámicos clásicos también para gases granulares. la hidrodinámica es no newtoniana, por lo que también
Las condiciones de transporte y segregación de impurezas analizamos las propiedades reológicas de la impureza.
inmersas en un gas granular se presenta como uno de los Otra propiedad interesante en mezclas granulares (in-
problemas de indudable interés técnico.2 Por ello, nos cluyendo el caso de impureza) es el cociente de tempera-
proponemos en este trabajo analizar las propiedades de turas granulares de las especies, que resulta ser una cons-
transporte de una impureza granular inmersa en un flujo tante en estados estacionarios. En la Fig. 1 mostramos
de Couette. los resultados que se derivan para esta magnitud en fun-
En nuestro sistema, disponemos de un gas granular ción de la tasa de cizalladura local a, magnitud que in-
de discos/esferas duras lisas idénticas, en el cual se hal- dica la intensidad del flujo de Couette en el gas granular
la presente, en una concentración despreciable, una es- y que también es una constante del sistema en estados
pecie de esferas/discos, también inelásticos, pero que estacionarios.
tiene propiedades diferentes a las del gas granular en
el que está inmerso. En nuestro sistema simplificado,
esas propiedades distintivas pueden ser: o bien el coefi-
1.6
ciente de restitución normal, que caracteriza la cantidad
de energı́a cinética perdida en las colisiones, o bien el 1.4
diámetro o masa de las partı́culas. Denotaremos, pues, 1.2 
  
con los subı́ndices 1 y 2 a la impureza y al gas granular,
Χ

respectivamente, de manera que denotamos como α1 y 1.0

  
α2 al coeficiente de restitución normal de la impureza y  
0.8
del gas granular, mientras que ω ≡ σ1 /σ2 y μ ≡ m1 /m2
indicarán, respectivamente, la relación entre diámetros 0.6
(σ) y masas (m) de ambas especies. 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Nuestro gas granular y la impureza se hallan entre dos a2
paredes paralelas e infinitas desde las cuales se calienta Figura 1. Ratio de temperaturas χ ≡ T1 /T2 para una im-
(a temperaturas diferentes) y cizalla el gas (geometrı́a del pureza con μ = 2 (sı́mbolos y lı́neas superiores en negro) y
flujo de Couette). En estas condiciones, el estado esta- μ = 0.5 (sı́mbolos y lı́neas inferiores en rojo) en función de la
cionario de la impureza queda determinado no sólo por tasa de cizalladura. Sı́mbolos: datos DSMC, lı́neas continuas:
sus propiedades fı́sicas sino también por las del gas en el modelo BGK, lı́neas discontinuas: método de Grad.
que está inmersa. En cambio, el estado final de la especie
del gas no se ve afectado por la presencia de la impureza.
De esta manera, y en nuestra geometrı́a, podemos escribir En nuestro trabajo presentaremos comparativamente
las ecuaciones de Boltzmann para el gas y la impureza, los resultados obtenidos por los tres métodos y discutire-
respectivamente, como mos las propiedades de los flujos de Couette de una im-
pureza granular. Mostraremos, además, una clasificación
∂ general de los mismos.
Vy f2 = J[f2 , f2 ], (1)
∂y
∂ ∗
Vy f1 = J[f1 , f2 ], (2) fvega@[Link]
∂y †
vicenteg@[Link]
‡
andres@[Link]
donde J[fi , fj ] es la integral colisional de la ecuación de 1
F. Vega Reyes, A. Santos, and V. Garzó, Phys. Rev. Lett.
Boltzmann entre las especies i y j. 104, 028001 (2010).
2
La solución analı́tica de las ecuaciones (1) y (2) la F. Vega Reyes, V. Garzó, and A. Santos, J. Stat. Mech.
hemos aproximado por dos vı́as alternativas: i) mediante P09003 (2008).

Contribución Oral O–22 Junio de 2011, Barcelona

 Parte III
Paneles
XVII Congreso de Fı́sica Estadı́stica FisEs11 55

Impact of subdiffusion on the formation and properties of morphogen gradients: a

reaction-diffusion approach

Santos Bravo Yuste , Enrique Abad† , K. Lindenberg ‡

 †
Departamento de Fı́sica y Departamento de Fı́sica Aplicada
Universidad de Extremadura
06006 Badajoz
‡
Department of Chemistry and Biochemistry
University of California, San Diego
La Jolla, CA 92093-0340

Recently, increasing evidence has been gathered on the nung et al.3 and allow for simultaneous degradation
importance of memory effects in systems where trans- and CTRW transport. In our model, the degrada-
port and reaction events coexist over biologically relevant tion reaction is assumed to be independent of transi-
time scales. The joint effect of reaction and anomalous tions between bound and unbound states and modeled
transport may bring about unexpectedly strong devia- via a first-order death process whose rate coefficient k
tions from results obtained for classical reaction-diffusion (reactivity) is in general space-dependent, k = k(x).
systems. Several authors have shown that a rigorous The underlying CTRW reaction-transport model leads
derivation of the relevant reaction-transport equations to a reaction-subdiffusion equation (RSE) with a mixed
from mesoscopic models will typically result in the oc- reaction-transport term containing a non-local integro-
currence of mixed reaction-transport terms1 . differential operator (the so-called Riemann-Liouville
An scenario where such equations could find a poten- fractional derivative). For a homogeneous k, our RSE
tial application is morphogen gradient formation2 , a pro- solved for the appropriate boundary conditions yields
cess of utmost importance in developmental biology. Tis- exponentially decaying stationary morphogen concentra-
sue patterning in multicellular organisms is often gov- tion profiles, as opposed to the results found by Hornung
erned by special signaling molecules called morphogens. et al. However, if the reaction is confined to a small re-
Morphogens are produced at localized sites and may un- gion around the morphogen source, a steady state can be
dergo degradation as they disseminate through the de- sustained only in the limit of normal diffusion; otherwise,
veloping tissue. The combined action of transport and the transport is too slow to stop morphogen accumula-
the degradation reaction results in a morphogen concen- tion outside the reactive region, leading to a discontinu-
tration gradient which cells across the tissue can sense ous concentration profile and a lack of a stationary profile
and interpret. The local morphogen concentration level (see fig. 1). In general, we find a large gamut of concen-
determines the number and the identity of the genes each tration profiles4 depending on the specific form of k(x).
cell will express, and thus it also determines its ultimate
fate. Hence, there is a direct relation between tissue pat-
terning and the shape of the morphogen gradient2 .
Hornung et al.3 recently suggested that in some cas-
es cell-to-cell morphogen transport involves successive
binding and unbinding events to specific cell surface re-
ceptors. Rather than ordinary diffusion, such stochas-
tic events can be viewed as trapping events whose wait-
ing times display a significant dispersion. Consequent-
ly, Hornung et al. constructed a 1-d transport model3 Figura 1. Typical evolution of concentration profiles for a
based on the so-called Continuous Time Random Walk step reactivity [k(x) = k0 H(R−x)] with k0 = 1/1000, R = 5.5
(CTRW), which is known to yield subdiffusive behavior
in the limit of a large number of jumps. Additionally, 
santos@[Link]
they allowed for morphogen degradation at a constant 1
V. Méndez, S. Fedotov, and W. Horsthemke, Reaction-
rate per transition between bound and unbound states.
Transport Systems: Mesoscopic Foundation, Fronts, and
However, they assumed that, while bound to the recep- Spatial Instabilities (Springer-Verlag, Berlin, 2010); E.
tors, morphogens were protected against degradation (re- Abad, S. B. Yuste, and K. Lindenberg, Phys. Rev. E 81,
action and transport are thus mutually exclusive). Their 031115 (2010).
approach yields transient concentration gradients, but 2
M. Ibáñes and J. C. Izpisúa, Molecular Systems Biology
not stationary morphogen profiles associated with sta- 4, 176 (2008).
ble tissue patterns, prompting certain authors to assert 3
G. Hornung, B. Berkowitz B, and N. Barkai, Phys. Rev.
that the latter were incompatible with anomalous (sub- E 72, 041916 (2005).
diffusive) transport. 4
S. B. Yuste, E. Abad, and Katja Lindenberg, Phys. Rev.
In this work, we relax the assumption made by Hor- E. 82, 061123 (2010).

Junio de 2011, Barcelona Panel P–1

56 XVII Congreso de Fı́sica Estadı́stica FisEs11

Propiedades del agua en condiciones extremas: región subenfriada y presiones

negativas

J.L.F. Abascal, M.A. González, C. Vega

Depto Quı́mica-Fı́sica I, Facultad de C. Quı́micas, Univ. Complutense, 28040 Madrid

El anómalo comportamiento del agua y sus consecuen- la temperatura de fusión con la presión observando que
cias son conocidos incluso fuera del ámbito estrictamente también muestra un reentrante aunque su pendiente es
cientı́fico. Es el caso por ejemplo de la existencia de notablemente más abrupta que la de la TMD4 .
un máximo en la densidad (TMD) a una presión dada.
Cuando nos alejamos de las condiciones de temperatu-
ra y presión ambiente, las anomalı́as se hacen aún más 2500

notables, especialmente en la zona subenfriada. Para ex-

plicar de forma unitaria el comportamiento del agua se 2000

ha propuesto la existencia de un segundo punto crı́tico

corrrespondiente a una transición lı́quido-lı́quido en la re- 1500

p (bar)
gión subenfriada (LLCP)1 . Aunque ciertos experimentos
parecen sugerir la existencia del LLCP, todavı́a no hay 1000
191K
una evidencia experimental concluyente. Por ello, los es- 195K
199K
tudios de simulación pueden arrojar luz sobre este tema. 500 203K
207K
El problema es que la simulación se basa en modelos de 215.8K
224K
agua simplificados por lo que cabe dudar de si represen- 0
0.96 1 1.04 1.08 1.12
tan el verdadero comportamiento del agua en condiciones 3
ρ (g/cm )
extremas.
4000
Trabajos experimentales recientes sobre la ecuación de Widom line
TMD
estado del agua en la región subenfriada permiten por 3000 spinodal
Tmelting
vez primera chequear capacidad predictiva de los mode- 2000
los en la zona en que se presume se encuentra el LLCP. LLCP
1000
La comparación de estos datos experimentales con las
p/bar

predicciones para el modelo TIP4P/20052 muestran una 0

excelente concordancia (tanto cualitativa como cuantita- -1000
tiva). Debe esperarse por tanto que los resultados de
-2000
simulación de este modelo se aproximen (incluso cuan-
titaviamente) a los del agua real. Por ello hemos lle- -3000
180 200 220 240 260 280 300 320
vado a cabo exhaustivas simulaciones para localizar el T/K
LLCP3 . Hemos calculado también la lı́nea de compre- Figura 1. Resultados de simulación para el modelo
sibilidad máxima (lı́nea de Widom) que emana desde el TIP4P/2005. Arriba: Isotermas en la región subenfriada.
punto crı́tico y se adentra en la región de presiones nega- Abajo: Lı́neas de Widom, TMD, spinodal y puntos de fusión
tivas. La lı́nea de Widom acaba uniéndose con la lı́nea de
densidades máximas justo en el punto en que ésta cambia
de pendiente (“retracing TMD”).
Consideraciones termodinámicas han demostrado que,
en el caso de TMD reentrante, la lı́nea de Widom debe ∗
jlfabascal@[Link]
continuar con otra de compresibilidad mı́nima. Nuestros 1
P. H. Poole, F. Sciortino, U. Essmann y H. E. Stanley,
cálculos con el modelo TIP4P/2005 muestran la con- Nature 360, 324 (1992)
tinuidad de ambas lı́neas y su contacto con la TMD a 2
J.L.F. Abascal y C. Vega, J. Chem. Phys. 123, 234505
aproximadamente -1000 bar. A partir de este punto la (2005).
pendiente de la TMD se hace muy pequeña de forma que 3
J.L.F. Abascal y C. Vega, J. Chem. Phys. 133, 234502
llega ser más o menos paralela a la espinodal lı́quido- (2010).
4
vapor. Finalmente, hemos investigado la dependencia de M.A. Gonzalez y J.L.F. Abascal (en preparacion).

Panel P–2 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 57

Depósitos y Flujo de Partı́culas con Caras Planas en un Silo

M. Acevedo∗ , I. Zuriguel, D. Maza

Departamento de Fı́sica y Matemática Aplicada, Universidad de Navarra, 31080 Pamplona.

Generalmente, cuando se estudia el comportamiento Adicionalmente a estos resultados, se presentarán los

de los medios granulares, ya sea de forma numérica o obtenidos cuando se varı́a la velocidad de carga (F eed
de forma experimental, las partı́culas que los constituyen Rate) comparándolos con los anteriormente expuestos.
son esferas o discos dependiendo si se trabaja en 3 ó 2
dimensiones. Sin embargo, estudios recientes muestran
que la geometrı́a de las partı́culas tiene un efecto im-
portante en propiedades granulares tales como: el factor
de empaquetamiento, el número de coordinación, la for-
mación de atascos y la propagación de estrés. Por otro
lado, el diseño y la explotación de silos de granos requiere
de un profundo conocimiento del comportamiento de las
partı́culas tanto en el almacenamiento como en la descar-
ga.
En este trabajo se estudia el depósito y flujo en el inte-
rior de un silo de partı́culas con caras planas y diferente
relación de aspecto. Pese al gran interés industrial de
este tipo de partı́culas, existe un gran desconocimiento
acerca de su comportamiento debido a la complejidad de
las interacciones entre las mismas.
Los principales resultados se exponen en la figura 1.
En a) se presenta la distribución de orientaciones de
partı́culas con una relación de aspecto de d = 5.4. De-
spués del proceso de carga, la orientación preferente es la
horizontal o paralela al sustrato (θ = 0 y θ = π), mientras
que las orientaciones verticales (θ = π/2) son muy poco
probables1,2. Después de descargar parcialmente el silo,
esta distribución de orientaciones cambia hacia una más
plana, indicando que es igualmente probable encontrar
cualquier orientación tras este proceso.
Los resultados que se muestran en b) corresponden a
partı́culas con relación de aspecto d = 2.4. Después del
depósito, la distribución es sensiblemente diferente al ca- Figura 1. Histogramas de Orientaciones para partı́culas
so de relación de aspecto d = 5.4, ya que la orientación con diferentes relaciones de aspecto: en a) d = 5.4 en b)
se desvı́a de la horizontal. Después de descargar par- d = 2.4 y en c) d = 1.
cialmente, se observa que la orientación preferente es la
vertical.
En c) se ilustran los resultados para relación de aspec-
to d = 1. Estos resultados muestran que, después de la
carga, la orientación más probable es θ = π/4 donde la ∗
macevedo@[Link]
diagonal es paralela a la gravedad. Después de la descar- 1
R. Cruz-Hidalgo, I. Zuriguel, D. Maza and I. Pagonabar-
ga parcial, resulta que estas partı́culas giran un ángulo raga, J. Stat. Mech. (2010) P06025.
2
de θ = π/4 provocando que la orientación preferida sea R. Cruz Hidalgo, I. Zuriguel, D. Maza and I. Pagonabar-
la horizontal (θ = 0). raga. Phys. Rev. Lett., 103, 1180001 (2009).

Junio de 2011, Barcelona Panel P–3

58 XVII Congreso de Fı́sica Estadı́stica FisEs11

Universal critical avalanches in the Coulomb glass

Martin Goethe∗ and Matteo Palassini†

Departament de Fı́sica Fonamental, Universitat de Barcelona
Diagonal 647, E–08028 Barcelona.

The Coulomb glass (CG) is a strongly correlated with the divergence of the screening length for T → 0.
regime observed at low temperature in Anderson insu- These results shed new light on the elementary excita-
lators such as dirty metals, amorphous semiconductors, tions in disordered insulators and on the long-debated
and lightly doped semiconductors. A defining feature of nature of the CG.
the CG is the extremely slow relaxation of the hopping
conductivity (phonon-assisted tunneling between local- 100
L= 8
ized states), and related glassy non-equilibrium effects -1 16
such as aging and memory1 . It was suggested2 that these 10 30
60
effects reflect an underlying equilibrium “glass phase”,

g(S)
10-2
which however was ruled out by recent numerical studies3
down to very low, but finite, temperature T .
10-3
In this work we search for glassiness directly at T = 0
by investigating charge avalanches in the CG for the first 10-4
time. We consider the standard CG model2 consisting 0 10 20 30 40 50 60
of N/2 electrons localized on N lattice sites, with 1/r S
pairwise interactions and random local potentials, in a 2
10 -3/2
x exp(-x)
uniform neutralizing background charge. Starting from 1 L= 16
10
a random configuration, we quench the system instan- 30
ζ(L) g(S)

taneously to T = 0 and let it evolve until it reaches a 100 60

local minimum, i.e. a stable configuration against all -1
10
one-electron jumps. We then perturb the system either ζ(L) = (2π)1/2 λL (1-exp(-λL)) (L/2)3/2
10-2
by inserting one extra electron at an empty site (charge -3 λL defined through λL - 1 - ln(λL) = 2/L
injection), or by displacing an electron from an occupied 10
site to a nearby empty site (dipole injection). Due to the 0.1 1
perturbation, other electrons may become unstable and S / (L/2)
relax to new positions. This in turn can trigger further Figura 1. (Top) Size distribution of the avalanches trig-
jumps so that an avalanche forms, until the system stops gered by charge injection, for different linear sizes L. The lines
upon reaching a new local minimum. are fits to compound Poisson distributions of Galton-Watson
It was shown recently5 , in agreement with earlier processes. (Bottom) Scaling plot of g(S) for avalanches trig-
simulations6 , that the avalanche size distibution g(S) for gered by dipole injection. In both cases g(S) has a power-law
similar avalanche processes in the infinite-range spin glass tail with an exponential cutoff, which diverges linearly with
and other mean-field models has a power-law behavior L.

g(S) ∼ S −τ exp[−(S/Sc )2 ] (1)

∗
where S is the number of jumps, τ > 0, and Sc a cutoff martingoethe@[Link]
†
that diverges for large system sizes. This signals a kind palassini@[Link]
1
A. Amir, Y. Oreg, and Y. Imry, in Annual Review of
of self-organized criticality6 , which was shown to be re-
Condensed Matter Physics Vol. 2: 235-262 (2011); arX-
lated to the marginal criticality5 of the equilibrium glass
iv:1010.5767.
phase in these models. 2
M. Müller and L. B. Ioffe, ibid. 93 (2004); S. Pankov and
Our goal is to elucidate whether a similar scenario is
V. Dobrosavljević, ibid. 94, 046402 (2005); M. Müller and
at play in the CG. Using a new algorithm4 that allows us S. Pankov, Phys. Rev. B 75, 144201 (2007) and references
to measure g(S) with high statistics for large systems therein.
(N = L3 ≤ 603 ), we find a power-law tail in agree- 3
M. Goethe and M. Palassini, Phys. Rev. Lett. 103, 045702
ment with Eq.(1) for both charge and dipole injection, (2009); B. Surer et al., ibid. 102, 067205 (2009)
with Sc diverging linearly with L (Fig. 1). We interpret 4
A. Glatz, V. M. Vinokur, J. Bergli, M. Kirkengen, and Y.
these results in terms of rearrangements of soft long-lived M. Galperin, J. Stat. Mech. P06006 (2008)
particle-hole pairs (dipoles), and analyse the possible ori- 5
P. Le Doussal, M. Müller, and K. Wiese, EPL 91, 57004
gin of the power-law behavior both from the point of view (2010)
6
of branching processes, and of glassy criticality. Finally, F. Pázmándi, G. Zaránd, and G. T. Zimányi, Phys. Rev.
we discuss the relationship of the scale-free avalanches Lett. 83, 1034 (1999)

Panel P–4 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 59

Stabilization of self-propelling particle clusters

F. Alarcón, I. Pagonabarraga∗
Departament de Fı́sica Fonamental
Facultat de Fı́sica
Universitat de Barcelona
Carrer Martı́ i Franqués, 1
08028 Barcelona

Suspensions of self-driven organisms are related to a ics4 . The theoretical prediction, based on linearized hy-
large variety systems of relevant interest which are com- drodynamics, that squirmer bands are intrinsically unsta-
posed by swimming organisms, such as fish shoals or bac- ble2 has not been analyzed. Previous numerical evidence
teria colonies. All these systems, referred to generically shows the development of dense squirmer bands coexist-
as active materials, are composed by constituents which ing with a dilute fluid suspension. As a result, nonlinear
absorb energy from their surroundings or from their in- couplings appear as a natural mechanism for band stabi-
ternal metabolism and dissipate it, usually carrying out lization which we analyze using a lattice Boltzmann (LB)
internal movements, leading to translational or rotation- model to describe the fluid and address the appropriate
al motion1 . These systems remain out of equilibrium, time scales which determine the coupling of the active
a feature which leads to unique properties and enhance particles and the fluid while the activity of the particles
the ability of these systems to self assemble and develop is characterized using the model described in Ref. [5 ].
patterns as a result of their intrinsic motion2 . We also address the role of squirmer interactions and ge-
We have studied the collective behavior of communities ometrical confinement to stabilize squirmer bands and
of active particles using a simple model in which the ef- discuss a hydrodynamically-controlled route to flocking.
fect of the internal metabolism of the microorganism can
be described through the effective fluid flow the particle
∗
generates on its surface. This squirmer model, which con- ipagonabarraga@[Link]
1
stitutes a model for ciliated microorganisms and certain Y. Hatwalne, S. Ramaswamy, M. Rao and R. Aditi Simha,
types of microrobots, accounts for the correct coupling Phys. Rev. Lett. 92, 118101 (2004).
2
of the selfpropelling particle to the fluid surrounding it. S. Ramaswamy and M. Rao, New Journal of Physics 9,
It is known that this hydrodynamic coupling promotes 423 (2007).
3
self-assembly through cluster formation3 , although their I. Llopis and I. Pagonabarraga, Euro. Phys. Lett. 75, 999
(2006).
stability has not been addressed systematically. 4
T. Vicsek, A. Czirók, E. Ben-Jacob, I. Cohen and O.
We address systematically the stability of clusters of
Shochet. Phys. Rev. Lett. 75, 1226 (1995).
squirmers, paying special attention to size spanning clus- 5
R. Matas Navarro and I. Pagonabarraga, Eur. Phys. J. E
ters, responsible for flocking transitions in ensembles of 33, 27 (2010).
selfpropelling particles in the absence of hydrodynam-

Junio de 2011, Barcelona Panel P–5

60 XVII Congreso de Fı́sica Estadı́stica FisEs11

Molecular dynamics study of polarizable ion models for molten copper halides

Olga Alcaraz∗ , Vicente Bitrián and Joaquim Trullàs

Departament de Fı́sica i Enginyeria Nuclear,
Universitat Politécnica de Catalunya, Campus Nord, B4-B5
08034 Barcelona, Spain

While the structural properties of molten alkali halides

are well reproduced by using rigid ion models (RIM) in
which the ions interact throughout the effective pair po-
tentials of Tosi and Fumi,1 the main features of the ex-
perimental structure factors S(k) for molten silver and
copper halides cannot be reproduced from the available
RIM. However, by using polarizable ion models (PIM),
in which the many-body interactions due to the anion-
ic induced polarization are added to those of the effec-
tive pair potentials in the form proposed by Vashishta
and Rahman,2 the S(k) for molten AgCl and AgBr is
improved,3,4 and that for molten silver iodide is suc-
cessfully reproduced.5,6 Recently, by molecular dynam-
ics (MD) simulation of a PIM for molten CuI, we have
Figura 1. S(k) for molten CuI from ND data of Drewitt et
also been able to reproduce the main trends of the corre-
al.8 (dotted line), and MD results for the PIM (solid line) and
sponding S(k), especially the first sharp diffraction peak RIM (dashed line) models.
at about 1 Å which indicates an intermediate range or-
dering related to the formation of voids with different
size due to the inhomogeneous distribution of cations.7
See Figure 1 where the experimental S(k) is compared
with that obtained by MD of a RIM and a PIM. In this ∗
[Link]@[Link]
work, we complete the series of copper halides with MD 1
M. P. Tosi and F. G. Fumi, J. Phys. Chem. Solids 25, 45
of PIM for molten CuBr and CuCl, and we report the (1964). R. Car and M. Parrinello.
corresponding structural and ionic transport properties. 2
P. Vashishta and A. Rahman, Phys. Rev. Lett. 40, 1337
The calculated S(k) also reproduce the main trends ob- (1978)
served in neutron diffraction (ND) data. Furthermore, we 3
J. Trullàs, O. Alcaraz, L. E. González and M. Silbert, J.
study the polarization effects on the properties of molten Phys. Chem. B 107, 282 (2003)
4
copper halides by comparing the PIM and RIM results. V. Bitrián and J. Trullàs, J. Phys. Chem. B 112, 1718
(2008)
5
V. Bitrián, J. Trullàs and M. Silbert, J. Chem. Phys. 126,
021105 (2007)
6
V. Bitrián and J. Trullàs, J. Phys. Chem. B 112, 1718
(2008)
7
V. Bitrián, O. Alcaraz and J. Trullàs, J. Chem. Phys. 134,
044501 (2011)
8
J. E. Drewitt, P. S. Salmon, [Link] and Y. Kawakita,
J. Phys.: Condens. Matter 21, 075104 (2009)

Panel P–6 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 61

Application of an effective medium theory to heterogeneous reaction-diffusion systems

Sergio Alonso∗ and Markus Bär

Physikalisch-Technische Bundesanstalt Department 8.4 Mathematical Modelling and Data Analysis Abbestr. 2-12 10587
Berlin Germany

The small scale of the heterogeneities usually allows predictions of the effective homogeneous medium theory.
an effective description of the inhomogeneous system. We also discuss the applicability of the effective medium
Specific homogenization theories have been already sug- theory to slow dynamical evolution of the domains 2.
gested, but we propose a general effective medium the- Some of the results are applied to chemical waves in
ory based on the homogeneization of reaction-diffusion microemulsions4 and to electric waves in heterogeneous
systems1,2 . We consider a system where domains of phase cardiac medium.
2 (heterogeneities) are randomly dispersed in a medium
of phase 1. The reactivity and diffusion of the reactants
∗
take different values if they are inside or outside of such [Link]@[Link]
1
domains. If the heterogeneities are small, we can calcu- S. Alonso, R. Kapral and M. Bär, Phys. Rev. Lett. 102,
late effective values for the diffusion and the reactivity. 238302 (2009).
2
S. Alonso, M. Bär and R. Kapral, J. Chem. Phys. 131,
Here, we validate the predictions of the effective medi- 214102 (2009).
um theory by the numerical calculation of the velocity 3
S. Alonso, J. Löber, M. Bär and H. Engel, Eur. Phys. J.
of a wave in reaction-diffusion systems under the pres- Special Topics, 187, 31 (2010).
4
ence of static obstacles. We compare the numerical re- S. Alonso, K. John and M. Bär, J. Chem. Phys. in press
sults obtained for bistable2 and excitable3 media with the (2011).

Junio de 2011, Barcelona Panel P–7

62 XVII Congreso de Fı́sica Estadı́stica FisEs11

Estudio Monte Carlo de la fase de spin-glass de sistemas diluidos de dipolos de Ising

Juan J. Alonso1,∗ , Julio F. Fernández2

1
Departamento de Fı́sica Aplicada 1, Universidad de Málaga, E-29071 Málaga (España)
2
ICMA, CSIC y Universidad de Zaragoza, E-50009 Zaragoza (España)

El comportamiento colectivo de sistemas de espines Presentaremos resultados Monte Carlo para sistemas
en los que las interacciones magnéticas dipolares son las de dipolos con dilución por sitios. Los dipolos ocupan
dominantes ha cobrado un renovado interés. Estos sis- una fracción x de los L3 sitios de una red SC, y to-
temas son raros de encontrar en la naturaleza, aunque dos ellos están alineados a lo largo de un mismo eje de
algunos materiales, como ferroeléctricos y cristales anistropı́a cristalina. Para xc < x ≤ 1, con xc  0.65 en-
magnéticos como el LiHoF4 pueden ser bien descritos contramos una fase antiferromagnética por debajo de una
como sistemas de dipolos. Parte de este interés en los temperatura de transición que tiende a 0 para x → xc .
sistemas dipolares proviene también de los avances en Para sistemas aún más diluidos encontramos una fase
la sı́ntesis de arrays de nanopartı́culas. En estos sis- de equilibrio de spin-glass por debajo de una temper-
temas, nanopartı́culas de hasta unos miles de magne- atura kB Tsg  xεd , donde εd es una energı́a de interac-
tones de Bohr se comportan como simples espines, y, ción dipolar entre vecinos próximos. Para caracterizar
cuando están empaquetados en arrays cristalinos, las in- el posible orden de spin-glass hemos estudiado el over-
teracciones dipolares crean orden magnético a temperat- lap de SG q. En particular, (a) la desviación cuadrática
uras accesibles experimentalmente. La anisotropı́a jue- relativa Δ2q , y (b) ξL /L, donde ξL es una longitud de
ga un papel importante en este orden dipolar. Nosotros correlación de SG. De su dependencia con la temperatu-
conside-ramos aquı́ sistemas en los que la barrera de en- ra y el tamaño del sistema determinamos Tsg . En la fase
ergı́a que un dipolo debe superar para invertir su signo de SG encontramos: (i) que | q |
y q 2
decrecen al-
es algo mayor que la energı́a dipolar, de forma que se gebraicamente con L al aumentar L, (ii) distribuciones
puedan observar efectos colectivos a temperaturas que bimodales anchas de q/ | q |
, (iii) que ξL /L crece con
no llegan congelar las direcciones de los espines. Este L hacia valores finitos. Todo ello es consistente con or-
tipo de sistemas pueden ser modelados como cristales de den débil de largo alcance en la fase de SG.1 . Finalmente,
dipolos de tipo Ising. mostraremos resultados para sistemas diluidos de dipolos
de Ising en redes bidimensionales cuadradas que indican
Nosotros estudiamos aquı́ el comportamiento de sis- que esta fase de equilibrio de SG no existe en sistemas
temas diluidos de dipolos de Ising. El propósito funda- bidimensionales.2
mental es investigar si la frustración que proviene del
carácter anisotrópico de la interacción dipolar, junta-
mente con el desorden que proviene de la dilución, puede
∗
dar lugar a la existencia de una fase de equilibrio de spin- jjalonso@[Link]
1
glass (SG). Experimentos con este tipo sistemas muestran J. J. Alonso and J. F. Fernández, Phys. Rev. B 81, 064408
comportamientos de no equilibrio de tipo SG, como ag- (2010).
2
ing, y relajación no exponencial. J. J. Alonso and B. Allés, Phys. Rev. B 82, 064425 (2010).

Panel P–8 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 63

Vidrios de espı́n en cuatro dimensiones en presencia de un campo magnético

R. Álvarez-Baños∗ , A. Cruz, L. A. Fernandez, J. M. Gil-Narvion, A. Gordillo-Guerrero, D. Iñiguez, A. Maiorano, F.

Mantovani, E. Marinari, V. Martin-Mayor, J. Monforte-Garcia, A. Muñoz-Sudupe, D. Navarro, G. Parisi, S.
Perez-Gaviro, J. J. Ruiz-Lorenzo, S. F. Schifano, B. Seoane, A. Tarancon, R. Tripiccione, D. Yllanes
Instituto de Biocomputación y Fı́sica de Sistemas Complejos
Mariano Esquillor, Edificio I+D
50018 Zaragoza

La solución exacta en la aproximación de campo medio

de los vidrios de espı́n predice caracterı́sticas muy intere-
santes, como la existencia de un número infinito de esta-
dos puros y su organización ultramétrica, que hacen de
estos materiales el paradigma de los sistemas complejos.
Actualmente se sabe que los vidrios de espı́n en tres
dimensiones presentan una transición de fase entre un
estado desordenado paramagnético y una fase de bajas Figura 1. Diagrama de fases en el espacio (T,h), de acuer-
temperaturas conocida como fase spin glass. Sin embar- do con la solución de campo medio (a) y el modelo droplet
go, la naturaleza de esta fase de bajas temperaturas es (b). Figura tomada de la referencia6 .
todavı́a cuestión de debate. El objetivo de las simula-
ciones numéricas se centra en determinar cuáles de las
predicciones de la solución de campo medio, exacta en
dimensión infinita, sobreviven en vidrios de espı́n más
∗
realistas de tres dimensiones, descritos generalmente por [Link]@[Link]
1
el denominado modelo de Edwards-Anderson1,2. M. Mezard, G. Parisi y M. A. Virasoro, Spin Glass theory
Una manera de llevar a cabo esta tarea consiste en el and beyond (World Scientific, 1994)
2
estudio del comportamiento de vidrios de espı́n en pres- G. Parisi, Field Theory, disorder and simulations (World
encia de un campo magnético. La solución de campo Scientific, 1994).
3
medio o solución de Parisi predice un transición de fase E. Marinari, G. Parisi, F. Ricci-Tersenghi, J. J. Ruiz-
a temperatura finita, de manera que existe una lı́nea de Lorenzo, F. Zuliani, Replica Symmetry Breaking in Short
transiciones en el espacio (T, h), conocida como lı́nea de Range Spin Glasses: A Review of the Theoretical Founda-
tions and of the Numerical Evidence, cond-mat/9906076
Almeida-Thouless, que separa la fase paramagnética a
(1999), J. Stat. Phys. 98, 973 (2000).
altos campos y temperaturas de la fase spin glass 3 . Por 4
D. S. Fisher, D. A. Huse, Phys. Rev. Lett. 56, 1601 (1986);
el contrario, los escenarios droplet y TNT predicen la
Phys. Rev. B 38, 386 (1988).
desparición de dicha transición incluso en presencia de 5
F. Krzakala, O. C. Martin, Phys. Rev. Lett. 85, 3013
un campo externo infinitesimal4,5 . Estudios numéricos (2000); M. Palassini, A. P. Young, Phys. Rev. Lett. 85,
previos ofrecen evidencias tanto a favor como en contra 3017 (2000).
de la existencia de la linea de AT. El objetivo de este 6
A. P. Young y Helmut G. Katzgraber, Phys. Rev. Lett. 93,
trabajo es determinar el comportamiento de vidrios de 207203 (2004).
espı́n en cuatro dimensiones, en los que se espera que
una transición de fase sea más fácil de detectar.

Junio de 2011, Barcelona Panel P–9

64 XVII Congreso de Fı́sica Estadı́stica FisEs11

Caos y arritmogénesis en la dinámica del calcio intracelular

E. Alvarez-Lacalle ∗ , I. R. Cantalapiedra, B. Echebarria, A. Peñaranda

Departament de Fı́sica Aplicada, Universitat Politècnica de Catalunya, Barcelona.

La muerte súbita debido a arritmias cardiacas es una tracelular de un batido a otro. La contracción celular
de las principales causas de muerte en el mundo indus- se activa con la depolarización de la célula, que provo-
trializado. El tipo más grave de arritmia es la fibrilación ca la apertura de los canales de Ca de la membrana,
que se produce cuando las fibras cardiacas se contraen permitiendo la entrada de Ca al interior de la célula.
de forma caótica y ası́ncrona, lo que impide el correcto El aumento del calcio intracelular origina la liberación
bombeo de sangre del corazón. Se ha estudiado exten- de Ca del retı́culo sarcoplasmático (RS) a través de los
samente el efecto de un desajuste en el balance de las canales de receptores de Rianodina (RyR) (Calcium in-
corrientes iónicas en las arritmias cardiacas, a través de duced Calcium release, CICR). Parte de este Ca se ad-
las variaciones que producen en el potencial de mem- hiere al sarcómero originando la contracción de la célula.
brana y la forma del potencial de acción. En la uĺtima La relajación hacia un nuevo ciclo se consigue mediante
década, sin embargo, se ha demostrado que desajustes la eliminación de Ca hacia el exterior mediante un inter-
en la regulación del calcio intracelular pueden contribuir cambiador y la reacumulación de Ca en el RS a través de
a la generación de arritmias. El acoplamiento electro- una bomba (SERCA).
mecánico hace que ambas causas estén interrelacionadas Experimentalmente se han observado situaciones pa-
tológicas con alternancia en la contracción entre un bati-
do y otro debido a la existencia de alternans en la con-
centración de Ca intracelular1 . Generalmente aparecen
cuando el ritmo cardiaco aumenta, aunque en algunos
casos, se pueden producir a bajas frecuencias de ex-
citación2,3 .
En esta contribución analizamos los efectos que tienen
posibles cambios en la activación, inhibición y recu-
peración de los receptores de rianodina en la aparición
de alternans en la concentración de calcio en un módelo
de corazón de conejo4 . Como se observa en la Fig.1
es posible conseguir sucesivas duplicaciones de periodo
modificando los parámetros de control del RyR, incluso
llegando a obtener dinámicas en apariencia caóticas .

∗
enrical@[Link]
1
Blater L.A. et al, J. Physiol 546 19 (2003)
Figura 1. Aparición de alternans en la concentración de Aistrup [Link] al, Cir. Res. 99 65, (2006)
2 ¯
calcio intracelular debido a la disminución del tiempo de in- Diaz M.E., O’Neill S.C. and Eissner D.A., Cir. Res. 99,
activación del receptor de rianodina. 740 (2006)
3
Picht E., DeSantiago J., Blatter L.A., Bers D.M., Cir. Res.
99 740 (2006)
4
En este trabajo estamos interesados en arritmias pro- Shannon T. R., Wang F., Puglisi, J., Weber, C., Bers,
ducidas por alternancia en la concentración de calcio in- D.M. Biophys. J. 87 87 5 (2004)

Panel P–10 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 65

Aggregation of superparamagnetic colloids in magnetic fields: the quest for the

equilibrium state

J. S. Andreu†,‡,∗ , J. Camacho† , J. Faraudo‡

‡
Institut de Ciència de Materials de Barcelona, Consejo Superior de Investigaciones Cientı́ficas (ICMAB-CSIC)
08193 Campus de la UAB, Bellaterra, Spain.

Figura 1. Left: Snapshot corresponding to the simulated system with φ0 = 1.05 × 10−3 (equivalent to c ≈ 1g/l) and Γ = 10.
We can observe that chains of colloids align parallel to the applied external uniaxial magnetic field. Right: Time evolution of
the mean chain length for Γ = 10 and different concentrations. After a fast transient regime, the mean chain length reaches a
constant value depending on the initial volume fraction of colloids φ0 and the magnetic coupling parameter Γ.

Previous experimental1 and simulation2 studies of su- This work is supported by the Spanish Govern-
perparamagnetic colloids in a strong external field have ment (grants FIS2009-13370-C02-02, PET2008-02-81-
systematically shown a nonequilibrium aggregation pro- 01/02 and CONSOLIDER-NANOSELECT-CSD2007-
cess in which chains of particles steadily grow in the direc- 00041), the Catalan Government (grant 2009SGR164)
tion of the applied external field with an average length and SEPMAG Tecnologies SL. We acknowledge com-
increasing with a power law with time. puter resources and technical assistance provided by the
Here we show3 , by employing Langevin dynamics sim- CESGA Supercomputing Center (Finisterrae Supercom-
ulations, the existence of a different behavior under the puter).
effects of an external magnetic field: after a transient
period of chain formation, the system attains an equilib-
∗
rium distribution of chain lengths. Furthermore, a ther- jandreu@[Link]
†
modynamic self-assembly theory supports the simulation Universitat Autònoma de Barcelona.
‡
results and it also predicts that the average chain length Institut de Ciència de Materials de Barcelona.
1
in the equilibrium state depends only on a dimension- P. Domı́nguez-Garcı́a et al., Phys. Rev. E 76, 051403
less parameter combining the volume fraction of colloids (2007).
2
φ0 and the magnetic coupling parameter Γ. The condi- F. Martinez-Pedrero et al., Phys. Rev. E 76 011405(2007).
3
tions under which this new behavior can be observed are J. S. Andreu, J. Camacho, J. Faraudo, Soft Matter
discussed. doi:10.1039/c0sm01424a.

Junio de 2011, Barcelona Panel P–11

66 XVII Congreso de Fı́sica Estadı́stica FisEs11

Experimental study of different LFF regimes in semiconductor lasers with an external

cavity

Andrés Aragoneses1 , Jordi Zamora-Munt, Jordi Tiana-Alsina, Nicolas Rubido, Cristina Masoller, Maria Carme
Torrent
Dept. Fı́sica i Enginyeria Nuclear2 , Universitat Politècnica de Catalunya. Rambla Sant Nebridi s/n, Terrassa, 08222 (Spain)

Optical feedback, due to an external mirror in a semi- times far from the fix point, showing longer intervals of
conductor laser, induces instabilities in the laser pow- LFF’s, and the rest of the time in the attractor, with cw
er output. When the laser works close to its solitary emission. At pump currents higher than 21.75 mA the
threshold and it is subjected to a small or moderate feed- system shows a complete LFF regime (e in figure 1).
back, sudden irregular power dropouts, followed by grad-
ual power recoveries, appear. These chaotic behaviour,
known as Low Frequency Fluctuations (LFF), may show
different statistics3–5 , depending on the system’s param- 2
PDF of cw intervals
10 (I=21.50mA)
eters. For pump currents just above the laser’s thresh-
old the system may be in a transient regime, where the
dropouts are rare and it is predominantly in continuous
wave (cw) emission. Increasing slightly the pump cur- 10
1

rent, the system gets to a coexistence regime of cw and

LFF. For higher currents, the system shows a sustained
LFF regime, with very short periods of cw emission.
For a constant feedback rate and different pump cur- 10
0

rentes (20.50mA to 22.00mA), we have studied the tran- 0 5 10 15 20

sition, as the system passes from the transient to the Time (μs)
coexistence and to the LFF regimes. In figure 1 the time Figura 2. Probability Distribution Function for the contin-
series of the power output, for the different regimes stud- uous wave intervals. The linear plot shown in the lin-log scale
ied, are shown. In order to characterize the transient can be described in terms of Kramers’ formula.
and the coexistence regimes, we have obtained long time
statistics of the bursts of LFFs and of the periods of sta-
ble emission. The probability distribution functions (PDF) we ob-
tain show an exponential decay behaviour for the cw in-
tervals (see figure 2). This can be explained in terms
(a) of the Kramers’ formula. But the PDF of the bursts of
power output (arb. units)

50000 LFF show some structure that makes it different from

(b)
an exponential decay. In order to distinguish between
(c) a chaotic or an stochastic origen of the jumps from the
0 bursts of LFF to the cw and vice-versa, we also follow
(d)
a complexity analysis of the time intervals in the coex-
istence regime. In this analysis we transform each time
-50000
(e)
series intervals into a symbolic series, so we can study
the probabilities of these symbols.
2510 2520 2530 2540 2550
Time (μs)
1
Figura 1. Time series for the different pump currents [Link]@[Link]
2
studied: Ia = 21.00mA, Ib = 21.25mA, Ic = 21.50mA, [Link]
3
Id = 21.75mA, Ie = 22.00mA. The threshold of the laser T. Heil, I. Fisher, W. Elsäβer, J. Mulet, C. R: Mirasso
was 20.64mA, and the feedback used reduced it 10% ”Statistical properties of low-frquency fluctuations during
single-mode operation in distributed-feedback lasers: ex-
periments and modeling,24, 18, p. 1275 (1999)
The transient regime, (a in figure 1), is interpreted as 4
A. Torcini, S. Barland, G. Giocomelli, F. Marin ”Low-
the system being near a fix point in the phase space. It frequency fluctuatinos in vertical cavity lasers: Experi-
sporadically jumps from it due to noise, and it quick- ments versus Lang-Kobayashi dynamics”, Phys. Rev. A,
ly returns and remains stably there for a long period. 74, p. 063801 (2006)
When increasing the pump current to around 21.25mA, 5
J. Zamora-Munt, C. Masoller, J. Garcı́a-Ojalvo ”Transient
the system gets to the coexistence of bursts of LFF and low-frequency fluctuations in semiconductor lasers with
cw emission (b-d of figure 1). The system spends longer ioptical feedback”, Phys. Rev. A, 81, p. 033820 (2010)

Panel P–12 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 67

Evolución temporal en el flujo longitudinal uniforme de un gas granular Estudio

mediante simulaciones en computación grid.

Antonio Astillero∗ , Andrés Santos†,‡

Departamento de Tecnologı́a de Computadores y Comunicaciones
Universidad de Extremadura, 06800 Mérida (Badajoz)

Un gas granular es un tipo particular de sistema com- ∂t f + vx ∂x f = J[f, f ]. (4)

plejo donde una de las caracterı́sticas más sobresalientes
que lo definen es la disipación de energı́a que se produce A continuación efectuamos el siguiente cambio de varia-
cuando tiene lugar una colisión entre partı́culas. Como bles
consecuencia, no se verifica el principio de conservación x
de la energı́a, haciendo que la dinámica interna del gas x→s= , (5)
1 + a0 t
sea mucho más rica que la correspondiente a su homólogo
molecular.
Una de las formas más extendidas de modelar un gas t → τ = a−1
0 ln (1 + a0 t) , (6)
granular es considerar que está formado por esferas duras
inelásticas, donde el grado de inelasticidad de las coli-
a0 x
siones viene determinado por el coeficiente de restitución v→V=v− ,
x (7)
normal α constante1 . 1 + a0 t
En un trabajo anterior2 , estudiamos la etapa donde a0 es el gradiente de velocidad inicial. Teniendo
hidrodinámica transitoria en el caso particular del flu- en cuenta el cambio anterior, la ecuación (4) queda como
jo longitudinal uniforme (ULF),3 tanto desde un punto
de vista analı́tico como computacional (DSMC)4 . En di- ∂  
cho trabajo, referimos las velocidades de las partı́culas ∂τ f + Vx ∂s f − a0 Vx f = J[f , f ]. (8)
∂Vx
al sistema de referencia lagrangiano que se mueve con
la velocidad media del fluido. Haciendo este cambio de Entonces, para poder generar el flujo longitudinal uni-
sistema de referencia nuestro problema original se trans- forme y realizar simulaciones DSMC inhomogéneas pode-
forma en uno independiente del espacio, lo cual permite mos utilizar las variables escaladas (5)–(7) y aplicar
simular la ecuación de Boltzmann utilizando DSMC de condiciones de contorno periódicas en s = ±L0 .
un modo relativamente sencillo sin necesidad de dividir Hemos efectuado simulaciones para α = 0.5 y a0 =
el sistema en celdas. −4 (τ 0 )−1 , −10 (τ 0 )−1 , analizando la evolución temporal
En el presente trabajo intentamos responder las sigui- de los perfiles espaciales de densidad, velocidad media y
entes cuestiones: ¿qué condiciones de contorno es preciso temperatura, respectivamente. Los resultados obtenidos
imponer para que podamos simular el flujo longitudinal muestran cómo el sistema, partiendo del estado inicial
inhomogéneo o, en otras palabras, dependiente del es- definido por (1)–(3), evoluciona durante la etapa cinética
pacio? ¿evoluciona un estado inicial inhomogéneo hacia hasta alcanzar las condiciones que definen el flujo longi-
el estado homogéneo? ¿cuál es la duración tı́pica de la tudinal uniforme.
relajación hacia el estado homogéneo? Partimos de la
condición inicial no homogénea definida por la función ∗
de distribución de velocidades5 : aavivas@[Link]
[Link]
†
n0 (x)  0 0
 Departamento de Fı́sica, Universidad de Extremadura,
f (r, V; 0) =
0 2 δ |V − u (x)| − V , (1) 06071 Badajoz
4πV ‡
andres@[Link]
donde los campos de densidad inicial y velocidad media [Link]
1
inicial están dados por N. Brilliantov and T. Pöschel, Kinetic theory of granular
gases (Oxford University Press, Oxford, 2004).
  2
A. Astillero and A. Santos, Régimen hidrodinámico transi-
1 2πx
n0 (x) = n 1 + sin , (2) torio de un gas granular en un flujo compresible, congreso
2 L
Partı́culas y flujos, Pamplona (2010).
3
A. Santos, in Rarefied Gas Dynamics: Proceedings of the
  26th International Symposium on Rarefied Gas Dynamics,
0 πx 2
u (x) = U cos − ,
x (3) Takashi Abe, ed. (AIP Conference Proceedings, vol. 1084,
L π Melville, NY, 2009), pp. 93-98
4
G. A. Bird, in Molecular gas dynamics and the direct sim-
respectivamente. Por contra, la temperatura inicial T 0 = ulation of gas flows, Oxford Science Publications, Oxford
2
mV 0 /3 es homogénea. La ecuación de Boltzmann para (1994).
5
un estado que sólo presenta gradientes a lo largo de la A. Astillero and A. Santos, Phys. Rev. E 72, 031309-1-20
dirección x tiene la forma (2005).

Junio de 2011, Barcelona Panel P–13

68 XVII Congreso de Fı́sica Estadı́stica FisEs11

Effects on the histeresis in martensitic transformations

V. Torra1 , A. Isalgue1 , C. Auguet1,2∗ , G. Carreras1

1
CIRG-ETSECCPB-DFA, Universitat Politècnica de Catalunya, Barcelona
2
EPSEB- DFA, Universitat Politècnica de Catalunya, Barcelona

Shape Memory Alloys (SMA) show a solid-solid struc- to changes in the internal state of the material: plastic
tural first order phase transformation from austenite to deformation in NiTi and plastic deformation plus stabi-
martensite when a stress is applied in austenite phase, lized martensite in Cu-based alloys.
producing changes in length relatively relevant (up to 8-
We studied the influence of the self-heating in the
10 %). When the stress is suppressed, the martensite
width of the hysteresis in NiTi wires with cycling fre-
retransforms to austenite. In this way, a hysteresis cy-
quency. In quasi-static cycling it is possible to see the
cle stress-strain is obtained. The necessary stress mean
increase of the stress due to the local self-heating. The
value to induce the transformation is dependent on the
hysteresis width, for several frequencies, shows the effects
temperature according to the Clausius-Clapeyron ther-
of coupling between temperature and stress by a Gaus-
modynamic equation: as higher is the local temperature,
sian shape. An outline of the hysteresis width for CuAlBe
higher is the necessary stress to produce the transforma-
is, also, determined. Minor dynamic actions associated
tion. The hysteresis cycle induces the damping property
to cycling-pause effects are also visualized1,2 .
of the SMA transforming the mechanical energy in heat,
and this property enables the use of SMA as dampers
(passive smart materials) in Civil Engineering. The hys- ∗
carlota@[Link]
teresis cycle induced by the differences between transfor- 1
V. Torra, A. Isalgue, F. Martorell, F. C. Lovey and Patrick
mation and retransformation stresses and the associated Terriault, Damping in Civil Engineering using SMA. Part
latent heat (dissipated and absorbed) produces the self- I: Particular Properties of CuAlBe for Damping of Family
heating in the SMA. Houses, Can. Metall. Quart., 49(2), 179-190 (2010)
Also, in the first cycles of working, it is observed that 2
V. Torra, A. Isalgue, C. Auguet, G. Carreras, F.C. Lovey,
the hysteresis cycles evolve progressively decreasing their P. Terriault, Damping in Civil Engineering using SMA.
width and reducing the available useful length of the Part II. Particular properties of NiTi for damping of
samples (SMA creep), and consequently, decreasing the stayed cables in bridges, submitted to Can. Metall. Quart.
damping effect. This is an intrinsic effect mainly related (2010)

Panel P–14 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 69

Buckling transition in quasi-spherical viral capsids

Marı́a Aznar, Antoni Luque, David Reguera

Universidad de Barcelona, Dpto. Fı́sica Fundamental, Martı́ i Franqués, 1,
08028 Barcelona Spain.

Viruses are fascinating biological organisms which can them to become infective, where the capsid undergoes
self-replicate and infect all types of cells. In their sim- a buckling transition in which its shape changes from
plest form, they are essentially composed of a protein spherical to a flattened icosahedron. In this poster, we
shell or capsid that encloses and protects the genetic ma- will present the results of our studies aimed to analyze the
terial, RNA or DNA. About half of all viral species have phenomenon of buckling in quasi-spherical viral capsids.
a quasi-spherical capsid with icosahedral symmetry char- In particular, we will discuss how this buckling transi-
acterized by a triangulation number T. This T number, tion depends on the virus triangulation number T and
which adopts a discrete set of values, serves as a struc- especially on its class P. We will also discuss the poten-
tural classification for viruses [1]. In turn, viruses with tial biological implications of this buckling phenomenon
different T numbers can be grouped into families with in the viral stability and life cycle.
similar geometrical arrangements of proteins, that are References [1] D.L.D Caspar and A. Klug, Quant.
labelled using a second number P. The mechanical prop- Biol. 27, 1-24 (1962)
erties of these viral capsids play a major role in their [2] R. Zandi and D. Reguera, Phys. Rev. E. 72, 021917
stability [2] and have been shown to depend on their (2005)
structure and T number [3]. [3] C. Carrasco, M. Castellanos, P.J. de Pablo, M.G.
Many viruses have a maturation stage, necessary for Mateu, Proc. Nat. Acad. Sci. USA 105, 4150-5 (2008).

Junio de 2011, Barcelona Panel P–15

70 XVII Congreso de Fı́sica Estadı́stica FisEs11

Voter models on weighted networks and limits of the mean-field approach

Andrea Baronchelli1 , Claudio Castellano2 , Romualdo Pastor-Satorras1

1. Departament de Fı́sica i Enginyeria Nuclear, Universitat Politècnica de Catalunya, C. Nord B4, 08034 Barcelona, Spain
2. Istituto dei Sistemi Complessi (ISC-CNR), and Dipartimento di Fisica “Sapienza” Università di Roma, [Link] A. Moro 2,
I-00185 Roma, Italy

The classical Voter model1 is a paradigm in the statis- point out that the interplay between the degree expo-
tical physics approach to social modeling2 . It describes nent and the link strength yields an extremely complex
the spreading of opinions/ideas based on a pure imita- phenomenology that can be rationalized in a phase plane
tion process: at each time step an individual is select- adopting simple assumptions. We also introduce the con-
ed and copies the opinion of a neighbor of hers. In the cept of annealed weighted network and discuss how the
most common version only two opinions compete in the validity of heterogeneous mean field theory is, strictly
population, and in finite systems all individuals ends up speaking, limited to these structures4 . However, we show
sharing the same one (i.e. consensus is reached). The that the same approach is able to describe the model
model has been extensively studied on different topolo- accurately also on quenched weighted graphs, provided
gies, ranging from regular lattices to complex binary net- the link strength is not too strong. We finally discuss
works. However, the role of heterogeneous connectivity the implications of our findings for the modeling of more
patterns has not been deeply investigated so far, even complex processes on complex weighted networks.
though many network structures found in technological,
biological or social contexts are intrinsically weighted, i.e.
∗
connections have widely varying strengths. [Link]@[Link]
1
P. Clifford and A. Sudbury, Biometrika 60, 581 (1973) and
Here we present a thorough analysis of the Voter mod- R. Holley and T. Liggett, Ann. Probab 3, 643 (1975).
el on complex weighted networks3. We write down het- 2
C. Castellano, S. Fortunato, and V. Loreto, Rev. Mod.
erogeneous mean field equations that allow us to esti- Phys. 81, 591 (2009).
mate the exit probability (i.e. the probability that a 3
A. Baronchelli, C. Castellano and R. Pastor-Satorras
given opinion will survive, depending upon the initial preprint arXiv:1011.2395 (2011).
4
conditions) and the consensus time. We focus on net- A. Baronchelli and R. Pastor-Satorras, Phys. Rev. E 82,
works characterized by power-law degree distribution and 011111 (2010).

Panel P–16 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 71

Percolative analysis of nanoconfined supercooled water

Valentino Bianco, Giancarlo Franzese∗

Departament de Fı́sica Fonamental, Universitat de Barcelona
Diagonal 647, 08028 Barcelona, Spain

Water is a fundamental substance for all processes a critical point thanks to a mapping into a percolating
involving life. It has many unusual properties com- model.
pared with other liquids. Recent studies of hydrated
biomolecules and nonoconfined water suggest that, sur- ∗
prisingly, the anomalous properties of water are consis- vbianco@[Link], gfranzese@[Link]
1
tent with the hypothesis of the occurrence of a first-order K. Stokely, M. G. Mazza, H. E. Stanley, G. Franzese, Proc.
Natl. Acad. Sci. U.S.A. 107, 1301-1306 (2010).
phase transition between two liquids with different den- 2
G. Franzese, F. de los Santos, J. Phys.: Cond. Matt. 21
sities at low temperature and high pressure. In this hy-
504107 (2009).
pothesis the phase transition line ends in a liquid-liquid 3
M. G. Mazza, K. Stokely, E. G. Strekalova, H. E. Stanley,
critical point. To validate this hypothesis experiments
G. Franzese, Computer Phys. Comunications 180, 497-502
in nanoconfinement could be very valuable, because they (2009).
allow to explore the water behavior in a region of temper- 4
P. Kumar, G. Franzese, and H. E. Stanley, Phys. Rev.
ature that would be otherwise inaccessible for bulk liquid Lett. 100, 105701 (2008).
water, as a consequence of inevitable crystallization. We 5
G. Franzese, H. Stanley, J. Phys.: Cond. Matt. 19, 205126
present here a coarse-grained model of water that give in- (2007).
sight into the physics of a nanoconfined water monolayer. 6
G. Franzese, H. Stanley, J. Phys.: Cond. Matt. 14, 2201-
Our model allows us to analyze the possible existence of 2209 (2002).

Junio de 2011, Barcelona Panel P–17

72 XVII Congreso de Fı́sica Estadı́stica FisEs11

Turbulent bubble dispersions in microgravity. Drop Tower experiments

Pau Bitlloch, Xavier Ruiz∗ , Laureano Ramı́rez-Piscina†, Jaume Casademunt

Departament d’Estructura i Constituents de la Matèria, Universitat de Barcelona
Av. Diagonal, 647, 08028 Barcelona

Dispersed multiphase flows are relevant to many tech- monodisperse turbulent bubble jets in microgravity2,3.
nological applications, often involving turbulent regimes. Here we present results of a new series of Drop Tower
Examples include particles suspended in liquid or gas experiments conducted at the ESA facilities at ZARM in
flows, droplets in gas flows or bubbly flows. This type of Bremen, that exploit the above bubble injection method
problems combines two of the most challenging problems to create a nearly uniform monodisperse bubble suspen-
in fluid mechanics and statistical physics: turbulence and sion carried by a turbulent duct flow.
multiphase flows. In particular, both for technological The experiment is designed to allow independent con-
and fundamental interests, it is important to character- trol of bubble size, bubble density and the degree of tur-
ize and understand the interaction between the carrying bulence of the carrying flow. Typical duct Reynolds num-
phase and the dispersed phase. This includes not only bers used are up to 12000. Typical bubble diameter is
how the carrying flow affects the transport and spatial 1.5 mm, in a duct of 80 cm long and 10x10 cm2 section,
distribution of the dispersed phase (one-way coupling) that is, significantly larger than the Kolmogorov turbu-
but also how the dispersed phase modifies the nature of lent scale, but significantly smaller than the scale of most
the turbulence of the carrying flow (two-way coupling)1 . energetic eddies. Bubble densities keep mean bubble dis-
tances comparable to the size of most energetic eddies.
This setup is designed to optimize spatial dispersion of
bubbles while minimizing the degree of coalescence. The
Webber number of bubbles is small enough to maintain
the spherical shape, but they cannot be considered as
point-like with respect to their interaction with the flow.
Under these circumstances, a significant degree of two-
way coupling between bubbles and turbulence is expect-
ed.
The experiment has proven succesful in creating near-
ly monodisperse bubble suspensions in turbulent flows in
microgravity, under controlled variations of the relevant
parameters. This is achieved for the first time. Quantita-
tive information is obtained through the appropriate dig-
ital image processing, based on particle tracking methods
that yield trajectories of individual bubbles. As a first
characterization, we have systematically measured the
mean velocity profile of bubbles and the velocity mean
square dispersion for different experimental conditions,
Figura 1. Snapshot of a monodisperse bubble suspension including variation of Reynolds number, bubble size and
of bubbles of diameter close to 1.5 mm in a turbulent flow in
bubble density. Preliminary results show that turbulent
microgravity.
kinetic energy tends to be decreased by the presence of
bubbles for increasing Reynolds number. In addition, our
data provide, for the first time, a quantitative character-
Here we address the problem of turbulent bubbly flows
ization of the decay of the so-called pseudo-turbulence,
in the absence of gravity. Efficient control of bubble for-
created by the bubbles previously formed with normal
mation and management of turbulent bubbly flows in mi-
gravity, before the drop release.
crogravity environments is indeed crucial for multiple ap-
plications in the rocket industry and for life support sys-
tems in space. However both practical procedures and ∗
Universitat Rovira i Virgili
fundamental aspects of the physics of bubbly flows in †
Universitat Politècnica de Catalunya
the absence of buoyancy are dramatically different from 1
S. Balachandar, J. K. Eaton, Annu. Rev. Fluid. Mech. 42,
those in normal gravity, and largely unknown because 111 (2010)
of lack of data. In particular, the controlled formation 2
J. Carrera, X. Ruiz, L. Ramı́rez-Piscina, J. Casademunt,
and management of a monodisperse bubble suspension M. Dreyer, AIAA Journal 46, 2010 (2008)
is by itself a practical challenge in the absence of buoy- 3
S. Arias, R. González-Cinca, X. Ruiz, L. Ramı́rez-Piscina,
ancy forces. Recently, a new method of bubble injection J. Casademunt, Colloids and Surfaces A: Physicochemical
has been successfully tested that allows the formation of and Engineering Aspects 365, 52-55 (2010)

Panel P–18 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 73

A nonlinear mechanism of cell motility in lamellar actomyosin fragments

Carles Blanch-Mercader∗ , Jaume Casademunt

Departament Estructura i Constituents de la Matèria, Universitat de Barcelona, Av. Diagonal 647, Barcelona, Spain

Directional cell locomotion is one of many biological formal mapping techniques. This is a very powerful tool
processes that depends on the development and mainte- for laplacian problems which allows to explore systemat-
nance of functional asymmetry between parts of a system ically the weakly nonlinear regime of the instability and
that were initially equivalent. Morphological instabili- thus have an analytical grasp of the nonlinear structure
ties may spontaneously break the symmetry but how the of the problem. At the same time, the method provides a
machinery of cell locomotion can couple to the morphol- powerful numerical scheme to find solutions deeply into
ogy and initiate/sustain motion is not well understood. the nonlinear regime.
In actin-based motility on a solid substrate, two main We have exploited this approach to pursue the insta-
mechanisms of force generation that are powered by the bility and have found explicit propagating solutions in
hydrolysis of ATP are present: the collective action of the problem without myosin, proving that motion can
molecular motors (myosin) that exerts contractile forces be sustained with only actin polymerization forces com-
on the actin network, and the polymerization of actin bined with the appropriate morphology of the fragment.
localized at the cell membrane. Both mechanisms are We show how a purely linear analysis of the morpho-
combined with friction (or adhesion) forces with the sub- logical instability misses this point, since motion is only
strate to achieve motion. initiated at the nonlinear level. Remarkably, preliminary
Lamellar fragments of keratocytes are pieces of the results also show that an adiabatic reduction of the prob-
motile machinery of these cells which lack nuclei, mi- lem exploiting the separation of time scales close to the
crotubules and most organelles, but retain the mini- instability threshold, may also miss the existence of a
mal set ingredients to generate motion. Experimental finite velocity, implying that the existence of a propagat-
observations1 show that the combination of these elemen- ing mode could be associated to a problem of so-called
tary acto-myosin machinery is indeed capable to gener- ’asymptotics beyond all orders’ such as that of the clas-
ate and sustain spontaneous directional motion through sical Saffman-Taylor problem4 .
some symmetry-breaking of these lamella. A theoreti-
cal understanding of this phenomenon is still lacking. In 1.0
particular it is not known whether myosin motors play
a fundamental role or are essentially auxiliary. Previous
theoretical studies2 based on the theory of active polar
gels3 have shown that a circular lamellar fragment under-
0.5

goes a morphological instability similar to that of viscous

fingering under centrifugal forcing, due solely to polymer-
ization forces, but no sustained motion have been found. 0.5 1.0 1.5

In the present work we demonstrate for the first time

that an inherently nonlinear mechanism of cell motili-
ty operates combined with the morphological instability 0.5

of lamellar fragments that is capable to sustain motion,

achieving a nontrivial stationary shape of the fragment.
We show that this mechanism is operative with only poly- 1.0

merization forces, in the absence of myosin. We base our Figura 1. Asymmetric shape of a cell fragment moving at
analysis on the framework of active polar gels in a con- finite velocity from right to left, under the polymerization
fined quasi-two-dimensional (Hele-Shaw) geometry. In dynamics of the actin gel, in a 2d confined geometry.
this regime and in the absence of myosin, the problem
can be mapped into a Laplacian free-boundary problem.
The model assumes a polar nematic continuous de-
scription of the gel of actin and assumes that the dy-
namics of the polymerization can be slaved to the slow ∗
blanch@[Link]
membrane dynamics. Under the appropriate conditions 1
A. B. Verkhovsky, T. M. Svitkina, and G. G. Borisy, Curr.
the flow of actin filaments associated to the treadmilling Biol. 9, 11 (1999)
process can be shown to satisfy Darcy’s law and therefore 2
A. C. Callan, J.F. Joanny, and J. Prost, Phys. Rev. Lett.
be reduced to a laplacian pressure field with appropriate 100, 258106 (2008)
boundary conditions at the moving boundary2 . Similarly 3
K. Kruse, J.F. Joanny, F. Jülicher, J. Prost, and K. Seki-
to the problem of viscous fingering in Hele-Shaw cells4 , moto, Eur. Phys. J. E 16, 5 (2005)
4
the problem can then be formulated by means of con- J. Casademunt, Chaos 14, 809 (2004)

Junio de 2011, Barcelona Panel P–19

74 XVII Congreso de Fı́sica Estadı́stica FisEs11

Camino aleatorio, subdifusión, memoria, series divergentes y funciones de Bessel de

primera especie

Santos Bravo Yuste∗ , Enrique Abad+

∗
Departamento de Fı́sica y + Departamento de Fı́sica Aplicada
Universidad de Extremadura
06006 Badajoz

La historia de la ciencia nos informa de los numerosos Λp [f ]/Λp [f ]x=0 sobre la función semilla fp,0 (x) = 1, sien-
hallazgos matemáticos que, de forma inesperada, surgen do
del estudio de problemas fı́sicos. En esta comunicación se
1 u
muestra cómo un problema relacionado con la difusión de du
Λp [f ] = zp2 dvv 2p+1 f (v).
cierta clase de caminantes aleatorios lleva al descubrim- x u2p+1 0
iento de una nueva forma de definir las funciones de
Bessel de primera especie Jp (x) y, además, al descubrim- (p)
Los polinomios Ban (x) o, equivalentemente, las fun-
iento de (infinitas) familias de polinomios cuyos miem- ciones Jp,n (x) pueden calcularse de forma recursiva me-
bros fp,n (x) son aproximaciones de Jp (x) que convergen diante una simple relación algebraica:2,3
hacia esta función cuando n crece.
El problema fı́sico a resolver consiste en hallar la prob- 
n
cn,n,p 2n+p
abilidad de supervivencia y la distribución espacial de Jp,n (x) = ck,n,p Jp,n−k (x) − x
caminantes aleatorios de tiempo de salto continuo (Con- 2p p!zp2n
k=1
tinuous Time Random Walk, CTRW) donde el tiempo
de espera entre saltos está descrito por una distribución Los coeficientes numéricos ck,n,p pueden evaluarse tam-
de cola larga: ω(t) ∼ t−1−γ con 0 < γ < 1. Los cami- bién de forma recursiva. Sorprendentemente, resulta que
nantes desaparecen cuando llegan a la superficie de una estas funciones constituyen aproximaciones polinómicas
esfera d-dimensional. Un modo eficiente de resolver prob- de la función de Bessel de primera especie Jp (x). De
lemas de este tipo pasa por expresarlos en términos de hecho es posible demostrar que Jp,n (x) → Jp (x) cuan-
ecuaciones difusivas. En el caso de distribuciones de cola do n → ∞, lo cual constituye una definición alterna-
larga, la ecuación difusiva de la concentración c(r, t) de tiva de la función Jp (x). Otras elecciones de funciones
caminantes es fraccional semilla fp,0 (x) conducen a otras familias de funciones
fp,n (x) = Λ[fp,n−1 (x)] las cuales también dan lugar a
dγ c(r, t)
= K∇2 c(r, t), familias de funciones que tienden a Jp (x) cuando n crece.
dtγ
El operador dγ /dtγ es una derivada fraccional de Ca-
puto. La ecuación puede resolverse por separación de 1.0

variables de modo que la solución se expresa como 0.8

una superposición de modos (sub)difusivos: c(r, t) =
 ∞ −nγ
n=0 an ψn (r)φn (t), donde φn (t) ∼ t
0.6
. La parte es-
pacial de los modos ψn (r) es interesante por dos razones. 0.4

Primero, ψn (r) contiene información de la condición ini- 0.2

cial para todo n. Por tanto, incluso para t → ∞, la condi-
x
ción inicial del problema puede reconstruirse mediante el 5 10 15 20

análisis de Fourier del modo espacial dominante ψ1 (r). 0.2

Esto es una manifestación clara de efectos de memoria de 0.4

largo alcance en problemas con caminantes de cola larga. Figura 1. Función J0 (x) (linea quebrada) y las primeras 21
El segundo aspecto interesante es que ψn (r) tienen la for- funciones polinómicas J0,n con n = 0, 1, . . . 20 (lı́neas contin-
ma de una serie de Fourier-Bessel divergente. En algunos uas).
problemas concretos hemos sido capaces de sumar las se-
ries resultantes.1 Por ejemplo, cuando la concentración
inicial de caminantes es constante c(r, 0) = c0 , la suma
de las series divergentes conduce a funciones polinómicas
Jd/2−1,n (x) definidas por ∗
santos@[Link]
  +
eabad@[Link]
xp (p) x 1
Jp,n (x) = p Ban S. B. Yuste, R. Borrego and E. Abad E Phys. Rev. E 81
2 p! zp (2010) 021105.
2
(p) S. B. Yuste and E. Abad J. Phys. A: Math. Theor. 44
donde Ban (x) son polinomios de grado 2n y zp es el (2011) 075203.
primer cero de Jp (x). Estos polinomios se generan medi- 3

ante la aplicación sucesiva del operador integral Λ̂p [f ] = [Link]

Panel P–20 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 75

Puzzling Out the Mechanical Stability and the Dynamics of the DV Organizer in the
Wing Imaginal Disc of Drosophila

Javier Buceta∗
Parc Cientific de Barcelona
CoSMo Lab (Computer Simulation and Modeling)
C/ Baldiri Reixac 10-12, 08028 Barcelona, Spain

During the course of development, tissues are sub- cycle, the cell growth, and the cellular interactions, in
jected to forces that shape the primordia. Within this order to shape the dorsal-ventral organizer as a robust
context we have studied the dynamics and stability of source of positional information and a lineage controller.
the dorsal-ventral organizer of the wing imaginal disc of All in all, our results provide novel insights into the de-
Drosophila as cell proliferation advances. Our approach velopmental mechanisms that drive the dynamics of the
is based on a generalized vertex model to perform in sil- DV organizer and set a definition of the so-called Notch
ico experiments that is fully dynamical and takes into fence model in quantitative terms.
account the available experimental data. Thus, in this
talk ’ll show that our results shed light on the complex
∗
interplay between the cytoskeleton mechanics, the cell [Link]@[Link]

Junio de 2011, Barcelona Panel P–21

76 XVII Congreso de Fı́sica Estadı́stica FisEs11

Grand Canonical Monte Carlo simulations of adsorption using functionalized

amorphous silica

Santiago Builes1,2 , Lourdes F. Vega1,2,3∗

1
MATGAS Research Center. Campus UAB, 08193 Bellaterra, Barcelona
2
Instituto de Ciencia de Materiales de Barcelona. (ICMAB-CSIC),Campus UAB, 08193 Bellaterra, Barcelona
3
Carburos Metálicos - Air Products Group. Campus UAB, 08193 Bellaterra, Barcelona

It is possible to merge the inherent sorptive behavior (surface silanols), we calculate the Rosenbluth factor for
of amorphous silica with organic groups; increasing the replacing the hydroxyl group for the first and second
adsorption capabilities of the solid silica. These function- bead in the chain. In our simulations, we consider as the
alized materials can be used to capture gases reversibly first atom in our chains the oxygen atom bonded to the
using adsorption. Effective design of these materials surface silica.
requires a method that can relate the structure of the
adsorbent to its performance. This level of understand- The results of the simulations are presented for the
ing can be achieved by using molecular simulations, as behavior of CO2 molecules captured on functionalized
they relate the microscopic behavior of the molecules amorphous silica. We calculate the adsorption isotherms
during the adsorption process to the macroscopic behav- and isosteric heats of adsorption. As well as specific the
ior of the system, allowing one to search for the best adsorption sites, and the orientation and distribution of
materials for separation purposes. the adsorbed molecules.

Although the silica material can be modeled as a Support for this work was provided by Air Products and
rigid structure, the functionalized chains in the surface the Spanish Government (project CEN2008-1027 CEN-
have to be allowed to move during the adsorption of IT SOST-CO2). Additional support from the Spanish
fluid molecules. The tethered molecules have branched Government (project CTQ2008-05370/PPQ) and by the
chains. Therefore, for the simulations of adsorption it is Catalan Government (2009SGR-666 and a TALENT to
necessary to use a method capable of moving branched S. Builes) is also acknowledged.
chains efficiently. The torsion and bending angles in the
surface groups can be handled using a coupled-decoupled
configurational bias algorithm1 . Additionally, we used ∗
vegal@[Link]
pregenerated Gaussian distributions for the probabili- 1
Martin, M. G.; Siepmann, J. I. J Phys Chem B 1999, 103,
ties of generating the bending and torsion angles for 4508.
2
the grafted molecules, which are then corrected in ac- Martin, M. G.; Frischknecht, A. L. Mol. Phys. 2006, 104,
ceptance rules2 . For all the possible substitution sites 2439

Panel P–22 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 77

Demographic growth and competition shape the size-area relationship for human
languages

Susanna C. Manrubia∗ , Jacob B. Axelsen, Damián H. Zanette†

Centro de Astrobiologı́a (INTA-CSIC), 28850 Torrejón de Ardoz, Madrid, Spain

According to linguists’ classifications, there are about The monotonically increasing function g(α) sets the de-
6.900 languages currently spoken on Earth1 . Among gree of correlation between size and area, and as such
them, 516 are classified as nearly extinct (with less than determines the value of the exponent z. We assume that,
100 speakers), while the ten most abundant languages contrary to the population size –which grows exponen-
are spoken by 2.6 × 109 people. The uneven distribution tially on the average, the total area available is constant.
of language sizes, measured as the number of speakers To fulfill this constraint we rescale at each time step the
per language, is well fit by a log-normal probability dis- areas, which amounts to dividing by the average value γ̄
tribution. Its functional form has been explained on the of γt : At+1 → At+1 /γ̄.
basis of a stochastic multiplicative growth of populations This model admits an analytical treatment in terms of
in the last thousand years2 . correlated random walks. The parameters of the mod-
We here present results on the correlation between the el can be put in correspondence with the empirical data
size Pi of a language i and the area inhabited by its and some conclusions on the degree of competition be-
speakers Ai . Our data set includes all living languages tween populations in different regions and the effect of
plus a set of about 700 recently extinct but listed in the the demographic pressure can be extracted.
most comprehensible database currently available: the
Ethnologue1 . We have first calculated the distribution
Size, Pi
of areas corresponding to all languages (see Figure 1) 0 2 4 6 8
10 10 10 10 10
and have observed that it is also compatible with a log- 1000
6
10
normal function. Further, the representation of the num-

Area, Ai
4
ber of speakers as a function of the area of a language 800 10
Frequency[Ln (variable)]

yields a dependence Ai ∝ Piz . The analysis of the previ- 10

2

600
ous quantities in five different geographic regions (Africa, 10
0

America, Asia, Europe, and Papua New Guinea) reveals

400
that, while the log-normal shape of the distributions for
areas and sizes, as well as the power-law dependence be- Language Area
200 Language Size
tween both quantities are maintained, the parameters of
the distributions (average size and dispersion) and the 0
0 4 8 12 16 20 24
exponent z are region-dependent. Ln (variable)
With these data at hand, we have devised a dynamical Figura 1. Distribution of language sizes and language areas
model for the change of language sizes and areas that re- on Earth. Both functions are compatible with a log-normal
lies on the demographic growth of the last ten centuries probability distribution. Inset: scatter plot of the area Ai and
and includes in an effective fashion the competition for the size Pi of each language i. The two variables are corre-
physical space. The model is defined by two rules speci- lated, indicating that the demographic growth conditions to
a large extent the area spanned by a language. The straight
fying the growth in language size and language area: 1.
line has slope 0.7.
At each time step (time is measured in years) the size of
a language is multiplied by αt , a value randomly drawn
for each language from a distribution Q(α): Pt+1 = αt Pt .
2. The change in area for that same language is set to
At+1 = γt At , where γt is chosen as follows. (2.i) If αt < 1, ∗
scmanrubia@[Link]
then γt = 1; ([Link]) if αt > 1 but at the same time there is †
Centro Atómico Bariloche and Instituto Balseiro, 8400 San
another, randomly chosen population whose growth fac- Carlos de Bariloche, Argentina
tor is larger than one, then γt = 1 again. ([Link]) When the 1
R. G. Gordon, ed., Ethnologue: Languages of the World,
size of a language grows and the second population has 15th edition (SIL International, Dallas, 2005). Online ver-
shrunk it is then possible for the population speaking the sion: [Link]
2
first language to increase the area it covers in an amount Zanette, D. H. Demographic growth and the distribution
directly related to the population growth: γt = g(αt ). of language sizes. Int. J. Mod. Phys. C 19, 237 (2008).

Junio de 2011, Barcelona Panel P–23

78 XVII Congreso de Fı́sica Estadı́stica FisEs11

A First Passage Time Analysis of Atomic-Resolution Simulations of the Ionic

Transport through a Bacterial Porin

Carles Calero∗† , Jordi Faraudo† , Marcel Aguilella-Arzo‡

†
Institut de Ciència dels Materials de Barcelona (ICMAB-CSIC), Campus de la UAB, E-08193 Bellaterra, Spain.
‡
Biophysics Group, Department of Physics, Universitat Jaume I, 12080 Castelló, Spain.

Understanding transport of charged solutes across the

cell membrane is a problem of paramount importance
in biophysics, since it is crucial to regulate many cell
functions1 . This task is undertaken by transmembrane
channels, so the characterization of the permeation of
charged particles, ions in particular, is essential. In ad-
dition, recently there has been a renewed interest to
comprehend transport through biological nanochannels
for possible applications in biotechnology2. Bacterial
porins are macromolecular proteins located at the outer
membrane that enable the diffusion of small molecules
through the lipid bilayer. As porins are well charac-
terized, both structurally and functionally, they repre- Figura 1. Flux of ions through the OmpF porin, as ob-
sent model systems to study transport through biological tained from molecular dynamics simulations. The inset shows
a diagram of the channel, where the directions of the external
nanochannels. In particular, the ionic transport proper-
electric field and of the flux of Cl− and K+ ions are repre-
ties of the OmpF porin, a transmembrane nanochannel
sented.
located at the outer membrane of Escherichia coli, have
been extensively studied since the determination of its
X-ray structure3 .
In this work, we study the transport of potassium and
chloride ions and of water through the OmpF porin5 .
Using the results of extensive all-atom molecular dynam- ∗
ics simulations of the system4 , we employ a first time 1
ccalero@[Link]
passage analysis to understand the transport of ions and B. Hille, Ion Channels of Excitable Membranes (Sinaure,
water through the channel in terms of a one dimensional Sunderland, 2001).
2
G. Maglia, A. J. Heron, W. L. Hwang, M. A. Holden, E.
biased diffusion model. We explore the applicability of
Mikhailova, Q. Li, S. Cheley and H. Bayley, Nature Nan-
such a description and extract the diffusion coefficients
otechnology 4, 437 (2009).
and effective forces characterizing the dynamics of the 3
Cowan, S. W., R. M. Garavito, J. N. Jansonius, J. A. Jenk-
particles in different regions of the channel. These re- ins, R. Karlsson, N. König, E. F. Pai, R. A. Pauptit, P. J.
sults can be used to describe the ionic transport through Rizkallah, J. P. Rosenbusch, G. Rummel and T. Schirmer,
the channel with the help of classical 1D models (such Structure 3, 1041 (1995).
as the Piosson-Nernst-Planck approach) with coefficients 4
J. Faraudo, C. Calero and M. Aguilella-Arzo, Biophys. J.
obtained taking into account all the complexity of the 99, 2107 (2010).
structure of the channel. From our analysis we recog- 5
C. Calero, J. Faraudo, M. Aguilella-Arzo, accepted
nize the appearance and quantify the magnitude of ef- in Phys. Rev. E (preprint at arXiv:1011.3952v1 [cond-
fective entropic forces, which are compared with existing [Link]]).
theories6 on the transport of particles through channels 6
R. Zwanzig, J. Phys, Chem. 96, 3926 (1992). D. Reguera
of varying cross section. and J. M. Rubı́, Phys. Rev. E 64, 061106 (2001).

Panel P–24 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 79

Solvation properties of mono- and di-valent cations in water from DFT molecular
dynamics simulations

Ausias-March Calvo†∗ , Elvira Guàrdia† , Marco Masia‡

†
Departament de Fı́sica i Enginyeria Nuclear,
Universitat Politécnica de Catalunya, Campus Nord, B4-B5
08034 Barcelona, Spain
‡
Dipartimento di Chimica, Università degli Studi di Sassari,
Sardinian Laboratory for Computational Materials Science
SLACS (INFM-CNR) and INSTM, Via Vienna 2, 07100 Sassari, Italy

Structural and dynamical properties of the solvation of

mono- and di-valent cations in water at ambient condi-
tions were studied using density functional theory based
molecular dynamics simulations. The simulated systems
were made of one single ion and 96 water molecules. The
ions studied include Li+ , Na+ , K+ , Mg2+ and Ca2+ . In
each case, the length of the cubic box was computed in
order to get a density ρ = 1g/cm3 . Periodic boundary
conditions were applied. The simulations were performed
using the Car-Parrinello scheme1 as implemented in the
CPMD package.2 The BLYP density functional3,4 was
used for the electronic structure calculations. The cut-
off for the wavefunctions was set to 80 Ry, the time step
was set to 4 a.u. and the fictitious electron mass was cho-
sen to be 400 a.m.u. Every five time steps, the Wannier
centers’ coordinates5,6 were computed (see Fig. 1). A
similar procedure was previously used in Ref. 7 to study
ion and molecular polarization of halides in water.

Figura 2. Ion-oxygen radial distribution functions (solid

lines) and running coordination numbers (dashed lines).

Figura 1. Electronic estructure of the water molecule. Bars

represent the oxygen-hydrogen bonds, purple spheres corre- ∗
[Link]@[Link]
spond to the location of the electronic pairs. 1
D. Marx and J. Hutter, in Modern methods and algorithms
of quantum chemistry, John von Neumann Institute for
Computing, Forschungszentrum Julich, pp.329–477, 2000.
2
CPMD version 3.11,Copyright IBM Corp. 1990-2006,
MPI für Festköorperforschung Stuttgart 1997-2001. For
downloads see [Link] (accessed Dec. 1,
Production runs of 15 ps in the microcanonical ensem- 2009).
3
ble followed NVT equilibration runs of 3 ps where the A. D. Becke, Phys. Rev. A 38, 3098 (1988).
4
temperature was set to 330 K; the initial configurations C. Lee, W. Yang, and R. G. Parr, Phys. Rev. B 37, 785
were generated with classical molecular dynamics simu- (1988).
5
lations of 200 ps. Fig. 2 shows the computed ion-oxygen N. Marzari and D. Vanderbilt, Phys. Rev. B 56, 12847
(1997)
radial distribution functions. A thorough analysis of the 6
P. L. Silvestrelli, and M. Parrinello, Phys. Rev. Lett. 82,
structure and dynamic properties of the first solvation
3308 (1999); erratum: Phys. Rev. Lett. 82, 5415 (1999).
shell molecules will be reported. 7
E. Guardia, I. Skarmoutsos, and M. Masia, J. Chem. The-
ory Comput. 5, 1449 (2009).

Junio de 2011, Barcelona Panel P–25

80 XVII Congreso de Fı́sica Estadı́stica FisEs11

Estudio del flujo sanguı́neo en una arteria estenótica

Marta Lage1 , Beatriz Rastrollo1, Alexis Cantizano1 , Mario Castro2

1
Departamento de Ingenierı́a Mecánica. Universidad Pontificia Comillas. Madrid, España.
2
Grupo Interdisciplinar de Sistemas Complejos (GISC). Universidad Pontificia Comillas. Madrid, España.

La Hemodinámica es la parte de la Biomecánica que mismo, por lo que se estudiarán la convergencia de

estudia el flujo sanguı́neo en el sistema circulatorio, la solución en situaciones biomédicamente realistas.
basándose en los principios fı́sicos de la dinámica de
fluidos1 . Estudia el movimiento de la sangre (campos de 4. Visualización y extracción de información:
velocidades y presiones) y las fuerzas que genera el flujo Por último, se analizarán los resultados del paso
en los elementos con los que interactúa (vasos sanguı́neos 3 tanto gráficamente (visualización del campo de
y corazón). presiones o del campo de velocidades en el interior
En este trabajo se simulará el flujo sanguı́neo en una de la arteria) ası́ como mediante el establecimien-
arteria con estenosis. La estenosis es un estrechamien- to de criterio de riesgo médico (relacionados, en su
to localizado de la arteria causado tı́picamente por gran mayorı́a, con magnitudes como el stress en
aterosclerosis2. La formación de placa aterosclerótica en las paredes de la arteria, la oscilación del stress o
arterias coronarias depende, entre otras variables, de la el tiempo de residencia del fluido en las distintas
hemodinámica del sistema arterial. partes de la geometrı́a del mismo).
Mediante el uso de software de fluidomecánica com- La validez del modelo numérico se evaluará comparan-
putacional (CFD, por sus siglas en inglés), este estu- do con resultados analı́ticos en situaciones sencillas (como
dio pretende proporcionar información relevante acerca el flujo de Womersley para un fluido newtoniano en un
de las variaciones de presión y distribución de tensiones conducto cilı́ndrico en régimen laminar oscilatorio) y con
en una arteria con estenosis, ayudando ası́ a entender, datos experimentales extraı́dos de estudios biomédicos5 .
predecir y solucionar este tipo de enfermedades vascu-
lares. Se utilizará la herramienta OpenFOAM, que es
un software de cálculo en 3D que puede simular y re-
solver numéricamente flujos complejos incluyendo reac-
ciones quı́micas, turbulencia y transmisión de calor, prob-
lemas estructurales o incluso electromagnetismo3.
El trabajo consta de cuatro etapas estándar4 :

1. Construcción del modelo: En esta etapa se de-

fine la geometrı́a del problema que se quiere estu-
diar. En nuestro caso, el flujo en una arteria es- Figura 1. Diagrama esquemático del flujo en las proximi-
tenótica. Por simplicidad, supondremos que la ar- dades de una estenosis.
teria es rı́gida (y por tanto que se pueden despreciar
los efectos de la elasticidad de la misma).

2. Generación del mallado: El software Open-

FOAM es un software que implementa distintos al- ∗
[Link]@[Link]
goritmos de resolución de las ecuaciones de Navier- 1
M. Thiriet, Biomechnics and Mechanics of Blood Flows
Stokes basados en volúmenes finitos. Por esta (vol I), (Springer Verlag, 2007).
razón, la construcción de un mallado adecuado es 2
F. Loth, P. F. Fischer, and H. S. Bassiouny, Ann. Rev.
crucial para la fiabilidad de los resultados. Fluid Mech. 40, 367 (2008).
3
OpenFOAM se puede descargar de la página web
3. Solución del problema: Como se ha menciona- [Link]
do, OpenFOAM integra el problema mediante la 4
C. A. Taylor and J. D. Humphrey, Comp. Meth. Appl.
técnica de volúmenes finitos. La estabilidad de es- Mech. Eng. 198, 3514 (2009).
5
tos esquemas depende del número de Reynolds del S. E. Logan, IEEE Trans. Biomed. Eng. 4, 327 (2007).

Panel P–26 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 81

Thermodynamics of feedback controlled systems

Francisco J. Cao∗
Departamento de Fı́sica Atómica, Molecular y Nuclear, Universidad Complutense de Madrid,
Avenida Complutense s/n, 28040 Madrid, Spain

We compute the entropy reduction in feedback con-

trolled systems due to the repeated operation of the
controller. This was the lacking ingredient to establish
the thermodynamics of these systems, and in particular
of Maxwell’s demons. We illustrate some of the conse-
quences of our general results by deriving the maximum
work that can be extracted from isothermal feedback con- 1−α
trolled systems. As a case example, we finally study a α 7
simple system that performs an isothermal information- α 6
fueled particle pumping. 5
4 L
3
1−α
2
1
Figura 1. Illustration of the Markovian particle pump with
n = 2 lattice sites between barriers. This is a simple exam-
ple of a feedback controlled system that extracts useful work
from the entropy reduction due to the information about the
system used by an external feedback controller.

∗
francao@fi[Link]
1
F. J. Cao, M. Feito, Thermodynamics of feedback con-
trolled systems, Phys. Rev. E 79, 041118 (2009).

Junio de 2011, Barcelona Panel P–27

82 XVII Congreso de Fı́sica Estadı́stica FisEs11

Scale invariance in marine population dynamics

José A. Capitán∗1 , Gustav W. Delius†2

1
Dept. Enginyeria Informàtica i Matemàtiques, Universitat Rovira i Virgili, 43007 Tarragona, Spain
2
Dept. of Mathematics, University of York, York, United Kingdom

(a) Predation (b) Reproduction

A special property of the open sea that has not been ex-
ploited yet from a theoretical point of view in the context +m ijk

of population dynamics is its approximate physical scale j i k k j i

invariance. The open sea looks similar at a wide range +m jik

of scales. Over many orders of magnitude there are no i j k k i j

physical features and no strong physical principles that +m jki

would single out a particular intermediate scale. This j k i i k j

kind of scale invariance is not present in terrestrial en- (c) Maintenance respiration (d) Intrinsic mortality
vironments, where not only the geographical constraints
set up a scale, but in addition the effects of gravity quick-
i i+1
ly become important for larger organisms.

i−1 i i

As for the marine ecosystem, the main observation-

al evidence for approximate scale invariance is that the
equilibrium size distribution of organisms in the open Figura 1. Individual stochastic processes involved in the
ocean is approximately given by a power law, valid over model that affect the number of organisms in different weight
brackets. Arrows indicate the movement of individuals be-
almost ten orders of magnitude. Several theoretical mod-
tween weight brackets as a consequence of each process.
els have been proposed in the past to derive this steady-
state size spectrum.1–3 In this contribution we interpret
the observed size spectrum as a consequence of the scale
invariance of the underlying population dynamics. Hence 0
we use this symmetry in the construction and solution ε = 0.9, μ = -8
of a size-structured dynamical population model. Start- ε = 1.1, μ = -10

ing from a Markov model encoding the basic process- -5

es of predation, reproduction, maintenance (respiration),
and intrinsic mortality (see FIG.1), we derive a partial 1
-10
λ(k)/u0a0

integro-differential equation describing the dependence of

abundance on weight and time. Our model represents an 0
extension of earlier models based on the McKendrick-von -15
Foerster equation.3 The model is scale-invariant provided -1
the rate functions of the elementary stochastic process-
-20
es fulfill precise scaling properties. We determine the -2
steady-state power-law solution, whose exponent is de- 0 1 2 3 4
termined by the relative scaling between the rates of the -25
0 10 20 30 40 50
density-dependent processes predation and the rates of k
the density-independent processes reproduction, mainte-
nance, and mortality.
Figura 2. Perturbations of the steady state can be stud-
ied using standard Fourier analysis. The inclusion of repro-
While the population dynamics leads to a power-law duction and maintenance processes stabilizes the steady state
steady-state solution, its dynamical stability is not nec- over the whole set of wavenumbers.
essarily ensured. The conditions under which this steady
state is an attractor are much less well understood. This
is an important issue, because the knowledge of what
makes marine ecosystems resilient or susceptible to ex- ∗
jcapitan@[Link]
ternal pressure has become extremely relevant at a time †
delius@[Link]
of overexploitation of marine resources. We study the 1
J. Camacho and R. V. Solé, Europhys. Lett. 55, 774
stability of the steady state against small perturbations (2001).
and find that inclusion of maintenance respiration and 2
E. Benoı̂t and M.-J. Rochet, J. Theor. Biol. 226, 9 (2004).
reproduction in the model has a strong stabilizing ef- 3
S. Datta, G. W. Delius, and R. Law, Bull. Math. Biol. 72,
fect (FIG.2). Furthermore, the steady state is unstable 1361 (2010).
against a change in the overall population density unless 4
J. A. Capitán and G. W. Delius, Phys. Rev. E 81, 061901
the reproduction rate exceeds a certain threshold.4 (2010).

Panel P–28 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 83

Harmonic vibration modifies the Turing pattern morphology from White-Spots to

Black-Spots

Jorge Carballido-Landeira∗ , Jacobo Guiu-Souto, Alberto P. Muñuzuri

Grupo de Fı́sica Non Lineal. Facultade de Fı́sica
Universidade de Santiago de Compostela
E-15782 Santiago de Compostela, Spain

We investigate the effects of a mesoscopic perturbation using topological functionals, which allow us to identi-
on the pattern formation. This perturbation is achieved fy the transition between the different Turing scenarios
by adding a vibrational potential term to the free Hamil- with quantitative measurements of digitized grey-scales
tonian of the system. We try to elucidate if this pertur- images.
bation is able to modify the dynamics of Turing patterns
on reaction-diffusion system. By using numerical simula-
∗
tions we observed two main effects induced by the forc- [Link]@[Link]
1
ing: changes from white to black spots towards stripe K. R. Mecke Phys. Rev. E 53 04794 (1996).
2
and labyrinthine configuration and moreover, modifica- J. Guiu-Souto, J. Carballido-Landeira, V. Pérez-Villar and
tions in the Turing wavelength, mostly observed in White A. P. Muñuzuri Phys. Rev. E 82 066209 (2010).
3
Spots with Hexagonal order. M. Okamoto and S. Kondo Nature Comunications 1 66,
The morphology of Turing patterns was analyzed by 1071 (2010).

Junio de 2011, Barcelona Panel P–29

84 XVII Congreso de Fı́sica Estadı́stica FisEs11

Detección automática de palabras clave en textos cortos

C. Carretero Campos∗ , P. Bernaola-Galván, A. V. Coronado y P. Carpena

Departamento de Fı́sica Aplicada II,Universidad de Málaga. 29071 Málaga

El problema de la detección automática de palabras medida se evalúa con un glosario elaborado manualmente
clave en textos literarios a partir de sus propiedades es- que contiene las palabras relevantes del texto.
tadı́sticas ha sido abordado de maneras muy variadas. Como una de las principales caracterı́sticas deseables
La primera aproximación propuesta por Luhn1 estaba en un detector de keywords es que funcione correcta-
basada en un análisis de la frecuencia de ocurrencia de mente cuando el texto a analizar es corto, consideraremos
las palabras en el texto. Sin embargo, métodos que sólo ese caso. La frecuencia de aparición de las palabras es
consideran la frecuencia con la que aparece cada palabra pequeña lo que complica la estimación estadı́stica. Sin
funcionan correctamente si se compara con una colección embargo, los textos cortos resultan ser de gran interés
de documentos de referencia (corpus), pero no son su- porque en él se encuadrarı́an, por ejemplo, los artı́culos
ficiente para un único documento. Si aleatorizamos el cientı́ficos y las páginas web. Tomamos como modelo
texto conservamos la frecuencia con la que aparece cada de texto corto el capı́tulo IV de “The Origin Of Species”
palabra, pero se destruye la información. De modo que para poder disponer de un glosario que nos permita com-
la relevancia de una palabra vendrá determinada no por parar la precisión de cada método. Definimos P r(n)
la palabra en sı́ misma, sino por cómo se distribuye a lo como el número de palabras del glosario entre las n
largo del texto. Esta es la hipótesis presentada en Ortuño primeras del ranking de relevancia y observamos en la
et al.2 , dónde se comprueba que existe una relación entre figura 1 como de las 50 primeras palabras que C0 de-
la cantidad de información que proporciona una palabra tecta como relevantes un 60% lo son, mientras que para
y su distribución espacial en el texto: las palabras rele- Enor (considerando los párrafos como partición) sólo un
vantes tienen una distribución inhomogénea y están con- 32%. Un análisis más profundo muestra que los resulta-
centradas en determinadas regiones del texto formando dos obtenidos para Enor dependen de la partición consid-
clusters, mientras que las irrelevantes están distribuidas erada y que los de C0 son tan buenos o mejores, tanto en
de manera aleatoria. Ası́ obtienen un método basado en un texto largo y sobre todo en un texto corto, y sin nece-
la distribución de las distancias entre apariciones sucesi- sitar previamente una elección arbitraria de la partición
vas de las palabras que detecta adecuadamente las pal- usada.
abras relevantes del texto, pero que presenta algunas de-
bilidades que pueden conllevar identificaciones incorrec-
tas. En Carpena et al.3 , se realiza una mejora significa- 50
tiva del método de Ortuño et al.2 , combinando la infor- C0
mación proporcionada por el clustering de una palabra 40 Enor
con la de su frecuencia. Sin embargo, estos métodos no
funcionan demasiado bien en textos cortos. 30
Pr(n)

En este trabajo, proponemos incluir la información

contenida en las fronteras del texto, que permiten de- 20
tectar un único cluster; y mejoramos la estimación es-
tadı́stica para palabras con frecuencia pequeña. Al mis- 10
mo tiempo, con el objetivo de evitar que el singular y el
plural de una palabra sean consideradas palabras distin- 0
tas, los identificamos implementando un algoritmo que 0 10 20 30 40 50
contemple las reglas de formación de plurales en inglés, n
asi como las principales excepciones. Estas mejoras se Figura 1. P r(n) frente a n para las medidas Enor y C0 .
implementan en la medida de relevancia C0 .
Por otro lado, también se habı́an propuesto otro tipo
de técnicas. Recientemente, Herrera and Pury4 hacen
uso de la entropı́a de Shannon para definir una medida ∗
ccarretero@[Link]
de relevancia(Enor ) basada en el contenido de informa- 1
H. P. Luhn, IBM J. Res. Devel. 2, (1958) 159
ción de la secuencia de ocurrencias de cada palabra y que 2
M. Ortuño, P. Carpena, P. Bernaola-Galván, E. Muñoz
necesita una partición previa para su evaluación. Usando and A. M. Somoza, Europhys. Lett. 57, (2002) 759
“The Origin Of Species” de Charles Darwin como corpus 3
P. Carpena, P. Bernaola-Galván, M. Hackenberg, A. V.
y los capı́tulos como partición natural, esta medida re- Coronado, and J. L. Oliver, Phys. Rev. E 79, (2009)
sultaba ser tan buena o mejor que otras publicadas pre- 035102(R)
4
viamente, que tomamos nosotros como referencia para J. P. Herrera and P. A. Pury, Eur. Phys. J. B 63, (2008)
evaluar la bondad de nuestras mejoras. La bondad de la 135

Panel P–30 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 85

Electrocinética DC de suspensiones de partı́culas concentradas

en medios electrolı́ticos generales

Félix Carrique1∗ , Emilio Ruiz-Reina2

1
Fı́sica Aplicada I, Universidad de Málaga, 29071 Málaga (Spain)
2
Fı́sica Aplicada II, Universidad de Málaga, 29071 Málaga (Spain)

En este trabajo se estudia la respuesta de una sus- hidrodinámicas entre las partı́culas en estos medios con-
pensión coloidal concentrada de partı́culas esféricas en centrados. Hemos aplicado el modelo al estudio de la
un medio electrolı́tico genérico en presencia de un cam- movilidad electroforética DC de las partı́culas y a la con-
po eléctrico estático. Se estudia en particular la veloci- ductividad eléctrica DC para diferentes concentraciones
dad estacionaria de las partı́culas por unidad de campo de partı́culas, densidades de carga superficial de las mis-
eléctrico aplicado, o movilidad electroforética. Nuestro mas y concentración del electrolito añadido. Se puede
modelo tiene en cuenta: 1) los contraiones liberados por concluir que la presencia de una concentración modera-
las partı́culas cuando se cargan en solución; 2) los iones da de sal apantalla la mayorı́a de los efectos asociados
H+ y OH− de la disociación del agua; 3) los iones pro- con los puntos 1 a 3, cruciales para medios desionizados,
ducidos por la contaminación del CO2 atmosférico dis- como cabrı́a esperar. Sin embargo, si la concentración
uelto en el medio; 4) los iones de un electrolito añadido de sal es baja se observa que: i) la contaminación por
externamente a la suspensión. El modelo presentado en CO2 juega un papel excepcionalmente relevante para ba-
este trabajo está basado en una nueva descripción realista jas concentraciones de partı́culas en la suspensión; ii) los
de la doble capa eléctrica de equilibrio que rodea a las contraiones añadidos por las partı́culas tiene un papel
partı́culas, recientemente desarrollada por los autores1 crucial si la concentración de partı́culas es alta, apan-
ası́ como en modelos electrocinéticos DC2,3 que incorpo- tallando progresivamente los efectos de la sal. En el fu-
ran los puntos anteriores 1 a 3. Aquı́ estamos intere- turo próximo se extenderá el estudio a la electrocinética
sados en los efectos asociados con la presencia de una AC para poder cuantificar en cada rango de frecuencias
sal externa añadida al sistema (punto 4). Se conoce que del campo eléctrico el peso de las diferentes contribu-
para suspensiones desionizadas realistas, los efectos so- ciones a la respuesta electrocinética.
bre las propiedades electrocinéticas de las suspensiones
que incluyen las consideraciones de los puntos 1 a 3 son
∗
enormes. De aquı́ nuestro interés en extender estos es- carrique@[Link]
1
tudios a medios salinos, que incluyen a los desionizados E. Ruiz-Reina and F. Carrique, J. Phys. Chem. B 112,
como caso lı́mite cuando la sal externa tiende a cero. El 11960 (2008).
2
modelo está basado en la estadı́stica clásica de Boltz- F. Carrique and E. Ruiz-Reina, J. Phys. Chem. B 113,
mann, las apropiadas reacciones quı́micas, las ecuaciones 8613 (2009).
3
electrocinéticas estándar y la aproximación de modelo F. Carrique and E. Ruiz-Reina, J. Phys. Chem. B 113,
10261 (2009).
de celda para tener en cuenta las interacciones electro-

Junio de 2011, Barcelona Panel P–31

86 XVII Congreso de Fı́sica Estadı́stica FisEs11

Self-similarity and chaos in coupled ratchets. Unfolding the complexity

of cooperative Brownian motors

Carles Blanch-Mercader, Javier G. Orlandi, Jaume Casademunt

Departament d’Estructura i Constituents de la Matèria
Universitat de Barcelona
08028 Barcelona

Brownian ratchets may be considered as a paradigm of complex structure that includes the so-called big bang
statistical physics where thermal noise rectification under bifurcations3 . Remarkably, while the two involved maps
nonequilibrium conditions leads to directed motion and cannot produce chaos by themselves, the stochastic com-
power generation. Some molecular motors in biology, bination of the two is capable to generate chaos. In this
such as monomeric kinesin, are known to exploit directly scenario, although the dynamics is stochastic, the phase
this noise-based mechanism. Recently it has been shown space of the system can be hierarchically organized in a
that one of the reasons to adopt such mechanism in the tree-like structure which is generated by a chaotic map.
biological context might be that the ratchet structure is In summary, we show that underlying an apparently
particularly fit to cooperative action. In particular, it has simple mechanism of cooperativity of coupled ratchets
been proved that motors are capable to spontaneously there is a complex mathematical structure. Combining
associate and dramatically enhance their collective effi- graph theory and concepts from dynamical systems we
ciency, in particular when noise is weak1,2 . Remarkably identify a variety of complex dynamical behaviour, in-
a motor cluster is capable to produce a finite power in cluding a high sensitivity response to the external force,
the limit of vanishing noise intensity, when individual self-similar behaviour, and the signature of deterministic
motors would be powerless. This is possible whenever chaos in the inner structure of a stochastic system.
two main conditions are satisfied: (i) the external force
is unequally loaded to the participant motors, and (ii)
the ratchet states change independently for each motor.
These conditions are naturally satisfied in the biological
context of intracellular traffic of soft cargoes and affect
crucially the collective behaviour of motors. In order to
isolate and gain understanding on the underlying deter-
ministic mechanism of motor cooperativity, we study the
weak noise limit of the two-motor problem.
We solve exactly the problem of two interacting mo-
tors when an external force is applied to the first one,
in the limit of vanishing noise. The motors switch states
between a fully asymmetric sawtooth potential and a flat
potencial. We show that this situation is particularly fit
to the collective improvement of the motor performance,
and because of its simplicity it can be considered as a
paradigm of cooperativity. We show that the problem
can be mapped into a random walk on a complex di-
rected graph. The topological structure of the graph de- Figura 1. Fractal velocity-force curve. We plot the mean
pends sensitively on physical parameters, in particular as velocity of a motor pair normalized by the velocity of one
the external force is varied. Consequently the velocity- motor at zero load vs the force normalized to the stall value.
force curve of the motor pair becomes fractal. Using The curve in exact for 1/3 ≤ f ≤ 1. For 0 < f < 1/3 it is
graph theory techniques we obtain general properties of continuous in a subset of irrationals and the set of discontinu-
the dynamics of the system. We show that the graphs ities is infinite with zero measure. The values on this region
may be finite or infinite and we are able to solve exactly are approximated by Monte Carlo sampling.
the random-walk statistics in most regions of parameter
space. We can thus determine the exact velocity-force
curves in most regimes, exhibiting fractal properties.
The sampling of the graph that defines the joint dy-
namics of the motor pair can in turn be seen as a ran- 1
J. Brugués, J. Casademunt, Phys. Rev. Lett. 102, 118104
dom alternation of two deterministic evolutions. The re- (2009)
spective Poincaré sections of those define two circle maps 2
J. G. Orlandi, C. Blanch-Mercader, J. Brugués, J. Casade-
which are piece-wise linear. The multi-parametric bifur- munt, Phys. Rev. E 82, 061903 (2010)
3
cation structure of those maps is analyzed, showing a V. Avrutin, M. Schanz, Nonlinearity 19, 531-552 (2006)

Panel P–32 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 87

Laminar oscillatory flow of Maxwell and Oldroyd–B fluids:

Theoretical analysis and experiments

L. Casanellas∗, J. Ortı́n†
Departament d’Estructura i Constituents de la Matèria
C. Martı́ Franquès 1, Facultat de Fı́sica, Universitat de Barcelona
08028 Barcelona

Wall-bounded oscillatory flows of newtonian and com- drical layers. However, as we add an increasing newto-
plex fluids are present in many different practical situa- nian solvent contribution to the model, measured by the
tions. Pulsating flows are of particular interest in physi- viscosity ratio X = ηs /η, the resonances progressively
ology in connection with the respiratory and circulatory disappear and the magnitude of the velocity and shear
systems of living beings, in industrial processes such as rate are drastically reduced (Fig. 1)2 .
fluid pumping, secondary oil recovery and filtration, in We compare our theoretical predictions to experimen-
acoustics, and in the rheological characterization of fluid tal results that we have obtained for oscillatory flow in a
properties. cylindrical tube with large aspect ratio. The experimen-
In this work we revisit the generic problem of wall– tal velocity profiles are measured for a wormlike micel-
bounded oscillatory flow of a viscoelastic liquid. We focus lar solution, CPyCl-NaSal [100:60], using a time resolved
on the periodic steady state of laminar flow, at small Re. PIV technique. At low amplitudes and frequencies of
In this approximation the governing equations are lin- the driving oscillation the profiles follow the predicted
ear and therefore accessible to analytical treatment. We trends. However, at high velocity amplitudes (high shear
consider as constitutive equations the upper convected rates) the experimental profiles deviate from the theoret-
Maxwell model (UCM) and the more general Oldroyd–B ical predictions, as a result of the non–linearities of the
model, that includes a newtonian solvent contribution. fluid rheology in this regime.
We first consider the fluid motion induced by the syn-
chronous oscillation of two parallel infinite walls, each
giving rise to a traveling transverse shaer wave. The 100
a.
X =0
flow is characterized by the damping length of the ampli- X = 10−3
ṽ(r = 0)

tude of oscillation of the shear waves, x0 , and their wave- 50

X = 10−2
X =1
length, λ0 /(2π). For a Newtonian fluid these two lengths
are equal and coincide with the thickness of the bound- 0
2π
ary layer, leading to overdamped shear waves. However, b.
for viscoelastic fluids x0 > λ0 /(2π) and underdamped 3 π /2
Φ(r = 0)

shear waves can propagate effectively before they are π

attenuated1 . In ‘narrow’ systems (small setup dimen- π /2

sions compared to x0 ) the viscoelastic shear waves extend 0
0 2 4 6 8 10 12 14 16
through the whole system and superpose themselves orig- a/(λ0 /(2π))
inating an interference pattern inside the fluid domain. Figura 1. Theoretical prediction of the dimensionless veloc-
This leads to a resonant behaviour with a huge increase ity magnitude at the tube axis (a) and time phase lag between
of the velocity amplitude at particular frequencies. Com- the fluid velocity at the tube axis and the pressure gradient
puting the phase lag between the velocity at the center of (b) vs. the dimensionless ratio a/(λ0 /(2π)), for an Oldroy-
the geometry and the acceleration of the moving plates d-B fluid. Different lines correspond to different values of the
we see that they move in–phase at these frequencies. viscosity ratio X.

A similar behaviour is obtained in the oscillatory flow

within an infinitely long straight cylinder of radius a, in-
duced by the oscillatory motion of a bottom piston and a
top endwall. We derive the expression for the vertical ve- ∗
laurac@[Link]
locity along the radial coordinate inside the tube in terms †
ortin@[Link]
of the characteristic lengths of the corresponding shear 1
Mitran, S.M., Forest, M.G., Yao, L., Lindley, B. & Hill,
waves. For Maxwell fluids the resonances are prominent D.B. 2008. J. Non-Newtonian Fluid Mech. 154, 120–135.
2
in ‘narrow’ systems and the velocity profiles show a com- Andrienko, Y.A., Siginer, D.A. & Yanovsky, Y.G. 2000.
plex inflectional structure, with multiple concentric cylin- Int. J. Non–Linear Mech. 35, 95–102.

Junio de 2011, Barcelona Panel P–33

88 XVII Congreso de Fı́sica Estadı́stica FisEs11

Control of spiral wave dynamics by low energy stimuli

Jorge Castro1 , Matias Rafti2 , Flavio H. Fenton3 , Jorge Carballido-Landeira1, Alberto P. Muñuzuri1∗
1
Group of Non Linear Physics Fac. de Fisicas. Universidade de Santiago de Compostela. Campus Sur. 15782 Santiago de
Compostela. Spain. 2 INIFTA. Fac. Cs. Exactas. Dpto. de Quı́mica.— Univ. Nac. de La Plata. Argentina. 3 Dept. of
Biomedical Sciences. Cornell University, Ithaca, NY, USA

Atrial fibrillation (AF) is the most common sustained We will analyze the same problem within the context of
cardiac arrhythmia worldwide affecting 2.2 million peo- chemical pattern forming chemical reactions (Belousov-
ple in the United States alone. Complications associated Zhabotinsky reaction). Effectively, BZ reaction is con-
with chronic AF include increased risk for both throm- sidered to be a paradigm for pattern formation research
boembolism and stroke. Left untreated, paroxysmal AF as the mechanisms here studied can be easily extrapolat-
often progresses to permanent AF, which is resistant to ed to more complex systems such as the cardiac tissue.
therapy. Although underlying anatomic or pathophysi- Spiral waves are believe to be responsible of Atrial Fib-
ological factors may fuel this progression, AF itself may rilation in the heart and can be easily observed in BZ re-
lead to its own perpetuation through electric, structural, action. In this system, we apply low-energy stimuli and
and metabolic remodeling of atrial tissue. The realiza- demonstrate the effectiveness of the method. The pres-
tion that AF begets AF has led to management strategies ence of conductivity discontinuities in the system will also
that are designed to avoid the progression of AF by re- be modeled and the results agree with the experiments
ducing the frequency and duration of AF episodes [1]. in cardiology.
Among these strategies, a new method to terminate fast Experimental results and numerical simulations will
atrial tachycardias and AF has been recently developed present clearly the deep mechanism underlying.
within the context of animal experimentation [2]. In this
method, electrodes located at a small distance from the
∗
heart deliver a train of low-voltage shocks at a rapid rate. uscfmapm@[Link]
1
The idea underlying is that during the low-energy shocks, Waktare JE. Cardiology patient page: atrial fibrilla-
small intrinsic conductivity discontinuities behave as in- tion. Circulation. 2002;106:14-16. Heart disease and stroke
ternal ”virtual”electrodes that actually serve as activa- statistics: 2007 update. Circulation. 2007;115: e69-e171.
tion sites (or secondary sources) if the field strength de- Nattel S, Opie LH. Controversies in atrial fibrillation.
polarizes the tissue beyond the excitation threshold. This Lancet. 2006;367: 262-272. Wijffels M, Kirchhof C, Dor-
method presents many advantages comparing with pre- land R, Allessie MA. Atrial fibrillation begets atrial fibril-
lation: a study in awake chronically instrumented goats.
vious methods.
Circulation. 1995;92:1954-1968.
Although the method has been successfully tested in 2
Fenton FH, Luther S, Otani NF, Cherry EM, Pumir A,
laboratory animals, a deep understanding of the mech- Bodenschatz E, Krinsky V, Gilmour RF Jr. ermination of
anisms underlying is still to be done. In such a sense, atrial fibrillation using pulsed low-energy far field stimu-
we will present in this contribution results aiming to un- lation Circulation 120, 467-476: 2009
3
derstand the previous mechanism in a completely dif- Matias Rafti, Flavio H. Fenton and Alberto P. Muũzuri,
ferent system although equivalent in many senses [3]. in preparation.

Panel P–34 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 89

First steps in the study of unconventional magnetic fluids

Juan J. Cerdà∗ , Pedro A. Sánchez, Tomás Sintes

Instituto de Fı́sica Interdisciplinar y Sistemas Complejos, IFISC (CSIC-UIB). Universitat de les Illes Balears. E-07122
Palma de Mallorca, Spain.

Unconventional magnetic fluids are suspensions in acting surfaces. Further insight about the influence of
which magnetic particles have been replaced by more external fields on the phase behavior of such magnetic
complex magnetic micro and nanostructures like for in- systems will be also provided.
stance magnetic filaments1 and shifted-dipole particles2 .
Those fluids are expected to exhibit magnetorheologial
∗
and tribological properties very different from ordinary joan@ifi[Link]
1
magnetic fluids. A first step towards their characteriza- Pedro A. Sánchez, Juan J. Cerdà, V. Balleneg-
tion is the study of very diluted suspensions in which such ger, T. Sintes, O. Piro, C. Holm, Soft Matter,
unconventional magnetic units are present. Our work in- doi:10.1039/C0SM00772B, (2011).
2
cludes the study of the phase behavior of such diluted Sofia Kantorovich, R. Weeber, Juan J. Cerdà, C. Holm,
suspensions in bulk as well as near neutral and inter- J. Magnetism Magnetic Materials,
doi:10.1016/[Link].2010.11.019, (2011);

Junio de 2011, Barcelona Panel P–35

90 XVII Congreso de Fı́sica Estadı́stica FisEs11

Nature of Waterlike Liquid-Liquid Criticality

Claudio A. Cerdeiriña, Pablo G. Debenedetti

Department of Chemical and Biological Engineering
Princeton University
08544 Princeton, New Jersey, USA

The study of phase transitions for pure substances in the isobaric thermal expansivity αp diverges to −∞ at
which two liquids of different density coexist in equilib- criticality, in accord with available information. Corre-
rium acquired great interest in 1992 with water’s sec- spondingly, ‘soft-core-like’ behavior (viz., dpσ /dT ¿0 and
ond critical point hypothesis by Stanley and coworkers1. αp → +∞ as T→Tc ) is obtained when low-density and
Such waterlike liquid-liquid phase transitions have been high-entropy states are energetically favorable.
observed experimentally for other fluids2–4 and studied In summary, we will show that (Ising-like) CCL mod-
via simulations of water models and of fluids character- els provide insights into liquid-liquid critical phenom-
ized by isotropic core-softened intermolecular potentials ena in pure substances. Some variants are consistent
(see, e.g., Refs. 6 and 7). At this stage, it seems appro- with what has been conjectured for water’s liquid-liquid
priate to investigate the nature of waterlike liquid-liquid critical point, others with what core-softened models
criticality. have evidenced. Implications of CCL’s for isostruc-
One may first inquire about the universality class of tural solid-solid transitions in Cs and certain rare earth
this sort of continuous phase transitions. This issue has metals13 as well as for crystalline systems with short-
an obvious intrinsic interest, but it is also relevant with range interactions14,15 will be briefly outlined.
a view to applying finite-size scaling techniques to sim-
ulation data8 . Kurita and Tanaka5 and Sciortino and ∗
coworkers9 have asserted that, like for the liquid-gas case, 1
calvarez@[Link]
waterlike liquid-liquid criticality should be of Ising char- P.H. Poole, F. Sciortino, U. Essmann, and H.E. Stanley,
acter. On the other hand, Brovchenko et al.10 have point- Nature 360, 324 (1992).
2
Y. Katayama, T. Mizutani, W. Utsumi, O. Shimomura, M.
ed out that it could belong to the universality class of
Yamakata, and K.I. Funakoshi, Nature 403, 170 (2000).
the random-field Ising model. To the best of our knowl- 3
G. Monaco, S. Falconi, W.A. Crichton, and M. Mezouar,
edge, there is neither strong experimental evidence nor
Phys. Rev. Lett. 90, 255701 (2003).
a detailed enough theoretical analysis on this particular 4
H. Tanaka, R. Kurita, and H. Mataki, Phys. Rev. Lett.
question. 92, 025701 (2004).
Here we tackle the problem by studying the properties 5
R. Kurita and H. Tanaka, Science 306, 845 (2004).
6
of a simple cell model. Consider the space divided in P.G. Debenedetti, J. Phys.: Condens. Matter 15, R1669
cells which contain one particle and have two accessible (2003).
7
volumes, say, v+ and v− . Particles in nearest-neighbor S.V. Buldyrev, G. Malescio, C.A. Angell, N. Giovambat-
cells interact via a discretized energy 0 , which is supple- tista, S. Prestipino, F. Saija, H.E. Stanley, and L. Xu, J.
mented by an additional one δ whenever two adjacent Phys.: Condens. Matter 21, 504106 (2009).
8
cells are in the ‘+’ state. Such compressible cell liquid Y. Liu, A.Z. Panagiotopoulos, and P.G. Debenedetti, J.
(CCL) model can be mapped onto the Ising model, and is Chem. Phys. 131, 104508 (2009).
9
in certain respects related to a previously reported water C.W. Hsu, J. Largo, F. Sciortino, and F.W. Starr, Proc.
model11 and in others to the so-called compressible cell 10
Natl. Acad. Sci. U.S.A. 105, 13711 (2008).
gases for asymmetric liquid-gas criticality12 . I. Brovchenko, A. Geiger, and A. Oleinikova, J. Chem.
Phys. 123, 044515 (2005).
Application to water’s second critical point entails in- 11
G. Franzese and H.E. Stanley, J. Phys.: Condens. Matter
corporating entropic effects ascribed to the geometric se- 19, 205126 (2007).
lectivity of hydrogen bonding. Such a task is accom- 12
C.A. Cerdeiriña, G. Orkoulas, and M.E. Fisher, in prepa-
plished by suitably imposing constraints on the free vol- ration.
ume that a particle can explore in its cell. Specifical- 13
J.M. Kincaid, G. Stell, and E. Goldmark, J. Chem. Phys.
ly, when local low-density and low-entropy states are 65, 2172 (1976).
energetically favorable (that is, when low-temperature 14
P. Bolhuis and D. Frenkel, Phys. Rev. Lett. 72, 2211
water’s essential microscopic features are implemented), (1994).
15
one finds that the slope of the coexistence line in the C.F. Tejero, A. Daanoun, H.N.W. Lekkerkerker, and M.
temperature-pressure plane dpσ /dT is negative and that Baus, Phys. Rev. Lett. 73, 752 (1994).

Panel P–36 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 91

Coordination of mitotic axes guides morphogenesis of epithelial cysts

Benedetta Cerruti1∗ , Alberto Puliafito1 , Annette M. Shewan2 , Keith E. Mostov2 , Guido Serini1 , Antonio Celani3 ,
and Andrea Gamba1,4
1
Institute for Cancer Research and Treatment, Str. Prov. 142 km 3.95, 10060 Candiolo, Torino, Italy
2
Department of Anatomy, University of California, San Francisco, California, United States of America
3
Institut Pasteur, Genomes and Genetics Department, Unit “Physics of Biological Systems”,
Centre National de la Recherche Scientifique Unité de Recherche Associée 2171, F-75015 Paris, France
1,4
Politecnico di Torino, Corso Duca degli Abruzzi 24, 10129 Torino, Italy

Almost every eucariothic cell is polarized, and the for- surface tensions corresponding to different cell-cell and
mation of tissues and organs requires the coordination cell-matrix adhesion strengths, and there are two char-
of this polarization in space and time1 . Typical exam- acteristic timescales, the mitotic time and the relaxation
ples of polarized tissues are epithelia, the 60% of mam- time between two mitotic events. Depending on the ra-
malian cell types being of epithelial origins. Epithelia tio between these two timescales there exist two possible
and epithelial-derived tissues serve to divide the organ- regimes: an equilibrium regime, in which cells can re-
ism into compartments, and are often characterized by lax to a global energy minimum before a new division
the presence of a central lumen or luminal network, as occurs, and an out-of-equilibrium regime. Topological
observed for instance in kidney, lung, blood vessels and analysis of our experiments on wild-type Madin-Darby
mammary glands. Lumens can serve to isolate specific canine kidney cells in 3d cultures shows that the physio-
functions, like in digestion, or to allow the movement of logical regime corresponds to the out-of-equilibrium case.
fluids, cells or gases, and lumen dysfunctions are at the In this regime, a mechanism of spatiotemporal correla-
basis of very common and often fatal human diseases, like tion of mitotic spindle axes is mandatory to achieve the
polycystic kidney diseases, hypertension or many epithe- formation of a central hollow lumen (fig.1a), while its
lial cancers. Despite the apparent differences between the absence leads to the formation of the pathological multi-
various epithelial tissues, the idea is arising that their for- lumen phenotype (fig.1b).
mation can be led by general principles, and thus there
exist common features underlying their dysfunctions2 .
In order to form multicellular lumens, cells must coor-
dinate the orientation of their apical surfaces, requiring
interaction of every cell with the other cells and with
the extra-cellular matrix. Hence a first general princi-
ple must involve cell-matrix and cell-cell recognition: po-
larity proteins are key regulators of apical surface and
lumen orientation3 , while sensing neighboring cells oc-
curs with a multitude of adhesion receptors. A second
general principle must involve apical-basal polarization, Figura 1. Simulated cyst topology: (a) monolumen and (b)
spatiotemporally coordinated with neighboring cells, and multilumen configurations.
the generation of a new luminal space: mitotic spindle
axes tend to align parallel to the lumen surface4 .
Inspired by this last phenomenologic observation, we
investigate if the coordination between successive mitoses ∗
[Link]@[Link]
is necessary and sufficient for the formation of a central 1
L.E. O’Brien, M.M. Zegers, and K.E. Mostov, Nature Rev.
and spherical luminal structure, as observed in physio- Mol. Cell Biol. 3, 531 (2002).
logical cysts, and hence if the absence of this correlation 2
A. Datta, D.M. Bryant, and K.E. Mostov, Curr. Biol. 21,
can lead to a multiluminal structure, observed in patho- R126-R136 (2001)
logical conditions like cancer and experimentally repro- 3
L.E. O’Brien, T.S. Jou, A.L. Pollack, Q. Zhang, S.H.
duced in cells where the spindle orientation machinery is Hansen, P. Yurchenco, and K.E. Mostov, Nat. Cell Biol.
disrupted by targeted mutations4 . We develop a simple 3, 831-838 (2001)
4
lattice model that simulates cystogenesis from its ear- Z. Zheng, H. Zhu, Q. Wan, J. Liu, Z. Xiao, D.P. Siderovs-
liest stages. In the model, cell evolution is driven by ki, and Q. Du, J. Cell Biol. 189, 275 (2010)

Junio de 2011, Barcelona Panel P–37

92 XVII Congreso de Fı́sica Estadı́stica FisEs11

Capillary rise between parallel plates:

Velocity dependence of the capillary pressure

Xavier Clotet∗ , Jordi Ortı́n

Dep. d’Estructura i Constituents de la Matèria, Fac. de Fı́sica, Universitat de Barcelona
C./ Martı́ i Franquès 1, 08028 Barcelona

Capillary rise is a process of spontaneous imbibition The first aim of the study is to check whether a dy-
in the presence of gravity. That is, the displacement at namic contact angle approach applies also to our models
constant external pressure of one fluid by a second immis- of porous medium. The second one is to measure the ex-
cible invading fluid that preferentially wets the medium, ponent x in the two models (with and without capillary
in a cavity under the effect of gravity. The dynamics pressure fluctuations) and investigate possible differences
of capillary rise was first studied in depth by Green and between them.
Ampt in 19111 and by Lucas and Washburn a few years The work consists of systematic experiments carried
later2,3 . out at different conditions of cell inclination (ψ) and ex-
The general equation for the pressure balance at the ternally applied pressure difference (P0 ). The analysis
average interface height for a fluid invading either capil- of the data is done by means of Eq.(1), neglecting the
lary tube of radius b/2 or a Hele-Shaw (H-S) cell of gap inertia term and replacing the capillarity contribution by
thickness b is:  
d
 2γ cos θ μ 2γ
cos θD = α + β(ḣ)x , (2)
ρ hḣ = − ρgh sin ψ − hḣ + P0 , b
dt b
      κ  
capillarity gravity viscous losses where α accounts for the static contribution and β for the
(1) dynamic one. Both coefficients together with the expo-
nent x are obtained from fitting the experimental data.
where the LHS accounts for the variation of the momen-
tum (inertial term) and the RHS is the sum of various
pressure contributions. ρ is the density of the fluid and
160
h(t) stands for the mean height of the front measured Experimental
140 Analytical
along the cell as a function of time. γ is the surface
120
tension of the air-fluid interface, θ is the apparent con- 150
100
h (mm)

tact angle at the solid-liquid-vapor contact line, g is the

80 100
acceleration of gravity and the angle ψ measures the in-
60
clination of the cell with respect to the horizontal. μ is 50
40
the dynamic viscosity of the fluid and κ the permeabili- 20 0
ty of the medium. P0 is the externally applied pressure 0
0 1000 2000 3000 4000 5000
0 500 1000 1500 2000 2500 3000 3500 4000
difference. t (s)
We have noticed that for capillary rise experiments Figura 1. Mean height of the front versus time. Thick line:
in vertical cylindrical tubes there is systematic devia- experimental data for an experiment in a H-S cell without
tion from the analytical or numerical curves proposed perturbations in the capillary pressure (κ = 0.018 mm2 )(main
for h(t) when considering a constant contact angle (Ex- plot) and with perturbations (κ = 0.013 mm2 )(inset), ψ = 5◦
tended Solution)4 . This deviation is partially corrected and P0 = 49 Pa. The fluid properties are μ = 49.9 mPa·s,
if a dynamic contact angle is assumed, cos θD ∝ β(ḣ)x , ρ = 998 kg/m3 and σ = 20.7 mN/m. Thin line: analytical
with x in the range5 0.33 ≤ x ≤ 1. solution given by the Extended Solution4 . Notice the devia-
In the present work we study these deviations for the tion of the analytical curve from the experimental results in
both cases.
displacement of air by a perfectly wetting fluid (silicone
oil) in laboratory models of porous media (Fig. 1). First,
the displacement of the fluid takes place in a rectangu-
lar Hele-Shaw cell with uniform gap thickness. Second,
the displacement takes place in a Hele-Shaw cell in which ∗
xclotet@[Link]
the gap spacing takes two possible values randomly dis- 1
W.H. Green and G.A. Ampt, J. of Agric. Sci. IV, 1 (1911).
tributed in space, mimicking an open fracture. In this 2
R. Lucas, Colloid & Polymer Science 23, 15 (1918).
medium the capillary pressure fluctuates along the air- 3
E.W. Washburn, Phys. Rev. 17, 273 (1921).
fluid interface. In both cases the flow is induced by an 4
N. Fries and M. Dreyer, J. Colloid Interface Sci. 320, 259
imposed externally pressure difference P0 , and the cell (2008).
5
is inclined so that a gravitational field g sin ψ acts as a A. Hamraoui et al., J. Colloid Interface Sci. 226, 199
volume stabilizing force. (2000), and references therein.

Panel P–38 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 93

Dynamics of passively phased fiber laser arrays

A. Jacobo1,2, P. Colet1 , E.J. Bochove3, Y. Braiman4 , A.B. Aceves5 , R. Deiterding6 , C.A. Miller7 , C. Rhodes8 , S.A.
Shakir9
1. IFISC, Instituto de Fı́sica Interdisciplinar y Sistemas Complejos, CSIC-UIB, 07122 Palma de Mallorca, Spain
2. Max Planck Institute for the Physics of Complex Systems, Nöthnitzer Strae 38, 01187 Dresden, Germany
¯
3. Directed Energy Directorate, Air Force Research Laboratory, Kirtland Air Force Base, New Mexico 87117, USA.
4. Center for Engineering Science Advanced Research Oak Ridge National Laboratory, Tennessee 37831, USA
5. Department of Mathematics, Southern Methodist University, Dallas, Texas 75275, USA
6. Department of Mechanical, Aerospace, and Biomedical Engineering, University of Tennessee, Knoxville, TN 37996, USA
7. Department of Mathematics, University of Tennessee, Knoxville, TN 37996, USA
8. Libration Systems Management, Inc., Albuquerque, New Mexico 87106, USA
9. TASC, Inc., Albuquerque, New Mexico 87106, USA

The goal of scaling fiber lasers to high power levels for where Knm are the coefficients of the coupling matrix,
applications has created an interest in coherently com- Lm is the length of fiber m and τnm are the delays in the
bine the beams of a number of fibers. However, success feedback loop.
in accurately controlling the frequencies and phases of By linear stability analysis and numerical simulations
the fields by either passive, active or hybrid means has we show that, for one fiber, the system has a stable CW
until now been limited. Studies of the dynamics of pas- regime above threshold, for low pumping powers. We al-
sively phased fiber arrays have been published, but the so show that the system can become unstable for high
understanding of their instabilities is limited. We consid- pump intensities first exhibiting pulsating behavior and
er a passively phased array of N rare earth doped fibers then becoming chaotic. For two fibers we show that phase
which are coupled through an external cavity. Each fiber locking in a stable CW regime is possible, although this
is independently optically pumped. The propagation of regime becomes unstable when the pump is increased.
the signal electrical field and the pump intensity within Instabilities appear for lower powers as the difference in
each fiber can be described by: length between the fibers is increased. Examples of these
 
g  regimes are shown in Fig. 1
∂En  i  ∂En
+ ik0 n2 |En |2 En (1)
n
+ β0 − g n = Γs
∂z 2 ∂t 2
∂Ipn ∂Ipn
+ βp = −Γp αn (2)
∂z ∂t
∂N2n αn |En |2 N2n
= − gn − (3)
∂t p s τ
gn = σe (Ω0 )N2n − σa (Ω0 )N1n (4)
αn = (σa (ωp )N1n − σe (ωp )N2n ) Ipn (5)
Where En (z, t) is the complex slowly varying envelope
of the signal electric field in the fiber n with respect
to the central frequency Ω0 , and propagation constant
β0 . Ipn (z, t) is the pump intensity at frequency ωp . The
primes denote derivatives with respect to the frequency
at ω = Ω0 or ω = ωp . β0 is the signal inverse group ve-
locity while βp is the pump inverse group velocity. σa and
σe are the effective emission and absorption cross sections
respectively. s and p are the energies of the photons at
the central and pump frequencies. Γs and Γp are the
signal-core and pump-core overlap factors. Finally N1n
and N2n are the lower and upper level populations of the
medium (with N1n + N2n = N0n ) and n2 is the nonlin-
ear refractive index. The above equations are completed
with the boundary condition associated to the external
cavity coupling

N
En (0, t) = Knm Em (Lm , t − τnm )ei(β0 Lm +Ω0 τnm ) (6) Figura 1. Time traces for two fibers. From top to bottom:
m=1 CW emission, pulsed behavior, chaotic behavior.

Junio de 2011, Barcelona Panel P–39

94 XVII Congreso de Fı́sica Estadı́stica FisEs11

Time delay identification using permutation information theory quantifiers

Miguel C. Soriano, Luciano Zunino, Ingo Fischer, Claudio R. Mirasso

IFISC (CSIC - UIB), Instituto de Fı́sica Interdisciplinar y Sistemas Complejos,
Campus Universitat Illes Balears, E-07122 Palma de Mallorca, Spain

An important problem in the analysis of a sequence of is associated with embedding delay values that minimize
data points (time series) originating from a given physical the permutation entropy and maximize the permutation
process is the identification of time delays. Delay phe- statistical complexity, simultaneously. The presence of
nomena are intrinsic to many dynamical processes and additional peaks at harmonics and subharmonics of the
the identification of a delay would give valuable informa- feedback time delay allow us to better distinguish the
tion about the interaction between the system compo- delay time.
nents.
Autocorrelation function (ACF) and delayed mutual
1
information (DMI) are conventional techniques widely
used to identify time delays. However, new alternatives 0.95
were introduced in recent years in order to perform this
task. We are particularly interested in the application of 0.9

a permutation information theory methodology to unveil

HS
delay phenomena from time series1 . This method relies 0.85

on the computation of the permutation entropy and the

permutation statistical complexity. 0.8

We analyze time series originating from an exper-

imental realization of a semiconductor laser with de- 0.75

layed optical feedback and from them construct a prob- 0.5

ability distribution of their amplitudes. We choose the 0.45

Bandt and Pompe method to construct the probability 0.4

distribution2 , which considers the order of neighboring 0.35

values rather than partitioning the amplitude into differ- 0.3

ent levels. The probability distribution of the generated

CJS

0.25

ordinal pattern for a given time series can be established 0.2

once an embedding dimension D and an embedding delay 0.15

time τ are chosen. The embedding dimension D refers 0.1

to the number of symbols that forms the ordinal pattern. 0.05

0
The embedding delay τ is the time separation between
257

441

772
309
386

515

1.544

Embedding
Embedding-delay-tau
delay (samples)

symbols which is directly related to the sampling time of

Figura 1. Permutation entropy HS (top) and permutation
the time series (see refs.1,2 for a detailed derivation and statistical complexity CJS (bottom) as a function of the em-
description of the quantifiers). bedding delay τ with embedding dimensions D = 8 for the
In our experimental realization of a laser with opti- experimental time series. Positions of the local extrema asso-
cal feedback, the laser exhibits chaotic intensity pulsa- ciated with the feedback time delay τS and its subharmonics
tions. Time series are recorded with a sampling time of are detailed.
δs = 25 ps. From the length of the feedback loop, the
delay time is estimated to be around τext = 38.5 ns3 .
In Fig. 1 we plot the normalized permutation entropy
and the permutation statistical complexity obtained from
the experimental time series as a function of the em- ∗
miguel@ifi[Link]
bedding delay τ for an embedding dimension D = 8. 1
L. Zunino, M. C. Soriano, I. Fischer, O. A. Rosso, and C.
We verify experimentally that the permutation entropy
R. Mirasso, “Permutation information theory approach to
is minimized and the permutation statistical complexity unveil delay dynamics from time series analysis,” Phys.
maximized when the embedding delay τ of the symbolic Rev. E, 82, 046212 (2010).
reconstruction takes values near τext , i.e. for τ close to 2
C. Bandt and B. Pompe, “Permutation entropy - a com-
1540 samples (τext /δs = 1540). We have also found other plexity measure for time series”, Phys. Rev. Lett., 88,
extrema when the embedding delay matches harmonics 174102 (2002).
and subharmonics of τext . 3
M. C. Soriano, L. Zunino, O. A. Rosso, I. Fischer, and
In summary, time delays present in the system dynam- C. R. Mirasso, “Time scales of a chaotic semiconductor
ics are detected through the presence of clear extrema laser with optical feedback under the lens of a permuta-
of the quantifiers when they are calculated as a func- tion information analysis,” IEEE J. Quantum Electron.,
tion of the embedding delay. The feedback time delay 47, 252–261 (2011).

Panel P–40 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 95

Quasi-stationary analysis of the contact process on scale-free networks

Silvio C. Ferreira,∗† Ronan S. Ferreira,† , Romualdo Pastor-Satorras

Departament de Fı́sica i Enginyeria Nuclear
Universitat Politècnica de Catalunya, Campus Nord B4, 08034 Barcelona, Spain

Complex network theory represents a general unify- probability distribution of activity both close to the criti-
ing formalism under which is possible to understand and cal point and in the off-critical regime, as well as to obtain
rationalize the intricate connectivity and interaction pat- high quality data for QS relevant quantities, such as the
terns of many natural and man-made systems. Among density of active sites or the characteristic times.
their characteristics, probably the most intriguing is the
discovery of the apparently ubiquitous scale-free (SF) na- 0.015
ture of the connectivity pattern of many systems, de-
γ=2.25
scribed by a probability P (k) that an element (vertex) γ=2.75
ME
is connected to other k elements (has degree k), scaling
as a power-law of the form P (k) ∼ k −γ , with a degree 0.01

exponent usually in the range 2 < γ < 3. In recent

Pn
years, the research community has also devoted a great
deal of attention to the study of the dynamical processes
0.005
on complex networks, which can have important implica-
tions in understanding of the behavior of real processes
such as the spread of epidemics in social systems or traf-
fic in technological systems as the Internet or transport 0
0 50 100 150 200
infrastructures. n
Figura 1. Examples of QS probability distributions ob-
Dynamical processes with absorbing configurations are
tained in simulations of the critical CP on annealed SF net-
subjects of outstanding interest in non-equilibrium sta-
works and in numerical solutions of the master equation
tistical physics that have also found a place in network
(ME).
science. The simplest model allowing absorbing config-
urations is the classical contact process (CP). In the
CP defined in an arbitrary network, vertices can be in The high accuracy of our data (Fig. 1) allows to iden-
two different states, either empty or occupied. The dy- tify strong corrections to the scaling in the critical quan-
namics includes the spontaneous annihilation of occupied tities that mask the correct finite size scaling exponents
vertices, which become empty, at unitary rate and the obtained analytically by means of an exact mean-field
self-catalytic occupation of an empty vertex i with rate solution. Both critical density and characteristic time
λni /ki , where ni is the number of occupied neighbors of show tenuous curvatures as a function of the network
i and ki is its degree. The model is thus characterized size N due to finite size corrections to scaling that may
by a phase transition at a value of the control parameter provide incorrect exponents if a simple power law decay
λ = λc , separating an active phase from an absorbing is assumed. In annealed networks, for which the criti-
phase devoid of active vertices. cal point is exactly known, we can determine the cor-
The configuration in which all vertices are empty plays rections to scaling analytically and thus recover the the-
a very particular role, since once the system has fallen oretical exponents in the finite size analysis, including
into this state, the dynamics becomes frozen. For this the abrupt change when the network loses its SF proper-
reason, these states are called absorbing and constitute a ty. The analysis of the supercritical region, on the other
central feature in the analysis of finite size systems since, hand, hints that those finite size corrections are relevant
in this case, the only actual stationary state is the ab- for very large network sizes, the asymptotic scaling being
sorbing one. Finite size and absorbing states must there- observable only for extremely large values of N .
fore be handled using suitable strategies, concomitantly
with an ansatz for the finite size scaling (FSS) of the
dynamics. A widely adopted procedure is the so called
quasi-stationary (QS) state,1 in which the absorbing con- ∗
silviojr@[Link]
figuration is suitably excluded from the dynamics. †
Departamento de Fı́sica, Universidade Federal de Viçosa,
In this work, we present a study of the QS state of 36571-000, Viçosa - MG, Brazil
CP on SF annealed networks, combining the QS nu- 1
M. M. de Oliveira and R. Dickman, Phys. Rev. E 71,
merical approach developed in Ref. 1, suitably extend- 016129, (2005).
ed to complex networks, with the theoretical analysis of 2
C. Castellano and R. Pastor-Satorras, Phys. Rev. Lett.
a approximated one-step process derived from mean-field 100, 148701 (2008).
theory2,3 and solving the corresponding master equation. 3
M. Boguñá, C. Castellano and R. Pastor-Satorras, Phys.
Our analysis allows us to obtain information about the Rev. E 79, 036110,(2009).

Junio de 2011, Barcelona Panel P–41

96 XVII Congreso de Fı́sica Estadı́stica FisEs11

Simple rules govern finite-size effects in scale-free networks

Sara Cuenda∗ , Juan A. Crespo†

Departamento de Economı́a Cuantitativa
Facultad de CC. Económicas y Empresariales
Universidad Autónoma de Madrid
Fco. Tomás y Valiente 5, 28049 Madrid.

The growing network with preferential attachment pro- 1

posed by Barabási and Albert1 introduced a simple, effec- <Fsyn(k,t)>
tive model of networks that would lead to a power-law Fe(k,t)
distribution of the degree of nodes. The results of this 0.1 Fm(k,t)
model should be understood within the limit of large Fsyn(k,t)
networks, as some investigations showed the existence
of finite-size effects that make the distribution of the de-

F(k,t)
0.01
gree of nodes depart from the power-law2–5 . These works
show/predict a cut-off degree from which distributions of
finite networks created with this model stop from behav-
ing as power-laws, and acknowledge the influence of the 0.001
initial nodes from which the network grows in the final
result. However, a general prediction of the degree dis-
tributions of finite networks in terms of the initial nodes 0.0001
has not been made so far. 1 10 100
k
In this work we find a general, theoretical prediction 1
of the final degree distribution of finite networks grow- b)
ing with preferential attachment in terms of the initial
degree distribution. We obtain the expression of the fi- Fsyn(k,t)
nal distribution using two different approaches: the well- Fe(k,t)
known deterministic mean-field approximation, and the Fm(k,t)
0.1
more accurate probability distribution of the degree of
1
each node, which considers the stochastic process.
F(k,t)

The model that we use is the original introduced by

Barabási and Albert1 , in which the probability of a node 0.1
0.01
with degree k of gaining a new link in a network of size
t when a new node arrives is πk,t = mk/ i ki , where m 0.01
is the number of undirected links of the new node to be
attached to the nodes in the network. Since this prob-
0.001
ability only depends on the state (degree) of each node, 1 10 100
0.001
we can model the dynamics of every node in order to 1 10 100
k
obtain the final distribution of the network. We did it Figura 1. Complementary, cumulative degree distribution
considering also the initial nodes in the network, which of the final network for different initial conditions.
are distributed according to the probability distribution
f0 (k).
Using the mean-field approximation we obtain an ana-
lytical expression of the final density distribution, which ∗
[Link]@[Link]
contains the effect of the initial nodes, but with no ef- †
[Link]@[Link]
fects of the dispersion in it. The expression obtained, 1
A.-L. Barabási , R. Albert and H. Jeong, Phys. A 272,
however, explains previous results on the universality of p. 173 (1999).
the cut-off function for finite systems starting from the 2
S.Ñ. Dorogovtsev, J. F. F. Mendes and A.Ñ. Samukhin,
same initial core2,3,5 . The theoretical calculation of the Phys. Rev. E 63, p. 62101 (2001).
stochastic dynamics of every node renders an exact ex- 3
P. L. Krapivsky and S. Redner, J. Phys. A: Math. Gen.
pression of the expected final degree distribution (exact 35, p. 9517, (2002).
within the limits of the accuracy of πk,t ). Numerical sim- 4
P. R. Guimarães et al., Phys. Rev. E 71, p. 37101 (2005).
5
ulations support very well our results, as shown in figure B. Waclaw and I. M. Sokolov, Phys. Rev. E 75, p. 056114
1. (2007).

Panel P–42 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 97

Replicator dynamics for the n-player Prisoner’s Dilemma with “moody” conditional
cooperators

Jelena Grujić1 , José A. Cuesta1 , Angel Sánchez1,2

1
GISC, Dpto. de Matemáticas, Universidad Carlos III de Madrid, Leganés
2
BIFI, Universidad de Zaragoza, Zaragoza

We have recently performed an experiment to test the probabilities for all combinations of actions of three or
emergence of cooperation in the presence of an under- four players. Thus, for instance, for m = 3 the payoff
lying structure1 . Human subjects played a Prisoner’s matrix will be a tensor of size 3 × 3 × 3 and the replicator
Dilemma (PD) with each of their neighbors in a 13 × 13 equation will be:
square lattice. The results show that the population ⎛ ⎞
consisted of cooperators and defectors, who respective-  
ly cooperate or defect with high probability regardless of ẋi = xi ⎝ Aijk xj xk − Ajkl xj xk xl ⎠ , (3)
their and their neighbors’ previous actions, and “moody” j,k j,k,l
conditional cooperators, whose behavior does depend on
those previous actions. and the obvious generalization will hold for m = 4.
Here we take a first step towards an evolutionary ex- For all the group sizes, the dynamics exhibits two at-
planation of the aforementioned experimental results. tractors: a population consisting only of defectors and an
Specifically, we use replicator dynamics to describe the interior point with population frequencies comparable to
evolution of a set of strategies that mimics the observa- those observed in the experiment. This interior point has
tions, in a simplified context consisting of a well-mixed a much larger basin of attraction than full defection, and
population of players confronted in iterated n-player PD thus it becomes the most probable evolutionary outcome
games. We consider three strategies: cooperators (coop- (Fig. 1). While this is the first hint that the experiment
erate with probability p), defectors (cooperate with prob- results may be understood from an evolutionary view-
ability 1 − p) and “moody” conditional cooperators (they point, more work is needed to ascertain the dependence
cooperate with probability pD following a defection, and of the size of the basis of attraction of the interior point
with a probability that increases linearly with the number as the group size increases.
of cooperating opponents, varying between pC0 and pC1
when none or all neighbors cooperated). Players played C
a PD game with each of their m − 1 opponents taking
only one action, either to cooperate (C) or to defect (D),
the action being the same against all the opponents; we
present results for m = 2, 3, and 4.
Denoting by xi , i = 1, 2, 3 the fractions of the three
strategies (x1 + x2 + x3 = 1), the dynamics of xi for the
case of pairwise PD (m = 2) is given by
 
ẋi = xi (Ax)i − x · Ax (1)

where A is the payoff matrix, and Aij is the average pay-

off that strategy i yields when confronted to strategy j
in an infinitely long iterated PD. To compute A, we need
to find first for every pair of strategies i, j the probability X D
that players play X’Y’ in the next round having played Figura 1. Population dynamics for the pairwise PD with
XY in the present round (X, X’, Y, Y’ can be C or D and defectors D, cooperators C and “moody” conditional cooper-
the first action listed corresponds to the strategy whose ators X. Black circles represent the attractors, gray ones are
payoff is computed, the second action being that of its saddle points and white ones are repellers. Trajectories are
opponent). This defines a Markov process with transi- plotted in gray and the arrows show the flow of the dynamics.
tion matrix M , whose stationary probability is obtained The thick solid line is the separatrix between the two basins
by solving the equation π = πM , leading to a payoff for of attraction.
strategy i vs strategy j will be given by2
Aij = RπCC + SπCD + T πDC + P πDD . (2)
Repeating the calculation for all pairs yields A, the payoff 1
J. Grujić, C. Fosco, L. Araujo, J.A. Cuesta, A. Sánchez,
matrix in Eq. (1). Social experiments in the mesoscale: Humans playing a
For the cases m = 3, 4, we apply the same procedure, spatial Prisoner’s Dilemma PLoS ONE 5, e13749 (2010).
except that M matrices will now be 8 × 8 or 16 × 16, re- 2
K. Sigmund, The Calculus of Selfishness (Princeton Uni-
spectively, since we now need to calculate the transition versity Press, Princeton, 2010).

Junio de 2011, Barcelona Panel P–43

98 XVII Congreso de Fı́sica Estadı́stica FisEs11

Coarse graining en dinámica browniana: modelos de difusión discreta

J. A. de la Torre1∗ , Pep Español1,2

1
Dpto. Fı́sica Fundamental, Universidad Nacional de Educación a Distancia, Apdo. 60141 E-28080 Madrid, España
2
Freiburg Institute for Advanced Studies, Freiburg im Breisgau, 79100, Alemania

Utilizando el formalismo de Zwanzig proponemos el de- tribución de probabilidad en distintos nodos) como el
sarrollo de diversos modelos de grano grueso para formu- comportamiento dinámico que se refleja en las correla-
lar modelos de difusión discreta. Trabajar con modelos ciones temporales (Fig. (1)).
sencillos como el de partı́culas brownianas no interac- Con el objetivo de hacer una simplificación mayor pro-
cionantes permite centrar el estudio del coarse graining ponemos considerar una energı́a libre aditiva local. Sim-
en aspectos esenciales del mismo: cómo, dada una de- ulaciones realizadas bajo esta aproximación demuestran
scripción microscópica, podemos obtener una descripción que si estamos interesados en propiedades globales del
macroscópica formalmente equivalente a la anterior con sistema o con una clara separación de escalas (nodos ale-
un coste computacional menor. Trabajos previos han uti- jados o tiempos largos) esta nueva hipótesis es óptima
lizado la triangulación de Delaunay y la teorı́a del coarse- para simplificar nuestro problema. Sin embargo las dis-
graining en un modelo hidrodinámico discreto a partir de tribuciones de probabilidad conjunta para nodos vecinos
la descripción de Smoluchowski1 . En el presente traba- no se ajustan a la dinámica real del sistema. En el caso
jo, establecido un nivel de descripción para un problema de las correlaciones se observa además que al incidir sobre
de difusión de partı́culas coloidales, se hace una descrip- los detalles del proceso (interacción entre nodos vecinos
ción de grano grueso en la que se trabaja no ya con las o tiempos cortos) aparecen inexactitudes que nos obligan
posiciones de las partı́culas coloidales sino con la concen- a considerar el modelo cuadrático.
tración, nμ (z), en el entorno de un nodo μ definido por
la triangulación de Delaunay. Nuestro objetivo es, a par-
tir de este formalismo, desarrollar de forma explı́cita las
1
ecuaciones de difusión discreta en forma de una ecuación BD
GA
diferencial estocástica que pueda ser simulada y compara- 0.75
LE
<δnμ δnν(t)> / <δnμ δnμ>

da posteriormente con los resultados de la dinámica real μ=ν

del sistema de partı́culas coloidales. 0.5
Si bien la teorı́a de Zwanzig da una solución formal al
μ=ν+2
proceso de granulado para la probabilidad en forma de 0.25
una ecuación de Fokker-Planck, ésta queda descrita en
términos de dos bloques de información computacional- 0
μ=ν+5
mente difı́ciles de calcular: la energı́a libre y la matriz
de fricción. En nuestro trabajo probamos que la matriz -0.25
0 10 20 30 40 50
de fricción, para perfiles de concentración suficientemente t
suaves, puede modelarse aproximando el promedio condi- Figura 1. Correlaciones de equilibrio para la dinámica
cional de la concentración por el valor medio de las con- browniana (BD), la aproximación gaussiana (GA) y la
centraciones en nodos vecinos. La energı́a libre queda a hipótesis de energı́a libre local (LE) para distintas vecindades
su vez bien modelada en el caso de un gran número de entre nodos μ y ν.
partı́culas coloidales por nodo por un modelo gaussiano
no local cuadrático en las concentraciones
Comparando simulaciones de este modelo con la
dinámica real obtenida mediante dinámica browniana
∗
comprobamos que se verifican tanto las propiedades es- jatorre@[Link]
1
tadı́sticas en el equilibrio (a partir del cálculo de la dis- P. Español e I. Zúniga, J. Chem. Phys 131, 164106 (2009).

Panel P–44 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 99

Phase diagrams of binary mixtures of patchy colloids with distinct numbers of patches

Daniel de las Heras, José Maria Tavares, Margarida M. Telo da Gama

Centro de Fı́sica Teórica e Computacional da Universidade de Lisboa
Avenida Professor Gama Pinto 2
P-1749-016, Lisbon, Portugal

We calculate the phase diagram and the connectivi- bonding drives the phase separation of two network flu-
ty of the coexisting fluid phases of binary mixtures of (1)
ids and if fA > 2 it changes the topology of the phase
patchy colloids, using Wertheim’s first order perturba- diagram of the mixture, from type I to type V. The dif-
tion theory and a generalization of Flory-Stockmayer’s ference in functionality also determines the miscibility at
theory of polymerization. The colloids are modelled by (2) (1)
high pressures: When fA − fA = 1 the mixture is com-
equisized hard spheres with a number of identical patch- pletely miscible, while closed miscibility gaps are present,
es on their surfaces, with species distinguished by the above the critical pressure of the less volatile component,
(1) (2)
number of patches -or functionality-, fA and fA (with (2) (1)
when fA − fA > 1. We argue that this rich phase be-
(2) (1)
fA > fA ). We found that the difference in functionali- haviour is driven by a competition between the entropy
ty is the key factor controlling the phase behaviour of the of mixing and the entropy of bonding.
(2) (1)
mixture. In particular, when fA > 2fA the entropy of

Junio de 2011, Barcelona Panel P–45

100 XVII Congreso de Fı́sica Estadı́stica FisEs11

Universality of rain event size distributions

Anna Deluca1∗ , O. Peters2,3, A. Corral1, J. D. Neelin3 , C. E. Holloway4

1
Centre de Recerca Matemàtica, Spain
2
Dept. of Mathematics and Grantham Institute for Climate Change, Imperial College London, UK.
3
Dept. of Atmospheric and Oceanic Sciences and Institute for Geophysics and Planetary Physics,UCLA, USA
4
Dept. of Meteorology, University of Reading, UK

Atmospheric convection and precipitation have been consistent with a power-law distribution and the esti-
hypothesised to be a real-world realization of self- mates of the apparent exponents are in agreement with
organized criticality (SOC). This idea is supported the hypothesis of a single exponent τs = 1.17(3), except
by observations of avalanche-like rainfall events1,2 and for three problematic data sets from Point Reyes, the
by the nature of the transition to convection in the Southern Great Plains and Alaska.
atmosphere4,5 . However, many questions remain open. Climatic differences between regions are scarcely de-
Here we ask whether the observation of scale-free tectable in event size distributions, which may be sur-
avalanche size distributions is reproducible using data prising on the grounds of climatological considerations.
from different locations and whether the associated fitted However, the cutoff sξ , representing the capacity of the
exponents show any sign of universality. climatic region around a measuring site to generate rain
We study rain data from all 10 available sites of the events, changes significantly from region to region, con-
Atmospheric Radiation Measurement (ARM) Program, firming meteorological intuition, and is easily extracted
see [Link], over periods from about 8 months to from the moments of the distributions. While the ex-
4 years3. The measurements are from climatically differ- ponents are not significantly different, the larger tropical
ent regions using a standardized technique, making them events are reflected in the greater large-scale cutoff of the
ideal for our purpose. Precipitation rates were recorded tropical distributions.
at one-minute resolution with an optical rain gauge. Similarly, the dry-spell durations (durations of
precipitation-free intervals) seem to follow another pow-
er law with τd = 1.2(1), and regional differences can be
seen in the strength of the diurnal cycle and the cut-
off dry spell duration. The broad range of event dura-
tions suggests a link to the lack of characteristic scales in
the mesoscale regime, where approximately scale-free dis-
tributions of clusters of convective activity, for example
cloud or precipitation, have been observed to span ar-
eas between O(1 km2 ) and O(106 km2 ), see for example
Ref.9 . The observation of scale-free rainfall event sizes
suggests long-range correlation in the pertinent fields, a
possible indication of critical behaviour near the transi-
tion to convective activity.

∗
adeluca@[Link]
1
Figura 1. Probability densities of event sizes, s in mm, and Andrade R F S, Schellnhuber H J, and Claussen M. Phys-
a power-law fit (black straight line). Inset: Precipitation rates ica A, 254(3-4):557–568, 1998.
2
including two rain events lasting 7 and 15 minutes respectively. In- Peters O, Hertlein C, and Christensen K. Phys. Rev. Lett.,
terpreting reported rain rates of less than 0.2 mm/h as zero, the 88(1):018701(1–4), 2002.
3
shaded areas are the corresponding event sizes. Peters O, Deluca A, Corral A, Neelin J D and Holloway C
E J. Stat. Mech., P11030, 2010.
Following Ref.2 , we define an event as a sequence of 4
Peters O and Neelin J D. Nature Phys., 2(6):393–396,
non-zero measurements of the rain rate, see inset in the
2006.
figure . The event size s is the rain rate, r(t), integrat- 5
Neelin J D, Peters O, and Hales K. J. Atmos. Sci.,
ed over the event, s = event dt r(t). For each data set, 66(8):2367–2384, 2009.
the probability density function Ps (s) in a particular size 6
Arakawa A and Schubert W H. J. Atmos. Sci., 31(3):674–
interval [s, s + Δs) is estimated. 701, 1974.
The distributions, shown in the figure, are visually 7
Tang C and Bak P. Phys. Rev. Lett., 60(23):2347–2350,
compatible with a power law (black straight line) over 1988.
most of their ranges. A procedure similar to that in 8
Clauset A, Shalizi C R, and Newman M E J. SIAM Rev.,
Ref.8 consisting of maximum-likelihood estimation plus 51(4):661–703, 2009.
9
a goodness-of-fit test confirms this result: over ranges Peters O, Neelin J D, and Nesbitt S W. J. Atmos. Sci.,
between 2 and 4 orders of magnitude, all data sets are 66(9):2913–2924, 2009.

Panel P–46 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 101

Teorema de fluctuación-disipación para un oscilador activo ruidoso, el haz ciliar del

oı́do interno.

L. Dinis1,2,∗ , P. Martin1 , J. Barral1, J. Prost1,3 , J.F. Joanny1

1
Institut Curie, CNRS UMR 168, 26 rue d’Ulm 75248 Paris Cedex 05 France
2
Grupo Interdisciplinar de Sistemas Complejos (GISC)
3
ESPCI, 10 rue Vauquelin, 75231 Paris Cedex 05 France

El teorema de fluctuación disipación (TFD) es un re-

100
sultado fundamental de la fı́sica estadı́stica. Relaciona la
80
respuesta de un sistema a una perturbación externa con
60
la correlación de fluctuaciones en el equilibrio térmico.
40
Sin embargo, los sistemas alejados del equilibrio no
20
cumplen en general el teorema. Uno de los casos más

Teff/T
0
espectaculares es el de los sistemas cerca de una bi- −20
furcación de Hopf. El haz ciliar (‘hair-bundle’) de las −40
células ciliadas del oı́do interno de los vertebrados propor- −60
ciona un ejemplo experimental de uno de estos sistemas1 . −80

El análisis de las oscilaciones tanto espontáneas como −100

forzadas externamente muestra una fuerte violación del 2 4 6 8

ν(Hz)
10 12 14 16

teorema: la temperatura efectiva definida a partir del

Figura 1. Temperatura efectiva (en unidades de la temper-
TFD diverge en torno a la frecuencia espontánea de os-
atura del sistema) en función de la frecuencia ν de la estimu-
cilación ν0 , como se muestra en la figura 1. La diver-
lación externa.
gencia corresponde a un cambio de signo de la parte
imaginaria de la función de respuesta lineal, que mide
la energı́a transferida al sistema por parte del forzamien-
to. Para estı́mulos de frecuencias menores a la frecuencia
de oscilación espontánea, el forzamiento externo recibe
3.5
en realidad energı́a del haz ciliar, indicando la presencia
de un proceso “activo” interno que mantiene al sistema 3

fuera del equilibrio.

2.5
En años recientes han aparecido varias generalizaciones
del teorema de fluctuación disipación para sistemas fuera 2
T’eff

del equilibrio. En particular, el teorema generalizado

(TFDG) de Prost et al.2 se aplica a sistemas de markov
1.5

que evolucionan a partir de un estado estacionario. Uti- 1

lizando los datos del experimento junto con una descrip-

0.5
ción del sistema en la forma normal de Hopf, mostramos
que el teorema generalizado se cumple aproximadamente 0
2 4 6 8 10 12 14 16
ν(Hz)
si se elige la variable conjugada de la fuerza externa
apropiada. El resultado se muestra en la figura 2 en la Figura 2. Resultado de la fracción πν C̃(ν)
en la variable
χ̃ (ν)
que se representa la fracción conjugada apropiada, calculada según cuatro procedimientos
diferentes. El teorema generalizado de fluctuación-disipación
 πν C̃(ν)
Teff ≡ (1) predice que la fracción vale 1 para sistemas de markov que
χ̃ (ν) evolucionan desde un estado estacionario. En cualquiera de
los cuatro métodos, la fracción es cercana a 1 al menos para
donde ν es la frecuencia, C̃(ν) representa el espectro frecuencias próximas a ν0  6Hz, la frecuencia propia de os-
de potencia de las fluctuaciones en la nueva variable y cilación.
χ̃ (ν) la parte imaginaria de la función de respuesta lin-
eal de la nueva variable a la perturbación externa. Como
se ve en la figura, su valor es aproximadamente 1 para
estı́mulos externos de frecuencias próximas a la frecuen-
cia espontánea. Si bien el resultado no es 1 para to- ∗
ldinis@fi[Link]
da frecuencia como predice el TFDG, la divergencia que 1
P. Martin, A. J. Hudspeth and F. Jülicher, Proc. Natl.
aparecı́a en la temperatura efectiva en la variable original
Acad. Sci. USA 98, 14380 (2001)
desaparece completamente. 2
J. Prost, J. F. Joanny and J. M. R. Parrondo, Phys. Rev.
Lett. 9, 090601 (2009)

Junio de 2011, Barcelona Panel P–47

102 XVII Congreso de Fı́sica Estadı́stica FisEs11

Information routing driven by background chatter in signaling networks

Núria Domedel-Puig, Pau Rué, Antonio J. Pons, Jordi Garcı́a-Ojalvo∗

Departament de Fı́sica i Enginyeria Nuclear
Universitat Politècnica de Catalunya
Edifici GAIA, Rambla de Sant Nebridi s/n
08222 Terrassa (Barcelona) Spain

Transformation of information into chemical change, chatter-dependent paths (Fig 1). In certain cases, back-
signal transduction, is a property of all living cells. Sig- ground activity allows input signals to be transmitted to
naling molecules stimulate their receptors, which trans- downstream nodes that would be unreachable otherwise.
mit the signal downstream through a series of protein- The existence of this chatter-induced information rout-
protein interactions that modify DNA expression and ing appears to rely on the structure and dynamics of a
protein levels. In this manner, information affects cell truly biological network, as we find that random versions
behavior. The description of signal transduction has of the system are unable to display this feature.
classically involved independent signaling cascades -or
pathways-, in which information is linearly transmitted
from membrane to nucleus. Contributing to this simpli-
fied view of cell signaling, experimental studies have usu-
ally analyzed pathway stimulation by single inputs, such
as variations in one chemical (nutrients, hormones, etc.)
or physical property (presence of light, pressure, etc.).
However, extracellular media usually contain a complex
mix of molecules that have the potential to feed the
signaling network with multiple inputs simultaneously1 .
Furthermore, it is now known that proteins of one signal-
ing cascade often interact with proteins of other path-
ways, forming a dense web of intracellular connections
both in eukaryotes and prokaryotes2. Finally, the fact
that cells process diverse signals under a fluctuating en-
Figura 1. Structure of the human fibroblast signaling net-
vironment provides a source of random variation at the
work. The network has 9 input nodes (top row), 6 output
input level that the cell must also deal with3 .
nodes (bottom), and 124 internal nodes. A periodic stimula-
Here we explore the impact of this signal diversity and tion has been applied to input stress under different chatter
variation upon the information processing capabilities of levels. The most utilized paths to node p38 are then identified
a cell signaling network as a whole. In particular, we based on correlation measures, and shown here in red (high
study how transmission of information coming from one score at low chatter), blue (high score at low and medium
single input is affected by the background activity, or chatter), and yellow (top scoring at high chatter levels).
chatter, provided by other network inputs. To address
this issue in a way that explicitly accounts for the com-
plexity of the system under consideration, we use one
of the most comprehensive dynamic models of cell sig-
naling currently available in the literature: a recently ∗
published Boolean network for the human fibroblast that [Link]@[Link]
1
involves over 130 protein species4 (Fig 1). The dynamics Natarajan M, Lin K, Hsueh RC, Sternweis PC, Ran-
of this network are implemented as a set of logic rules, an ganathan R (2006) A global analysis of crosstalk in a
mammalian cellular signalling network. Nature Cell Bi-
approach that -despite its simplicity- represents a good
ology, 8:571–580.
choice when building a detailed kinetic model is unfeasi- 2
Oda K, Kitano H (2006) A comprehensive map of the Toll-
ble. Indeed, Boolean networks have successfully been ap- like receptor signaling network. Molecular Systems Biolo-
plied to modelling numerous biological processes, show- gy, 2:2006.0015.
ing that sequences of events can be reproduced by this 3
Suel G, Kulkarni R, Dworkin J, Garcia-Ojalvo J, Elowitz
type of discrete dynamic models5 . M (2007) Tunability and noise dependence in differentia-
We have characterized the response of this network to tion dynamics. Science, 315:1716–1719.
periodic signals under different chatter levels by perform- 4
Helikar T, Konvalina J, Heidel J, Rogers J (2008) Emer-
ing extensive numerical simulations. Our findings suggest gent decision-making in biological signal transduction net-
that the level of background activity shapes the response works. Proc Natl Acad Sci USA, 105:1913–1918.
of the entire network to the external signal. Indeed, spe- 5
Huang S, Eichler G, Bar-Yam Y, Ingber D (2005) Cell fates
cific levels of background chatter activate different areas as highdimensional attractor states of a complex gene reg-
of the network, causing the information to travel through ulatory network. Phys Rev Letters, 94:128701.

Panel P–48 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 103

Del colapso gravitatorio a la descomposición espinodal

Alvaro Domı́nguez1∗ , Johannes Bleibel2 , Martin Oettel3 , Siegfried Dietrich2

1
Fı́sica Teórica, Universidad de Sevilla, Apdo. 1065, E–41080, Sevilla
2
Max–Planck–Institut für Metallforschung, Heisenbergstr. 3, D–70569, Stuttgart
3
Institut für Physik, Johannes–Gutenberg–Universität Mainz, WA 331, D–55099, Mainz

Un gas de Yukawa bidimensional consiste en esferas ción de estructuras en el régimen no lineal (inhomogenei-
duras (diámetro σ) que se atraen según el potencial dad alta). Los resultados se pueden resumir en términos
del cociente λ/L, donde la escala L( σ) caracteriza el
Vatr (d) = −V0 K0 (d/λ) (V0 , λ > 0) tamaño de la distribución de partı́culas:
en términos de la función de Bessel modificada K0 , que • En el lı́mite λ/L  1 se observa un proceso de
exhibe un decaimiento exponencial para d λ (corres- descomposición espinodal según la distribución de
pondiente a una interacción de corto alcance), pero un partı́culas se separa en una fase gaseosa y otra
comportamiento logarı́tmico para d  λ (formalmente lı́quida.
análogo a la gravitación newtoniana bidimensional y por
tanto no integrable en el sentido de la mecánica es- • Por el contrario, en el lı́mite λ/L 1 se observa
tadı́stica del equilibrio). Los estudios previos del gas de una inestabilidad análoga a la del colapso gravita-
Yukawa han considerado invariablemente σ ≈ λ; nuestro torio con la formación de cúmulos densos en una
objetivo es estudiar el caso σ  λ, de forma que el efecto distribución diluı́da de partı́culas.
del comportamiento logarı́tmico sea patente.
Aparte de su interés intrı́nseco, este modelo es de im- La evolución en este segundo lı́mite ocurre sobre una es-
portancia práctica porque describe la interacción por cala de tiempo caracterı́stica independiente de λ y mucho
fuerzas capilares de partı́culas coloidales atrapadas en más corta que la de descomposición espinodal. A pe-
una interfase fluida. Se trata de un sistema que hoy en dı́a sar de esto, es posible identificar claramente dos etapas
se genera y estudia rutinariamente en el laboratorio, para reminiscentes de los regı́menes de separación de fases y
los cuales V0 y λ dependen de parámetros experimental- crecimiento de dominios, respectivamente, de la descom-
mente controlables con relativa facilidad. En particular, posición espinodal. La evidencia sugiere que la transición
una configuración tı́pica involucra σ ≈ 1 μm (partı́culas entre estas dos dinámicas lı́mites es continua como fun-
coloidales) y λ ≈ 1 mm (longitud capilar). ción del parámetro λ.
Nos hemos centrado en el estudio de la inestabilidad de
una distribución homogénea de partı́culas. El problema
se ha abordado teóricamente mediante una aproximación ∗
dominguez@[Link]
de campo medio1 , ası́ como con ayuda de simulaciones de 1
Domı́nguez, Oettel, and Dietrich, Phys. Rev. E 82, 011402
dinámica molecular2 . Estas últimas permiten la medición (2010).
2
de cuantificadores particularmente sensibles a la forma- Bleibel, Domı́nguez, Oettel, and Dietrich, preprint (2011).

Junio de 2011, Barcelona Panel P–49

104 XVII Congreso de Fı́sica Estadı́stica FisEs11

Supernormal conduction in cardiac tissue promotes concordant alternans and action

potential bunching

Blas Echebarria1∗ , Georg Röder2 , Harald Engel3 , Jörn Davidsen4 , Markus Bär5
1
Departament de Fı́sica Aplicada, Universitat Politècnica de Catalunya, E-08028 Barcelona, Spain
2
Max-Planck-Institut für Physik komplexer Systeme, 01187 Dresden, Germany
3
Institut für Theoretische Physik, Technische Universität Berlin, 10623, Germany
4
Department of Physics and Astronomy, University of Calgary, Calgary, Alberta, Canada AB T2N 1N4,
5
Physikalisch-Technische Bundesanstalt, 10587 Berlin, Germany.

Propagation of pulse trains in excitable media is usual- nations in the duration of the excited phase of the AP)
ly characterized by a so-called dispersion curve that gives in excitable cardiac tissue. We use bifurcation analysis
the velocity of a pulse in a periodic train as a function and simulations to study attraction between propagating
of its wavelength. Since a pulse typically is followed by a APs caused by SNC, that leads to AP pairs and bunch-
refractory zone of decreased excitability, the pulse veloc- ing. It is shown that SNC stabilizes concordant alternans
ity monotonically increases with increasing wavelength in arbitrarily long paced one-dimensional cables. As a
(=normal dispersion). In the context of excitable cardiac consequence, spiral waves in two-dimensional tissue sim-
tissue, normal dispersion corresponds to normal conduc- ulations exhibit straight nodal lines for SNC in contrast
tion. Supernormal conduction (SNC) in excitable cardiac to spiraling ones in the case of normal conduction1 .
tissue refers to an increase of pulse (or action potential)
velocity with decreasing distance to the preceding pulse.
∗
Here, we employ a simple ionic model to study the effect blas@[Link]
1
of SNC on the propagation of action potentials (APs) B. Echebarria, G. Röder, H. Engel, J. Davidsen, Markus
and the phenomenology of alternans (beat-to-beat alter- Bär, Physycal Review E, in press.

Panel P–50 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 105

Local routing of the Internet based on a linear projection of complex networks

Pau Erola∗ , Sergio Gómez, Alex Arenas

Departament d’Enginyeria Informàtica i Matemàtiques
Universitat Rovira i Virgili, 43007 Tarragona (Spain)

Internet is the largest synthetic complex system ever between nodes. First, we construct the matrices Dz with
built, with a collection of more than 30,000 networks, the N × N distances between any pair of nodes of the
each one known as an Autonomous System. In the last network at stage z. Then, we compute the matrices of
few years, Internet is experiencing an explosive growth weighted distances S z using a gaussian distribution that
that is compromising its navigation scalability1,2 due to prioritizes the stability of closer nodes over the distant
its dependence on the Border Gateway Protocol (BGP). ones,
The BGP routing protocol requires to maintain an updat-
z 2 2
ed view of the network topology, involving a huge amount z z −Dij /2(σR0i )
Sij = Dij e ,
of data exchange and significant convergence times. The
scale-free topology of Internet makes the complex net- where Ri0 is the module of the node i, and σ is a constant.
work theory the natural framework to analyze its prob- Finally, the local measure of reliability is computed as
lems. Here, we propose a local alternative to this rout-
ing protocol based on complex networks. Our approach ⎛ ⎞⎛ ⎞−1
guarantees a high reliability over time and the simula- 
N 
N 
N 
N
Elocal = ⎝ |Sij
z 0 ⎠⎝
− Sij | 0 ⎠
|Sij | ,
tion results show that we are in high percentage close to
i=1 j=1 i=1 j=1
optimal paths.
In many strongly clustered networks, without a global
where S 0 and S z represent the matrices of weighted dis-
view of the topology, a message can be routed efficiently.
tances of the initial network and the grown network.
Assuming the hypothesis that the mesoscopic structure
We have selected two snapshots of the AS network6 ,
provides meaningful insights on this routing property, we
June and December 2009, as our test set. With them we
analyze the contribution of each node of a network to
construct two undirected and unweighted networks. We
modules using the projection technique introduced by
have simulated 106 paths with a success rate of 93% and
Arenas et al.3 . This projection is based on a rank 2
an average path of 6 steps (due to the long tail of the
TSVD and constructs a plane U2 where each node has
distribution). The global error between both projected
a coordinate pair. For each pair we calculate the polar
networks is 18%, however, the local error is only 0.007%.
coordinates (R, θ) and the values Rint and Rext . R and
This reinforces the idea that our projection is very ro-
θ enlighten about the degree and connectivity pattern
bust against evolving data. Likewise, the Figure below
of each node, Rint informs about the internal contribu-
shows that the distribution of path length of our greedy
tion of nodes to their corresponding modules, and Rext
algorithm does not significantly degrade.
reflects the boundary structure of modules.
Using these values, we propose a greedy routing algo- 60
rithm that at each step chooses the neighbor that mini- 6
β=10 , λ=0.7 All Shortest Paths
Greedy June 2009
mizes the function 50 Greedy December 2009
Avg Shortest Paths
40
costk = β(λ + |Δθk→j |)/Rintk if k ∈ αj ,
frequency %

= |Δθk→j |/Rextk if k ∈ αj ∧ Rextk > 0, 30

= (λ + |Δθk→j |)/Rintk otherwise; 20

where |Δθk→j | is the angular distance between the neigh- 10

bor k and destination j; αj is the destination community;
0
and λ and β are constants. For more details see 3,4 . 0 5 10
path length
15 20 25
This study is also concerned about the reliability of
this projection: if data is continuously changing the pro-
jection might become obsolete. In our earlier work5 we
defined two measures to quantify the differences between
two projections. The first measure is of global reliabili- ∗
[Link]@[Link]
ty and indicates the amount of change in the position of 1
RFC 4984, The Internet Architecture Board (2000)
nodes in U2 . It is computed by the relative error between 2
Nature Physics 5, 74–80 (2009)
the U2 coordinates of the initial network and the grown 3
New J. Phys. 12, 053009 (2009)
network. Nonetheless, the neighborhood of each node in 4
Int. J. of Complex Systems in Science 1, 37–41 (2011)
the U2 plane may differ and be hardly reflected by the 5
arXiv:1101.2064 [[Link]-an]
6
measure defined above. Thus, we proposed a norm that
reflects these local changes using the weighted distances [Link]

Junio de 2011, Barcelona Panel P–51

106 XVII Congreso de Fı́sica Estadı́stica FisEs11

Changes in buoyancy-driven instabilities by using a reaction-diffusion system

Darı́o Martı́n Escala∗ , Alberto Pérez Muñuzuri

Departamento de Fı́sica da Materia Condensada, Grupo de Fı́sica Non Lineal†
Facultade de Fı́sica ,Univerisdade de Santiago de Compostela
15782 Santiago de Compostela

At the interface generated in the mixing of misci- of two fluids of different density and diffusion coefficient.
ble fluids1 , instabilities can displayed by the difference
between the fluids densities and diffusion coefficients. The mathematical models involved in these phenom-
These instabilities generate characteristic patterns that ena are solved using numerical methods such as finite
affect the mass transport between the two species. differences, finite volumes and finite elements.
BZ reaction (Belousov-Zhabotinsky) is a chemical re-
action where, due to the autocatalysis of its intermedi-
aries and the difference between diffusion coefficients of
the same, are generated chemical oscillations and waves ∗
dmescala@[Link]
that result in pattern formation when the reaction is †
[Link]
carried out in two-dimensional media. 1
T. Bansagi, Jr., D. Horvath, A. Toth, J. Yang, S. Kalli-
adasis, and A. De Wit. Phys. Rev. E 68, 055301(R) (2003)
2
The aim of this study is to analyze the influence of re- S. Kalliadasis & J. Yang,A. De Wit, Phys. Fluids, 16,
action diffusion on the instabilities caused by the contact 5,(2004)

Panel P–52 Junio de 2011, Barcelona

XVII Congreso de Fı́sica Estadı́stica FisEs11 107

A coarse-grained model for water in hydrophobic confinement

Cristian Esplugas∗ , Giancarlo Franzese

Departament de Fisica Fonamental,
Universitat de Barcelona, Diagonal 647,
08028 Barcelona, Spain

Hydrophobic confinement allows to supercool liquid the case of an arbitrary number of water layers between
water well below its melting temperature and is relevant hydrophobic plates. Here we present our preliminary re-
in biology and technological applications. By perform- sults about the possible liquid-liquid phase transition, a
ing Monte Carlo simulations, we study the effect of hy- feature that has been predicted at low temperature and
drophobic confinement on the thermodynamics and dy- high pressure for bulk water by several theories and mod-
namics of supercooled water. To this goal, we extend els.
a coarse-grained model previously developed for the case
of a water monolayer in hydrophobic nanoconfinement to ∗
[Link]@[Link]

Junio de 2011, Barcelona Panel P–53

108 XVII Congreso de Fı́sica Estadı́stica FisEs11

Noise-tolerant signal detection in genetic circuits

Javier Estrada, Raú