Professional Documents
Culture Documents
Textbook Jamming and Rheology Constrained Dynamics On Microscopic and Macroscopic Scales 1St Edition Liu Ebook All Chapter PDF
Textbook Jamming and Rheology Constrained Dynamics On Microscopic and Macroscopic Scales 1St Edition Liu Ebook All Chapter PDF
https://textbookfull.com/product/stochastic-control-and-
filtering-over-constrained-communication-networks-qinyuan-liu/
https://textbookfull.com/product/technology-of-dispersed-systems-
and-materials-physicochemical-dynamics-of-structure-formation-
and-rheology-1st-edition-naum-b-uriev/
https://textbookfull.com/product/theories-of-molecular-reaction-
dynamics-the-microscopic-foundation-of-chemical-kinetics-2nd-
edition-niels-e-henriksen/
https://textbookfull.com/product/multivariable-dynamic-calculus-
on-time-scales-1st-edition-martin-bohner/
Knowledge Proof and Dynamics The Fourth Asian Workshop
on Philosophical Logic Logic in Asia Studia Logica
Library 1st Edition Fenrong Liu (Editor)
https://textbookfull.com/product/knowledge-proof-and-dynamics-
the-fourth-asian-workshop-on-philosophical-logic-logic-in-asia-
studia-logica-library-1st-edition-fenrong-liu-editor/
https://textbookfull.com/product/fractional-dynamic-calculus-and-
fractional-dynamic-equations-on-time-scales-svetlin-g-georgiev/
https://textbookfull.com/product/advances-in-food-rheology-and-
its-applications-1st-edition-jasim-ahmed/
https://textbookfull.com/product/advances-in-food-rheology-and-
its-applications-1st-edition-jasim-ahmed-2/
https://textbookfull.com/product/rheology-concepts-methods-and-
applications-3rd-edition-alexander-ya-malkin/
JAMMING AND RHEOLOGY
JAMMING AND RHEOLOGY
Edited by
Andrea ]. Liu
Sidney R. Nagel
This book contains information obtained &om authentic and highly regarded sources. Reasonable efforts have
been made to publish reliable data and information, but the author and publisher cannot assume responsibility
for the validity of all materials or the consequences of their use. The authors and publishers have attempted to
trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if
permission to publish in this form has not been obtained. If any copyright material has not been acknowledged
please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted,
or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying, microfilming, and recording, or in any information storage or retrieval system, without
written permission &om the publishers.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only
for identification and explanation without intent to infringe.
Publisher's Note
This book was prepared from camera-ready copy supplied by the editors.
Overviews 7
The viscous slowing down of supercooled liquids and the glass transition:
phenomenology, concepts, and models
G. TARJUS AND D. KIVELSON 20
Statistical physics of the jamming transition: the search for simple models
S. F. EDWARDS AND D.V. GRINEV 80
PART 1
Force chains and fluctuations 95
vii
CONTENTS
PART 2
Force propagation in a granular pile 131
PART 3
Elasticity and inhomogeneous response to stress of dense amorphous
packings 165
viii
CONTENTS
PART 4
Experimental manifestations of mode coupling 219
PART 5
Glass transition: signatures and models 299
Scaling concepts for the dynamics of viscous liquids near an ideal glassy state
T.R. KIRKPATRICK, D. THIRUMALAI AND P.G. WOLYNES 301
PART 6
Inhomogeneities near jamming 349
Ē
CONTENTS
PART 7
Unsteady rheology 385
PART 8
Effective temperature 441
Theory of powders
S.F. EDWARDS AND R.B.S. OAKESHOTT 443
X
CONTENTS
Energy flow, partial equilibration and effective temperatures in systems with slow
dynamics
L.F. CUGLIANDOLO, J. KURCHAN AND L. PELITI 473
PART 9
Traffic 507
Xİ
ACKNOWLEDGEMENTS
Figure acknowledgements
“Jamming and the Statics of G ranular M aterials” (Dov Levine).
Figures 2 and 5 American Physical Society. Figure 4 American Association for the Advancement of
Science.
“ Rheology, and how to stop aging” (Jorge Kurchan)
Figure 1 American Physical Society and Elsevier Science. Figures 2 and 5 EDP Sciences and Elsevier
Science.
“The Viscous Slowing Down of Supercooled Liquids and the Glass Transition: Phenomenology,
Concepts, and M odels” (Gilles Tarjus and Daniel Kivelson).
Figures 3 and 6 American Physical Society. Figure 5 EDP Sciences. Figures 7 and 10 Elsevier Science.
Figure 8 American Institute of Physics. Figure 13 Academic Press. Reproduced with permission.
“Jamming, Friction and Unsteady Rheology” (M ark O. Robbins).
Figure 7 American Association for the Advancement of Science. Figure 8 American Chemical Society.
Reproduced with permission.
Reprint acknowledgements
“ Model for force fluctuations in bead packs,” S. N. Coppersmith, C.-h. Liu, S. M ajumdar, O.
N arayan, and T. A. Witten, Phys. Rev. E 53, 4673^1685 (1996). Copyright 1996 by the American
Physical Society.
“ Force Distributions in Dense Two-Dimensional G ranular Systems,” F. Radjai, M. Jean, J.-J. Moreau
and S. Roux, Phys. Rev. Lett. 77, 274-277 (1996). Copyright 1996 by the American Physical Society.
“ Force Transmission in G ranular M edia,” C. Thornton, KONA 15, 81-90 (1997). Copyright 1997
Hosokawa M icron International Inc. Reproduced with permission.
“ Force Distribution in a G ranular M edium ,” D. M. M ueth, H. M. Jaeger, and S. R. Nagel, Phys. Rev. E
57, 3164-3169 (1998). Copyright 1998 by the American Physical Society.
“Jamming and Static Stress Transmission in G ranular M aterials,” M. E. Cates, J. P. Wittmer, J.-P.
Bouchaud and P. Claudin, Chaos 9, 511-522 (1999). Copyright 1999, American Institute of Physics.
Reproduced with permission.
“ Statistical Mechanics of Stress Transmission in Disordered G ranular Arrays,” S. F. Edwards and D. V.
Grinev, Phys. Rev. Lett. 82, 5397-5400 (1999). Copyright 1999 by the American Physical Society.
“ Elastoplastic arching in two dimensional granular heaps,” F. Cantelaube, A. K. Didwania and J. P.
G oddard, in Physics o f Dry G ranular Media, ed. H. J. Herrmann, et al. (Kluwer Academic
Publishers, Amsterdam, 1998), 123-128. Reproduced with kind permission of Kluwer Academic
Publishers.
“ Stress propagation through frictionless granular m aterial,” A. V. Tkachenko and T. A. Witten, Phys.
Rev. E 60, 687-696 (1999). Copyright 1999 by the American Physical Society.
“ Rigidity loss transition in a disordered 2D froth,” F. Bolton and D. Weaire, Phys. Rev. Lett. 65, 3449-
3451 (1990). Copyright 1990 by the American Physical Society.
xiii
ACKNOWLEDGEMENTS
‘O sm otic pressure and viscoelastic shear moduli of concentrated em ulsions/’ T. G. Mason, M.-D.
bacasse, G. S. Grest, and D. Levine, J. Bibette and D. A. Weitz, Phys. Rev. E 56, 3150-3166 (1997).
Copyright by the American Physical Society.
“ Universality and its Origins at the Am orphous Solidification Transition,” W. Q. Peng, H. E. Castillo,
P. M. G oldbart, and A. Zippelius, Phys. Rev. В 57, 839-847 (1998). Copyright by the American
Physical Society.
“ Anomalous viscous loss in emulsions,” A. J. Liu, S. Ramaswamy, T. G. M ason, H. Gang, and D. A.
Weitz, Phys. Rev. Lett. 76, 3017-3020 (1995). Copyright 1995 by the American Physical Society.
“ Yielding and Rearrangements in Disordered Emulsions,” P. Hebraud, F. Lequeux, J. P. Munch, and
D. J. Pine, Phys. Rev. Lett. 78, 4657^1660 (1997). Copyright 1997 by the American Physical Society.
“ Dynamics of viscoplastic deform ation in am orphous solids,” M. Falk and J. S. Langer, Phys. Rev. E
57, 7192-7205 (1998). Copyright 1998 by the American Physical Society.
“ The Mode Coupling Theory of Structural Relaxations,” W. Götze and L. Sjogren, Transport Theory
and Statistical Physics 24, 801-853 (1995). Reproduced with permission of Marcel Dekker, Inc.
“ Light scattering spectroscopy of orthoterphenyl-idealized and extended mode coupling analysis,” H. Z.
Cummins, G. Li, W. M. Du, Y. H. Hwang, and G.Q. Shen, Prog. Theor. Phys. S126, 21-34 (1997).
Copyright 1997, The Progress of Theoretical Physics. Reproduced with permission.
“ Glass transition in colloidal hard spheres: mode-coupling theory analysis,” W. Van Megen and S. M.
Underwood, Phys. Rev. Lett. 70, 2766-2769 (1993). Copyright 1993 by the American Physical
Society.
“ Emulsion glasses: a dynamic light-scattering study,” H. Gang, A. H. Krall, H. Z. Cummins, and D. A.
Weitz, Phys. Rev. E. 59, 715-721 (1999). Copyright 1999 by the American Physical Society.
“ Scaling concepts for the dynamics of viscous liquids near an ideal glassy state,” T. R. Kirkpatrick, D.
Thirumalai, and P. G. Wolynes, Phys. Rev. A 40, 1045-1054 (1989). Copyright 1989 by the American
Physical Society.
“ A Topographic View of Supercooled Liquids and Glass Form ation,” F. H. Stillinger, Science 267,
1935-1939 (1995). Copyright 1995 American Association for the Advancement of Science. Reprinted
with permission.
“A thermodynamic theory of supercooled liquids,” D. Kivelson, S. A. Kivelson, X. Zhao, Z. Nussinov,
and G. Tarjus, Physica A219, 27-38 (1995). Reproduced with permission of Elsevier Science Ltd.
“ Long-Lived Structures in Fragile Glass-Forming Liquids,” A. I. Melcuk, R. A. Ramos, H. Gould, W.
Klein, and R. D. M ountain, Phys. Rev. Lett. 75, 2522-2525 (1995). Copyright 1995 by the American
Physical Society.
“On a Dynamical Model of Glasses,” J. P. Bouchaud, A. Comtet, and C. M onthus, J. Phys. (France) I,
5, 1521-1526 (1995). Copyright 1995 by Springer-Verlag.
“ High-Frequency Asymptotic Shape of the Primary Relaxation in Supercooled Liquids,” R. L. Leheny
and S. R. Nagel, Europhys. Lett. 39, 447^152 (1997). Reproduced with permission of EDP Sciences.
“ Ergodicity Breaking Transition and High-Frequency Response in a Simple Free-Energy Landscape,”
M. Ignatiev and B. Chakraborty, Phys. Rev. E. 60, R21-R24 (1999). Copyright 1999 by the American
Physical Society.
“ Dynamic heterogeneity in ortho-terphenyl studied by multidimensional deuteron N M R ,” R. Bohmer,
G. Hinze, G. Diezemann, B. Geil, H. Sillescu, Europhys. Lett. 36, 55-60 (1996). Reproduced with
permission of EDP Sciences.
“ How Long Do Regions of Different Dynamics Persist in Supercooled o-terphenyl” C. Y. W ang and M.
D. Ediger, J. Phys. Chem. В 103, 4177-4184 (1999). Copyright 1996 American Chemical Society.
Reprinted with permission.
“ Origin of the difference in the tem perature dependences of diffusion and structural relaxation in a
supercooled liquid,” D. N. Perera and P. Harrowell, Phys. Rev. Lett. 81, 120-123 (1998). Copyright
1998 by the American Physical Society.
“ Dynamics of highly supercooled liquids: heterogeneity, rheology and diffusion,” R. Yamamoto and A.
Onuki, Phys. Rev. E 58, 3515-3529 (1998). Copyright 1998 by the American Physical Society.
“ Liquid to Solidlike Transitions of Molecularly Thin Films Under Shear,” M. L. Gee, P. M.
McGuiggan, J. N. Israelachvili and A. M. Homola, J. Chem. Phys., 93, 1895-1906, (1990). Copyright
1990, American Institute of Physics. Reproduced with permission.
“ Structure and Shear Response in N anom eter Thick Films,” P. A. Thompson, M. O. Robbins and G. S.
XIV
ACKNOWLEDGEMENTS
Grest, Israel Journal of Chemistry 35, 93-106 (1995). Copyright 1995 Laser Pages Publishing Ltd.
Reproduced with permission.
“Glasslike Transition of a Confined Simple Fluid,” A. L. Demirei and S. Granick, Phys. Rev. Lett. 77,
2261-2264 (1996). Copyright 1996 by the American Physical Society.
“ Nonlinear Bubble Dynamics in a Slowly Driven Foam ,” A. D. Gopal and D. J. Durian, Phys. Rev.
Lett. 75, 2610-2613 (1995). Copyright 1995 by the American Physical Society.
“ Kinetic Theory of Jamming in Hard-Sphere Startup Flows,” R. S. Farr, J. R. Melrose and R. C. Ball,
Phys. Rev. E 55, 7203-7211 (1997). Copyright 1997 by the American Physical Society.
“ Friction in G ranular Layers: Hysteresis and Precursors,” S. Nasuno, A. Kudrolli and J. P. Gollub,
Phys. Rev. Lett. 79, 949-952 (1997). Copyright 1997 by the American Physical Society.
“ Kinematics of a two-dimensional granular Couette experiment at the transition to shearing,” C. T.
Veje, D. W. Howell, and R. P. Behringer, Phys. Rev. E 59, 739-745 (1999). Copyright 1999 by the
American Physical Society.
“Theory of Powders,” S. F. Edwards and R. B. S. Oakeshott, Physica A 157, 1080-1090 (1989).
Reproduced with permission of Elsevier Science Ltd.
“ Density Fluctuations in Vibrated G ranular M aterials,” E. R. Nowak, J. B. Knight, E. Ben-Naim, H.
M. Jaeger, and S. R. Nagel, Phys. Rev. E 57, 1971-1982 (1998). Copyright 1998 by the American
Physical Society.
“ Diffusing-Wave Spectroscopy o f Dynamics in a Three-Dimensional G ranular Flow,” N. Menon and
D. J. D urian, Science 275, 1920-1922 (1997). Copyright 1997 American Association for the
Advancement of Science. Reprinted with permission.
“ Rheology o f Soft Glassy M aterials,” P. Sollich, F. Lequeux, P. Hebraud, and M. E. Cates, Phys. Rev.
Lett. 78, 2020-2023 (1997). Copyright 1997 by the American Physical Society.
“ Energy flow, partial equilibrium and effective temperatures in systems with slow dynamics,” L. F.
Cugliandolo, J. K urchan and L. Peliti, Phys. Rev. E 55, 3898-3914 (1997). Copyright 1997 by the
American Physical Society.
“ Frequency-domain study of physical aging in a simple liquid,” R. L. Leheny and S. R. Nagel, Phys.
Rev. В 57, 5154-5162 (1998). Copyright 1998 by the American Physical Society.
“ Sheared Foam as a Supercooled Liquid?” S. A. Langer and A. J. Liu, Europhysics Lett. 49, 68-74
(2000). Reproduced with permission of EDP Sciences.
“An Exact Solution of a One-Dimensional Asymmetric Exclusion Model with Open Boundaries,” B.
Derrida, E. Domany, and D. Mukamel, Journal of Statistical Physics, 69, Nos. 3-4, 667-687 (1992).
Reproduced with permission of Plenum Publishing Corporation.
“Jamming and Kinetics of Deposition-Evaporation Systems and Associated Quantum Spin Models,”
M. Barma, M.D. Grynberg, and R. B. Stinchcombe, Phys. Rev. Lett. 70, 1033-1036 (1993).
Copyright 1993 by the American Physical Society.
“ Slow Relaxation in a Model with M any Conservation Laws Deposition and Evaporation of Trimers on
a Line,” M. Barma, D. Dhar, Phys. Rev. Lett., 73, 2135-2138 (1994). Copyright 1994 by the
American Physical Society.
“ Spontaneous Symmetry-Breaking in a One-Dimensional Driven Diffusive System,” M. R. Evans, D. P.
Foster, C. Godreche, and D. Mukamel, Phys. Rev. Lett. 74, 208-211 (1995). Copyright 1995 by the
American Physical Society.
“ Spontaneous jam m ing in one-dimensional systems” , O. J. O ’Loan, M. R. Evans, and M. E. Cates,
Europhys. Lett. 42, 137-142 (1998). Reproduced with permission of EDP Sciences.
“ Self-Organization and a Dynamic Transition in Traffic-Flow M odels,” O. Biham, A. A. Middleton, D.
Levine, Phys. Rev. A 46, R6124-R6127 (1992). Copyright 1992 by the American Physical Society.
XV
Jamming and Rheology: An Introduction
Andrea J. Liu
D ept of Chemistry and Biochemistry, University o f California, Los Angeles, CA 90095
Sidney R. Nagel
James Franck Institute, The University o f Chicago, Chicago, IL 60637
Things get jammed. Traveling on a highway we tion or to jamming transitions; the relaxation time
get caught in traffic jams. We are frustrated at the diverges near a second-order phase transition, as
whole-foods supermarket when grains and beans well. However, for such a transition, there is a
jam in the process of flowing out of the hopper into structural length scale, the correlation length, which
our bags. On a larger scale, factories using powders diverges. Because long length scale perturbations
as the raw materials are plagued by the jamming relax more slowly, the diverging correlation length
that occurs as the powders become clogged in the gives rise to a diverging relaxation time. Near the
conduits that were designed to have them flow glass transition, by contrast, no structural correla
smoothly from one side of the factory floor to the tion length has yet been found and it is not clear if
other. there is any length scale that can be used to char
In these systems, one can see the structural ar acterize the growth of the glassy response.
rest of macroscopic particles. Vibrations are neces Is there really a meaningful connection among
sary to induce any motion in materials such as these different jamming and glass transitions? This
granular media, since thermal energy is insufficient is still an open question, but at first glance the an
to lift a particle by its own diameter. As the vibra swer appears to be “No” because the jamming and
tion intensity is lowered, these materials undergo a glassy phenomena apparently have very different
transition from a state reminiscent of a fluid or gas physical origins. In supercooled liquids, the in
' to one more akin to a solid with a yield stress. crease in the relaxation time with decreasing tem
In colloidal suspensions of smaller particles, perature clearly has a thermal origin. In colloidal
such jamming also occurs when the pressure or suspensions, increase in relaxation time with in
density is sufficiently large. Above this transition, a creasing density has an entropie origin. Finally, in
colloidal suspension is a disordered solid with a athermal systems such as Bingham plastics (foams
yield stress. A dense suspension of deformable and emulsions) or granular systems of macroscopic
particles such as an emulsion or foam is an example solid particles, the increase in the relaxation time
of a class of soft amorphous solids known as Bing appears to have a purely kinetic origin.
ham or Herschel-Bulkeley plastics, which flow In spite of these differences, however, there are
under high shear stress. When the applied shear striking similarities in the phenomenology of these
stress is lowered below the yield stress, these sys different systems. For example, the dynamics of
tems stop flowing. Thus, foams and emulsions can colloids near the colloidal glass transition have
unjam when either the pressure (or density) is low many characteristics in common with the dynamics
ered, or the shear stress is raised. of glass-forming liquids and the same theoretical
Even molecular systems can become stuck in a framework has been used to analyze the dynamical
part of phase space that does not support macro data from experiments on all these different sys
scopic motion. The prototypical example is a liquid tems. [See the papers in the section: Experimental
that ordinarily flows readily but that, as the tem Manifestations o f Mode Coupling.] In addition, the
perature is lowered, turns into a rigid glass - an dynamics appear to become spatially inhomogene
amorphous solid with a yield stress. One of the ous [see papers in the section: Inhomogeneities
oldest unsolved problems in condensed matter near Jamming] and temporally intermittent [see
physics is the nature of this glass transition [1]. As papers in the section: Unsteady Rheology] near the
the temperature is lowered, the stress relaxation jamming or glass transition in all systems. The
time increases dramatically. Rapidly increasing stress response in the jammed or glassy state also
relaxation times are not unique to the glass transi appears to be inhomogeneous in many different
1
ANDREA J. LIU AND SYDNEY R. NAGEL
2
JAMMING AND RHEOLOGY: AN INTRODUCTION
(Indeed recent data from David Weitz and his Fluctuations, temperature and the fluctua
group indicate that the shape might be convex.) It is tion-dissipation theorem
also very likely that the shape of the surface is not
always simple. For a shear-thickening material like One question that appears over and over again in
corn starch in water, one would expect reentrant all out-of-equilibrium systems is whether there is a
behavior: there would be a liquid at low shear useful concept of temperature. [See the papers in
stress, a jammed state at the same density at high the section: Effective Temperature.] This appears
shear rates (high shear stresses), and a liquid again naturally in the subject of supercooled liquids
at even higher shear stress when the jammed state where the term “ficti ve temperature” has long been
breaks up. used to describe the characteristic out-of
There are several benefits of having such a dia equilibrium thermal characteristics of the system.
gram. On the one hand, it suggests a number of More recently, the question has been addressed as
different experiments. For example, there is virtu to whether there exists for these systems a fluctua
ally no study of how the introduction of a tem tion-dissipation theorem. The existence of such a
perature (i.e., kinetic energy) can affect the possi theorem relies on there being a temperature. Gener
bility that a system will jam at a given stress. In a alizations of such theorems have been produced for
typical foam, the thermal energy is a million times supercooled systems although there have been no
too small to cause rearrangements of bubbles. experimental confirmations of the predictions.
However, if one were able to increase the thermal There have been other attempts to define a “tem
energy by a factor of a million without destroying perature” for athermal systems such as granular
the foam, the phase diagram predicts that the yield materials, foams and colloids. One question is
stress would decrease. Likewise, the introduction of whether the fluctuations in a driven steady-state
vibrations in a granular material can lead to a low system that is far from equilibrium can be used to
ered yield stress. In a similar vein, there is little define a temperature. Are all the temperatures that
known, especially in simple supercooled liquids, one defines in this way consistent with one an
about how the introduction of a shear stress can other? Is there a zeroeth law of thermodynamics
influence the glass transition temperature. The (i.e., is there a notion of equilibrium)? For static
phase diagram predicts that if one could apply a systems in some metastable state, can one define a
large shear stress to a supercooled liquid (compara quantity similar to temperature which describes the
ble to the thermal energy at the glass transition state of the system?
divided by the molecular volume), the glass transi
tion would shift to lower temperatures. Experi Force chains in jammed systems
ments that elucidate such effects will be important
for gaining a better understanding of both the jam When a system becomes jammed, it is held rigid
ming and the glass transitions. due to propagation of forces from one boundary
Another benefit of having such a phase diagram surface to another. One common feature of many
is that it suggests that the jamming and the glass jammed systems is that this propagation of forces is
transitions may be linked at an underlying level. very heterogeneous. [See papers in the section:
That is, the same processes that are important for Force Chains and Fluctuations. ] Thus, for exam
jamming a material under an applied stress may be ple, in a granular system one can actually see the
important for arresting the dynamics of a liquid as “force chains” by placing the stressed material be
the temperature is lowered. tween two crossed circular polarizers. Simulations
There are several categories of questions that of granular materials under a variety of loads also
can be asked about jamming, (i) What is the nature clearly show this heterogeneous force propagation.
of the jammed state? (ii) How does one get into the These fluctuations in the forces can be very large -
jammed state? (iii) Are there similarities between as large as the average force itself - and a high
jamming in driven, athermal systems and jamming fraction of the contacts between particles carry
in quiescent thermal systems (such as the glass almost no force. Experimental measurements and
transition in molecular liquids)? In this volume we numerous computer simulations seem to indicate
have collected a series of articles that have ad that the distribution of the forces is robust - that is
dressed what we think are some of the important it varies only slightly between one system and
issues in the fields of jamming and glassy dynam preparation condition and another. Several ques
ics. We have organized them loosely into topics, tions naturally arise. Are these fluctuations remi
but many of the papers could be placed under more niscent of a temperature? How does one model
than one topic. these fluctuations? Clearly there is disorder, but it
is not clear how one puts disorder into a theory. In
3
ANDREA J. LIU AND SYDNEY R. NAGEL
particular, it is not apparent how to model the dis mentally justified? The issue is still a subject of
order: should one assume complete randomness or heated debate.
should one build in correlations? (In a thermal sys
tem, by contrast, there is an ergodic assumption Friction in confined systems
that all accessible states are equally possible.) Most
of the studies of “force chains” have been on When a fluid is confined in a narrow region, a
granular systems. Are similar force heterogeneities small stress may not be sufficient to force the sys
seen in other jammed systems such as foams and tem to move: there is a yield stress. For systems
molecular glasses? Ideally, one could develop a driven just above the yield stress, there is
comprehensive theory that would describe how the “stick/slip” motion where the fluid flows in short
force distributions vary from one system to another. bursts. [See the papers in the section: Unsteady
For example, why is the distribution of forces ex Rheology.] As the force on the driving spring is
ponential at large forces in a granular material but increased still further, the motion becomes more
Gaussian for a foam? and more smooth. Such behavior is seen in many
other systems such as granular materials (in ava
Force propagation in granular materials lanches as well as in forced flow between plates),
foams, and colloids, and goes by different names
In addition to asking how to treat force hetero depending on the context. On the largest scale, this
geneities, one can ask how to treat the average stick/slip motion is felt in earthquakes. The relation
propagation of forces in a granular material. [See to jamming is obvious: the flow only 'stops because
the papers in the section: Force Propagation in a the particles get jammed and new force chains are
Granular Pile.] Is it appropriate to use conven set up that span the entire system. Questions that
tional theories of elasticity or elasto-plasticity to have been asked are: What is the nature of the rear
treat these very hard but delicately balanced mate rangement events during one of these processes?
rials? Recently, in the literature, there has been an Can the stick/slip event be predicted from other
assault on this approach in which a different set of external observations of the system? How do the
constitutive relations have been proposed to close stick/slip events change their character as the stress
the equations. These new equations are not based is increased? Recently, experiments and simula
on elastic behavior but rather are based solely on tions have been able to observe the nature of the
relations between various components of the stress individual “avalanche” events. It appears that the
Censor in the material. One question that has been size of these events does not diverge as the stress
investigated at length is whether the new ap decreases towards the yield stress.
proaches allow one to explain experimental results
that could not be understood by the more conven Glass transition
tional theories. That is, are the two approaches
really fundamentally different? Another question is The nature of the glass transition is one of the
what is the nature of the plastic region in the more oldest unsolved questions in condensed matter
conventional theories and what experimental impli physics. [See the papers in the sectioa* Glass Tran
cations does the plastic region have? In relating this sition: Signatures and Models.] In a fundamental
topic to the previous one on force inhomogeneities, sense this is a jamming phenomenon: as the tem
one is also led to ask whether plasticity is the con perature of a liquid is lowered, it becomes more
sequence of a broad spectrum of force fluctuations. sluggish. This jamming is an equilibrium phe
In particular, many observed force distributions nomenon and is caused by the decrease in thermal
show that there is a high fraction of particle pairs energy available for molecular rearrangements.
that have almost no force between them. If so many Thus, as distinct from many of the cases discussed
contacts are on the verge of breaking, does this above, it is not caused by the kinetics of driving the
imply that the assembly is particularly sensitive and system out of equilibrium. Despite this difference,
that ordinary elastic response theory might fail? A are there similarities between these jamming phe
technical issue is whether the equations that govern nomena? Certainly many of the questions asked are
the static stress patterns are elliptic or hyperbolic. similar: What notion of temperature can be used to
The more recent theories assert that the equations describe the system as soon as it has left equilib
are hyperbolic so that stresses will propagate along rium? Is there a fluctuation-dissipation theorem
lines like a wave. This has the consequence that the below the glass transition temperature? As the glass
stress needs only to be specified along a (e.g., top) transition temperature is approached, does one be
surface and the stress then propagates (downward) come increasingly more sensitive to kinetic jam
due to those applied forces. Is this scenario experi ming due to shear or other forcing of the system?
4
JAMMING AND RHEOLOGY: AN INTRODUCTION
Many experiments appear to indicate that super The idea for this volume grew out of the pro
cooled liquids are very heterogeneous as are the gram on Jamming and Rheology that was held at
other jammed systems (e.g., force chains, inelastic the Institute for Theoretical Physics at Santa Bar
collapse). Several recent experiments indicate that bara in the Autumn of 1997. The conference that
the time scales for translational diffusion in liquids was held as part of the program had a somewhat
decouple from those important for rotational re unusual feature. Because the subject matter for the
laxation and viscosity. Are there different “tem conference spanned many different fields, no one
peratures” that are important as soon as the system could be expected to be an expert in all relevant
begins to jam that govern the different properties? specialties. For that reason the organizers planned a
Recent experimental results have also indicated that series of tutorial talks in conjunction with the re
structural glasses have a great deal more in com search talks in the different areas. These overviews
mon with spin glasses than had previously been were designed to give a survey of a particular area
suspected. There is a tail, extending to high fre and indicate how the work in that subfield related
quencies, that appears in the susceptibility of both to the topic of jamming in the wider context of the
systems (as well as in plastic crystals). This ubiq workshop. We have used that same idea in this
uitous feature is not understood although it may volume. At the outset we have a series of overview
imply a definite correspondence between the phys papers, written specifically for this book, that were
ics of these systems. meant to provide a broader review of the field. The
second part contains a series of reprints that cover
Kinetic jamming and rheology many of the topics that we consider to be the most
relevant in a discussion of jamming. These papers
Glassy systems have a peculiar rheology. How are divided into sections dealing with individual
does one account for a yield stress and how does topics.
one model the response of the system above the We thank all of the staff at the ITP for their
yield threshold? How does the shear modulus de generous hospitality. In particular we would like to
pend on frequency as the frequency is lowered? thank David Gross, Director of the Institute, who
[See the papers in the section: Elasticity and Inho first suggested this volume. We particularly thank
mogeneous Response to Stress o f Dense Amor Deborah Storm for helping us to get the overview
phous Packings.] Many soft systems (foams, col talks transcribed. We also thank her and Daniel
loids etc.) have a very non-Maxwellian frequency Hone for helping to arrange the details of the initial
dependence with both real and imaginary parts of program. We thank Judy Sweeney at UCLA for all
the modulus varying as a fractional power of the her help in putting this volume together and track
frequency as the frequency approaches zero. Mo ing down all the loose ends. We are extremely
lecular glass-forming liquids do not have this char grateful to Tom Witten for his encouragement and
acteristic as far as we can tell. Can this anomalous support of the entire project from its inception.
frequency dependence be understood in terms of Without him, it is doubtful that there would have
some simple model? Several attempts at construc been a workshop at all. Finally we would like to
tion such a model have been tried. Some of these thank our co-organizers of the workshop, Sam Ed
focus on the microscopies at the level of the indi wards and Mark Robbins, who helped in so many
vidual bubbles or particles and the forces con ways to give substance and shape to this program.
straining them. Can one relate the weak regions that
occur in these systems to the nature of the force References
distributions that were discussed above or is the
nature of the weak regions most accurately de [1] P.W. Anderson, Science 267, 1615 (1995). “The deep
scribed by a geometric effect? Other theories oper est and most interesting unsolved problem in solid state
ate on a more phenomenological, coarse-grained theory is probably the theory of the nature of glass and
the glass transition. This could be the next breakthrough
level; energy can flow between different cells ac
in the coming decade. The solution of the problem of
cording to a specified set of rules. Both kinds of spin glass in the late 1970s had broad implications in
theories can produce non-Maxwellian behavior. unexpected fields like neural networks, computer algo
Another whole class of theories, is based on jam rithms, evolution, and computational complexity. The
ming due to essentially free-volume effects such as solution of the more important and puzzling glass
the random sequential deposition model. These problem may also have a substantial intellectual spin
models can be generalized to treat other dynamical off. Whether it will help make better glass is question
systems such as traffic flow problems. [See the able.”
papers in the section: Traffic.] [2] A. J. Liu and S. R. Nagel, “Jamming is not just cool any
more,” Nature 396, 21-22 (1998).
5
OVERVIEWS
Jam m in g and the S tatics of G ranular M aterials
Dov Levine
Department of Physics, The Technion, Haifa, 32000, Israel
The effect of jamming on the static properties of granular materials is discussed. Various models
for stress propagation, such as elasto-plasticity, isostaticity, and fragility are examined critically.
Analogy is made to (frictionless) emulsion rheology, and the effect of friction on the results is
considered. The relevance of the specifics of the construction history is emphasized.
9
DOV LEVINE
grain forces for which all the grains are in mechanical traditional way of doing this is to introduce a strain field
equilibrium. The reason for this is that the number of (see below) and relating it to the stress. These constitu
variables in the system, for example, the number of in tive relations, as they are traditionally called, are meant
tergrain forces, may well be larger than the number of to reflect the physics of the system under examination,
(mechanical) equilibrium conditions, being the equations and may be derived, in principle, from microscopic con
of force and torque balance on each grain. Such ideas siderations. Thus, gels, wood, plastics, and metals, all
are not new, but they have been recently applied in the of which satisfy Eq. (1), each has its own distinctive
context of packings [8-13]. Briefly stated, the argument constitutive relation.
runs as follows: Any given grain in a d dimensional pack As stated above, it is often useful to introduce a dis
ing has, if it is not perfectly round, d translational de placement field, Щ, (where і indexes the components of
grees of freedom (these correspond to d equations of force the vector u) which gives the displacement of a small vol
balance) and d{d~l) rotational degrees of freedom (corre ume element of material relative to a reference state. In
sponding to d(d~l) equations of torque balance). Thus, traditional elasticity theory, one begins from a reference
state which is stress-free [16], and assumes that the free
for each grain, there are d+ equations constraining
energy F of the system may be expanded to quadratic
the possible forces between the grains [14]. If the forces
order in the strain field Uij [17]. This is simply the gen
are not simply central [15], but are nontrivial vectors (as
eralization of the Hooke law for a simple spring, where
is the case if there is friction), then, per grain, there are
the energy of a displacement x is given by kx2, where k
d žļ2 (scalar) components of intergrain force, where ž is
is the spring constant. In an elastic solid, k is replaced
the average number of contacts that a grain has. Thus,
by a tensor relating stress and strain, reflecting the ten-
in order for the system to be perfectly defined, that is,
sorial nature of the deformation. In principle, this tensor
to have a unique equilibrium solution, the number of un
can vary in space, in which case it is a tensor field, which
knowns (the components of the forces) must equal the
we shall call, for want of a better name, the elasticity
number of constraint equations, giving z — d + 1. A
field [18]. Since the stress and the strain are conjugate
system which has exactly one such solution is called iso
variables,
static. If ž < d + 1, then there are too few variables: the
system is overconstrained, and in general there is no so
lution. If z > d -f 1, there are fewer constraint equations ( 2)
than unknowns, and there is a multitude of solutions
consistent with the packing. What this means is that, where T is the temperature of the system, which for us,
in the last case, for a given (underdetermined) packing, as stated earlier, is effectively zero. Eq. (2) represents
there are many different stress patterns which are consis relations between σ,-ј and Uįj, with a,j as unknown
tent with the packing being in mechanical equilibrium. quantities. However, since uų is derived from a vector
Note that we have tacitly assumed throughout that all
field Щ, being its derivative, we have d+ d(d+li unknowns
the constraint equations are linearly independent, which
(that is, Щ and σ,-j), and the same number of governing
is not always the case, though for a disordered packing
equations, Eqs. (1) and (2). Provided that the applied
of variously shaped grains this is probably true. Note
stress is not so large as to impart irreversible (plastic)
too the interesting special case of grains which interact
deformations, elasticity theory describes most solids ex-
solely via normal forces. In this case, if the grains are
temely well [19]. However, for an isostatic system, there
obliquely shaped, then the critical coordination number
are no additional relations necessary, since the system is
is z — d(d -f 1), while if the grains are all spherical,then
well-defined and fully determined by force balance alone,
z — 2d.
and so an isostatic system need not obey elasticity theory.
Another way to understand static indeterminacy is
from the continuum equations defining the stress ten It has recently been proposed [20] that, for granular
sor crti7(r). For static systems, we have d force balance materials, in lieu of the aforementioned procedure of elas
ticity theory, there may exist local, linear relations be
equations
tween components of the stress tensor, which may be
(1) written (in two dimensions)
where fi is the body force (e.g. gravity) acting on the (3)
system. These equations hold for any static system for
which a continuum picture applies. On the other hand, (additional relations are needed in three dimensions). In
there are independent components of the stress particular, for a sand pile constructed from a point source,
tensor (by virtue of its being symmetric), leaving d(d~H it is argued that the principal axes of the stress tensor of a
components undetermined. One now needs to introduce given grain [22] are frozen at the moment it stops moving,
this number of relations among the stress components in and that the addition of susbsequent grains will not af
order to determine the state of stress of the system. The fect this. This approach has been criticized as simplistic
10
JAMMING AND THE STATICS OF GRANULAR MATERIALS
by Savage [23], but it has one interesting defining fea ular material). This question arises because of the fas
ture: Stress is propagated along linear paths (See Fig. 1) cinating feature of many sandpiles [29,30] that directly
related to the characteristics of the differential equation beneath the apex, there is a pressure dip, contrary to
for the stress obtained by supplementing Eq. (1) with normal intuition. One may now inquire whether this
the linear closure Eq. (3). This equation is hyperbolic in pressure distribution, at the base of the pile, can be cal
nature, like a wave equation, and as such requires specifi culated. For the hyperbolic models, there is no problem -
cation of the stress on a portion of the boundary in order the boundary condition is specified at the free surface of
for a solution to exist. In a sense, such models are cousins the pile, and this (supplemented by a scaling hypothesis
of the simple stress propagation models [24-28] which en for the stress field) fully determines the stress distribu
visage that a grain passes stress down (in the direction of tion at the base. For certain choices of the vector a,j
gravity) to its neighbors beneath it. This stands in stark in Eq. (3), such a dip is found. On the other hand,
contrast to elastic models, where the differential equa if one treats the pile as an elastic medium (see the dis
tion for the stress field is elliptic, whose solution requires cussion in [31,23]) (analogous to, though different from,
the specification of data (stress or displacement field) on a pile-shaped piece of rubber or metal) then the ellip
the entire boundary of the material. When a weight is tic nature of the equations for the stress field require us
placed on a simple elastic material, the entire material to specify boundary conditions everywhere, in particu
deforms, not just the portion beneath the disturbance. lar at the base of the pile itself. If the stresses must
We note here (see also [21]) that there is a precedent for be prescribed, then the meaning is that they cannot be
such behavior: if an elastoplastic material is at yield (see calculated. One may instead specify the displacements
below and next section), then stresses propagate in it ac at the base; clearly, this requires a reference state (see
cording to a hyperbolic equation. The difference is that next section). It is unclear that such a reference state
in the aforementioned proposal, the local, linear relation is well-defined; certainly it is not a stress-free state as
is posited whether the material is at yield or not. is standard in elasticity theory. Moreover, for extremely
rigid grains, specification of displacements may be ill-
defined [32]. Whatever the resolution of this controversy,
attempting to treat the sandpile, or any other granular
system, as an elastoplastic material without regard to its
construction history seems misguided. It is precisely dur
ing construction that the system develops stress paths,
and in the elastic picture, this is when the elasticity field
is set up. Without knowledge of this field, the elastic
ity theory is not defined. One way of incorporating the
construction history is to solve the problem by adding
layer upon layer of material, at each stage solving the
elasto-plastic problem (say, by finite elements) assuming
a constitutive relation. The way in which new material
is added to the existing pile defines the construction his
tory. Such a study has been carried out by Savage [23],
who concludes that the dip in the pressure beneath the
apex is not inevitable, but depends on the nature of the
support upon which the pile sits. If the support is rigid,
no dip is obtained, while if the table is allowed to sag a
bit, a dip is apparant in the data. This debate should,
as all scientific debates should, be settled experimentally,
and careful experiments are essential.
FIG. 1. Schem atic depiction o f force propagation in a sand- The experimental results that do exist are fascinating.
pile w ith a linear closure as in Equation 3. Forces are prop In a precise experiment on a granular material in a long,
agated along the characteristics of a wave-like equation. To rather small diameter vertical cylinder [33,34] of diameter
solve for the stress distribution, it is necessary to specify a 3.8 cm (filled with 2 mm diameter glass balls), the appa
boundary condition at the free surface (the top) of the pile.
rant mass at the bottom of the column was measured as
a function of the total mass of material in the cylinder.
The nature of the constitutive or closure relations, and Since some of the weight is taken up by the sidewalls,
the attendant issue of boundary conditions has fueled the apparant mass M e is different (typically less, since
sharp debate, the principal question being whether one friction is usually mobilized upwards unless special mea
can or cannot, in principle, calculate the stresses at the sures are taken) from the total, actual mass Mt (which
base of a sandpile (or a differently shaped mass of gran is the weight of each grain times the number of grains).
11
DOV LEVINE
The simple, classical argument of Janssen [35,36] predicts formation. [39] A “better”, more stable packing may be
an initial linear dependence of M e for small Μχ, which obtained by “annealing” , by gentle agitation. Of course,
asymtotes to a value Мто for large Μχ. If a weight is the marginal state hypothesis is not necessarily correct,
placed on the top of the column (not touching the cylin and in the following I discuss some possible descriptions
der walls), then the above result is only slightly modified, of the statics of granular materials.
so as to give the correct pressure at the top layer of the There are several interesting possibilities for the state
granular material. In any case, the function Mj (M e ) is of such as-poured granular material. First, it may be
monotonie increasing, as intuition would suggest. It is isostatic, having exactly the correct number of contact
therefore a great surprise that for substantial overloads, points so as to define the forces uniquely [9-11]. Or,
an oscillatory dependence of M e on Μχ was observed, it may be fragile [40,41], meaning that it is able to sup
with a period which appears to be roughly the Janssen port only those external stresses which were imposed dur
decay length. For (at least) one of the models [20] em ing the construction process, and no others-these oth
ploying a linear closure of the form of Eq. 3, such os ers, even if infinitesimal, would generate irreversible re
cillatory behavior is predicted, and derives from the fact arrangements in the bulk. We might term this “strong
that in this model, stress is ’’reflected” off the sidewalls, fragility”. A slightly relaxed definition of fragility might
in wavelike fashion. It will be interesting to see whether allow for the support of a certain subset of external
elasto-plastic theories, incorporating construction histo stresses, perhaps close to those the system grew up with,
ries mimicking the experiment, can reproduce this un but the application of other, different external stresses
usual dependence. It seems clear that these inriguing would cause the system to rearrange irreversibly. An
results, supplimented by further experiments, will go a other possibility might be called “weak fragility” - this
long way in sorting out the various proposals. would be the ability to sustain all infinitesimal applied
stresses save one; that is, only if we push in a certain
way will the system rearrange irreversibly. If the state
II. ISOSTATICITY, FRAGILITY, AND generated was marginally stable, it would presumably
ELASTICITY fall into one of these categories. Another possibility,
the traditional view [31,23], is that the the system is,
The case of an isostatic system of spherical particles in fact, an elasto-plastic medium. An elasto-plastic ma
with central forces only is of particular interest since it is terial is one which possesses regions which behave ac
what is expected for an emulsion at critical volume frac cording to elasticity theory, until they are so stressed
tion Фс. An emulsion consists of liquid drops (say oil) that they reach a ’’yield criterion”, at which they deform
immersed in another liquid (say water); the drops be irreversibly. The accepted yield criterion for granular
ing stabilized by a surfactant (say detergent). Фс is the materials is the Coulomb criterion [42,36], which states
volume fraction at which droplets touch in point con that yield is reached when the maximum ratio of shear
tacts, with no droplet deformation. In the case of an stress to normal stress equals μ, the friction coefficient
emulsion whose volume fraction is slowly increased to Фс [43]. The basis for the elasto-plastic picture is that each
from the liquid state, there is ample time for the system individual grain is itself a solid body obeying the dic
to find a good local free energy minimum, though the tates of elasticity theory. This being said, it is clear that
global minimum (believed to be a close-packed crystal for stress is propagated in a granular medium via grain de
a monodisperse emulsion) is usually too far away to be formation, and so why should the material as a whole be
accessed. Under these conditions, the emulsion achieves anything but elastic [23,44]? The reason that this is not
random close packing (RCP) of spheres at Фс « .64 [6]. so clear is that if the material is fragile or isostatic, it
Indeed, for a three dimensional emulsion at Фс, computer is unable to suffer certain applied stresses, regardless of
simulation [37] indicates that z — 6. their strength, without internal rearrangement. Since the
What happens for granular materials? If the system grains, although elastic, are so hard, to a certain extent
consists of sand (for example) poured into a container in they may be expected to behave like ideal hard particles,
some particular fashion, then due to the rapid dissipa and it is here that the ideas of fragility and isostaticity
tion of kinetic energy, the density of the resulting pack may enter.
ing is considerably lower than Фс = .64. In this case, Why should a granular material be isostatic? Consider
it is interesting to ask whether the packing achieved is again an emulsion, which at low volume fractions behaves
marginally stable, that is, on the boundary separating like a viscous liquid, the viscosity enhanced by the pres
stable from unstable in the space of states of the system. ence of the droplets. As the volume fraction increases,
A marginally stable state may be thought of as the most the viscosity increases, diverging at Фс, when the system
unstable of stable states, and we may speculate that it develops a shear modulus [5,45,46]. Just at this point,
is what occurs for as-poured sand since the system jams the droplets are undeformed spheres, with average coor
up, stopping at the first state [38] which can support dination 6. If the volume fraction is increased further,
whatever external load is present during the aggregate’s the average coordination number increases (growing as
12
Another random document with
no related content on Scribd:
33. Changes in the United States and Japan[109]
Every war has two aspects, the defensive and the offensive, to each of
which there is a corresponding factor of activity. There is something
to gain, the offensive; there is something to lose, the defensive. The
ears of men, especially of the uninstructed, are more readily and
sympathetically open to the demands of the latter. It appeals to the
conservatism which is dominant in the well-to-do, and to the
widespread timidity which hesitates to take any risk for the sake of a
probable though uncertain gain. The sentiment is entirely
respectable in itself, and more than respectable when its power is
exercised against breach of the peace for other than the gravest
motives—for any mere lucre of gain. But its limitations must be
understood. A sound defensive scheme, sustaining the bases of the
national force, is the foundation upon which war rests; but who lays
a foundation without intending a superstructure? The offensive
element in warfare is the superstructure, the end and aim for which
the defensive exists, and apart from which it is to all purposes of war
worse than useless. When war has been accepted as necessary,
success means nothing short of victory; and victory must be sought
by offensive measures, and by them only can be ensured. “Being in,
bear it, that the opposer may be ware of thee.” No mere defensive
attitude or action avails to such end. Whatever the particular mode
of offensive action adopted, whether it be direct military attack, or
the national exhaustion of the opponent by cutting off the sources of
national well-being, whatsoever method may be chosen, offense,
injury, weakening of the foe, to annihilation if need be, must be the
guiding purpose of the belligerent. Success will certainly attend him
who drives his adversary into the position of the defensive and keeps
him there.
Offense therefore dominates, but it does not exclude. The necessity
for defense remains obligatory, though subordinate. The two are
complementary. It is only in the reversal of rôles, by which priority of
importance is assigned to the defensive, that ultimate defeat is
involved. Nor is this all. Though opposed in idea and separable in
method of action, circumstances not infrequently have permitted the
union of the two in a single general plan of campaign, which protects
at the same time that it attacks. “Fitz James’s blade was sword and
shield.” Of this the system of blockades by the British Navy during
the Napoleonic wars was a marked example. Thrust up against the
ports of France, and lining her coasts, they covered—shielded—the
operations of their own commerce and cruisers in every sea; while at
the same time, crossing swords, as it were, with the fleets within,
ever on guard, ready to attack, should the enemy give an opening by
quitting the shelter of his ports, they frustrated his efforts at a
combination of his squadrons by which alone he could hope to
reverse conditions. All this was defensive; but the same operation cut
the sinews of the enemy’s power by depriving him of sea-borne
commerce, and promoted the reduction of his colonies. Both these
were measures of offense; and both, it may be added, were directed
upon the national communications, the sources of national well-
being. The means was one, the effect twofold....
[It is shown that, in the case of insular states, offense and defense
are often closely combined, home security depending on control of
the sea assured by offensive action of the national fleet.—Editor.]
An insular state, which alone can be purely maritime, therefore
contemplates war from a position of antecedent probable superiority
from the twofold concentration of its policy; defense and offense
being closely identified, and energy, if exerted judiciously, being
fixed upon the increase of naval force to the clear subordination of
that more narrowly styled military. The conditions tend to minimize
the division of effort between offensive and defensive, purpose, and,
by greater comparative development of the fleet, to supply a larger
margin of disposable numbers in order to constitute a mobile
superiority at a particular point of the general field. Such a decisive
local superiority at the critical point of action is the chief end of the
military art, alike in tactics and strategy. Hence it is clear that an
insular state, if attentive to the conditions that should dictate its
policy, is inevitably led to possess a superiority in that particular kind
of force, the mobility of which enables it most readily to project its
power to the more distant quarters of the earth, and also to change
its point of application at will with unequalled rapidity.
The general considerations that have been advanced concern all
the great European nations, in so far as they look outside their own
continent, and to maritime expansion, for the extension of national
influence and power; but the effect upon the action of each differs
necessarily according to their several conditions. The problem of sea-
defense, for instance, relates primarily to the protection of the
national commerce everywhere, and specifically as it draws near the
home ports; serious attack upon the coast, or upon the ports
themselves, being a secondary consideration, because little likely to
befall a nation able to extend its power far enough to sea to protect
its merchant ships. From this point of view the position of Germany
is embarrassed at once by the fact that she has, as regards the world
at large, but one coast-line. To and from this all her sea commerce
must go; either passing the English Channel, flanked for three
hundred miles by France on the one side and England on the other,
or else going north about by the Orkneys, a most inconvenient
circuit, and obtaining but imperfect shelter from recourse to this
deflected route. Holland, in her ancient wars with England, when the
two were fairly matched in point of numbers, had dire experience of
this false position, though her navy was little inferior in numbers to
that of her opponent. This is another exemplification of the truth that
distance is a factor equivalent to a certain number of ships. Sea-
defense for Germany, in case of war with France or England, means
established naval predominance at least in the North Sea; nor can it
be considered complete unless extended through the Channel and as
far as Great Britain will have to project hers into the Atlantic. This is
Germany’s initial disadvantage of position, to be overcome only by
adequate superiority of numbers; and it receives little compensation
from the security of her Baltic trade, and the facility for closing that
sea to her enemies. In fact, Great Britain, whose North Sea trade is
but one-fourth of her total, lies to Germany as Ireland does to Great
Britain, flanking both routes to the Atlantic; but the great
development of the British sea-coast, its numerous ports and ample
internal communications, strengthen that element of sea-defense
which consists in abundant access to harbors of refuge.
For the Baltic Powers, which comprise all the maritime States east
of Germany, the commercial drawback of the Orkney route is a little
less than for Hamburg and Bremen, in that the exit from the Baltic is
nearly equidistant from the north and south extremities of England;
nevertheless the excess in distance over the Channel route remains
very considerable. The initial naval disadvantage is in no wise
diminished. For all the communities east of the Straits of Dover it
remains true that in war commerce is paralyzed, and all the resultant
consequences of impaired national strength entailed, unless decisive
control of the North Sea is established. That effected, there is
security for commerce by the northern passage; but this alone is
mere defense. Offense, exerted anywhere on the globe, requires a
surplusage of force, over that required to hold the North Sea,
sufficient to extend and maintain itself west of the British Islands. In
case of war with either of the Channel Powers, this means, as
between the two opponents, that the eastern belligerent has to guard
a long line of communications, and maintain distant positions,
against an antagonist resting on a central position, with interior
lines, able to strike at choice at either wing of the enemy’s extended
front. The relation which the English Channel, with its branch the
Irish Sea, bears to the North Sea and the Atlantic—that of an interior
position—is the same which the Mediterranean bears to the Atlantic
and the Indian Sea; nor is it merely fanciful to trace in the passage
round the north of Scotland an analogy to that by the Cape of Good
Hope. It is a reproduction in miniature. The conditions are similar,
the scale different. What the one is to a war whose scene is the north
of Europe, the other is to operations by European Powers in Eastern
Asia.
To protract such a situation is intolerable to the purse and morale
of the belligerent who has the disadvantage of position. This of
course leads us straight back to the fundamental principles of all
naval war, namely, that defense is ensured only by offense, and that
the one decisive objective of the offensive is the enemy’s organized
force, his battle fleet. Therefore, in the event of a war between one of
the Channel Powers, and one or more of those to the eastward, the
control of the North Sea must be at once decided. For the eastern
State it is a matter of obvious immediate necessity, of commercial
self-preservation. For the western State the offensive motive is
equally imperative; but for Great Britain there is defensive need as
well. Her Empire imposes such a development of naval force as
makes it economically impracticable to maintain an army as large as
those of the Continent. Security against invasion depends therefore
upon the fleet. Postponing more distant interests, she must here
concentrate an indisputable superiority. It is, however, inconceivable
that against any one Power Great Britain should not be able here to
exert from the first a preponderance which would effectually cover
all her remoter possessions. Only an economical decadence, which
would of itself destroy her position among nations, could bring her
so to forego the initial advantage she has, in the fact that for her
offense and defense meet and are fulfilled in one factor, the
command of the sea. History has conclusively demonstrated the
inability of a state with even a single continental frontier to compete
in naval development with one that is insular, although of smaller
population and resources. A coalition of Powers may indeed affect
the balance. As a rule, however, a single state against a coalition
holds the interior position, the concentrated force; and while
calculation should rightly take account of possibilities, it should
beware of permitting imagination too free sway in presenting its
pictures. Were the eastern Powers to combine they might prevent
Great Britain’s use of the North Sea for the safe passage of her
merchant shipping; but even so she would but lose commercially the
whole of a trade, the greater part of which disappears by the mere
fact of war. Invasion is not possible, unless her fleet can be wholly
disabled from appearing in that sea. From her geographical position,
she still holds her gates open to the outer world, which maintains
three fourths of her commerce in peace.
37. Bearing of Political Developments on Naval
Policy and Strategy[115]
The external activities of Europe, noted a dozen years ago and before,
have now to a certain extent been again superseded by rivalries
within Europe itself. Those rivalries, however, are the result of their
previous external activities, and in the last analysis they depend
upon German commercial development. This has stimulated the
German Empire to a prodigious naval programme, which affects the
whole of Europe and may affect the United States. In 1897 I summed
up two conspicuous European conditions as being the equilibrium
then existing between France and Germany, with their respective
allies, and the withdrawal of Great Britain from active association
with the affairs of the Continent. At that date the Triple Alliance,
Austria, Germany, Italy, stood against the Dual Alliance, France and
Russia; Great Britain apart from both, but with elements of
antagonism against Russia and France, and not against the German
monarchies or Italy. These antagonisms arose wholly from
conditions external to Europe,—in India against Russia, and in Africa
against France. Later, the paralysis of Russia, through her defeat by
Japan, and through her internal troubles, left France alone for a
time; during which Germany, thus assured against land attack, was
better able to devote much money to the fleet, as the protector of her
growing commerce. The results have been a projected huge German
navy, and a German altercation with France relative to Moroccan
affairs; incidents which have aroused Great Britain to a sense of
naval danger, and have propelled her to the understandings—
whatever they amount to—with France and Russia, which we now
know as the Triple Entente. In short, Great Britain has abandoned
the isolation of twenty years ago, stands joined to the Dual Alliance,
and it becomes a Triple Entente.
To the United States this means that Great Britain, once our chief
opponent in matters covered by the Monroe Doctrine, but later by
the logic of events drawn to recede from that opposition, so that she
practically backed us against Europe in 1898, and subsequently
conceded the Panama arrangement known as the Hay-Pauncefote
Treaty, cannot at present count for as much as she did in naval
questions throughout the world. It means to the United States and to
Japan that Great Britain has too much at stake at home to side with
the one or the other, granting she so wished, except as bound by
treaty, which implies reciprocal obligations. Between her and Japan
such specific obligations exist. They do not in the case of the United
States; and the question whether the two countries are disposed to
support one another, and, if so, to what extent, or what the attitude
of Great Britain would be in case of difficulty between Japan and the
United States, are questions directly affecting naval strategy.[116]
Great Britain does indeed for the moment hold Germany so far in
check that the German Empire also can do no more than look after
its European interests; but should a naval disaster befall Great
Britain, leaving Germany master of the naval situation, the world
would see again a predominant fleet backed by a predominant army,
and that in the hands, not of a state satiated with colonial
possessions, as Great Britain is, but of one whose late entry into
world conditions leaves her without any such possessions at all of
any great value. The habit of mind is narrow which fails to see that a
navy such as Germany is now building will be efficacious for other
ends than those immediately proposed. The existence of such a fleet
is a constant factor in contemporary politics; the part which it shall
play depending upon circumstances not always to be foreseen.
Although the colonial ambitions of Germany are held in abeyance for
the moment, the wish cannot but exist to expand her territory by
foreign acquisitions, to establish external bases for the support of
commercial or political interests, to build up such kindred
communities as now help to constitute the British Empire, homes for
emigrants, markets for industries, sources of supplies of raw
materials, needed by those industries.
All such conditions and ambitions are incidents with which
Strategy, comprehensively considered, has to deal. By the successive
enunciations of the Monroe Doctrine the United States stands
committed to the position that no particle of American soil shall pass
into the hands of a non-American State other than the present
possessor. No successful war between foreign states, no purchase, no
exchange, no merger, such as the not impossible one of Holland with
Germany, is allowed as valid cause for such transfer. This is a very
large contract; the only guarantee of which is an adequate navy,
however the term “adequate” be defined. Adequacy often depends
not only upon existing balances of power, such, for instance, as that
by which the British and German navies now affect one another,
which for the moment secures the observance of the Doctrine.
Account must be taken also of evident policies which threaten to
disturb such balances, such as the official announcement by
Germany of her purpose to create a “fleet of such strength that, even
for the mightiest naval power, a war with Germany would involve
such risks as to jeopardize its own supremacy.” This means, at least,
that Great Britain hereafter shall not venture, as in 1898, to back the
United States against European interference; nor to support France
in Morocco; nor to carry out as against Germany her alliance with
Japan. It is a matter of very distinct consequence in naval strategy
that Great Britain, after years of contention with the United States,
essentially opposed to the claims of the Monroe Doctrine, should at
last have come to substantial coincidence with the American point of
view, even though she is not committed to a formal announcement to
that effect.[117] Such relations between states are primarily the
concern of the statesman, a matter of international policies; but they
are also among the data which the strategist, naval as well as land,
has to consider, because they are among the elements which
determine the constitution and size of the national fleet.
I here quote with approval a statement of the French Captain
Darrieus:
“Among the complex problems to which the idea of strategy gives
rise there is none more important than that of the constitution of the
fleet; and every project which takes no account of the foreign
relations of a great nation, nor of the material limit fixed by its
resources, rests upon a weak and unstable base.”
I repeat also the quotation from Von der Goltz: “We must have a
national strategy, a national tactics.” I cannot too entirely repudiate
any casual word of mine, reflecting the tone which once was so
traditional in the navy that it might be called professional,—that
“political questions belong rather to the statesman than to the
military man.” I find these words in my old lectures, but I very soon
learned better, from my best military friend, Jomini; and I believe
that no printed book of mine endorses the opinion that external
politics are of no professional concern to military men.
It was in accordance with this changed opinion that in 1895, and
again in 1897, I summed up European conditions as I conceived
them to be; pointing out that the distinguishing feature at that time
was substantial equilibrium on the Continent, constituting what is
called the Balance of Power; and, in connection with the calm thus
resulting, an immense colonizing movement, in which substantially
all the great Powers were concerned. This I indicated as worthy of
the notice of naval strategists, because there were parts of the
American continents which for various reasons might attract upon
themselves this movement, in disregard of the Monroe Doctrine.
Since then the scene has shifted greatly, the distinctive feature of
the change being the growth of Germany in industrial, commercial,
and naval power,—all three; while at the same time maintaining her
military pre-eminence, although that has been somewhat qualified
by the improvement of the French army, just as the growth of the
German navy has qualified British superiority at sea. Coincident with
this German development has been the decline of Russia, owing to
causes generally understood; the stationariness of France in
population, while Germany has increased fifty per cent; and the very
close drawing together of Germany and Austria, for reasons of much
more controlling power than the mere treaty which binds them. The
result is that to-day central Europe, that is, Austria and Germany,
form a substantially united body, extending from water to water,
from North Sea to Adriatic, wielding a military power against which,
on the land, no combination in Europe can stand. The Balance of
Power no longer exists; that is, if my estimate is correct of the
conditions and dispersion which characterize the other nations
relatively to this central mass.
This situation, coinciding with British trade jealousies of the new
German industries, and with the German naval programme, have
forced Great Britain out of the isolation which the Balance of Power
permitted her. Her ententes are an attempt to correct the
disturbance of the balance; but, while they tend in that direction,
they are not adequate to the full result desired. The balance remains
uneven; and consequently European attention is concentrated upon
European conditions, instead of upon the colonizing movements of
twenty years ago. Germany even has formally disavowed such
colonizing ambitions, by the mouth of her ambassador to the United
States, confirmed by her minister of foreign affairs, although a dozen
years ago they were conspicuous. Concerning these colonizing
movements, indeed, it might be said that they have reached a
moment of quiet, of equilibrium, while internally Europe is
essentially disquieted, as various incidents have shown.
The important point to us here is the growing power of the
German Empire, in which the efficiency of the State as an organic
body is so greatly superior to that of Great Britain, and may prove to
be to that of the United States. The two English-speaking countries
have wealth vastly superior, each separately, to that of Germany;
much more if acting together. But in neither is the efficiency of the
Government for handling the resources comparable to that of
Germany; and there is no apparent chance or recognized inducement
for them to work together, as Germany and Austria now work in
Europe. The consequence is that Germany may deal with each in
succession much more effectively than either is now willing to
consider; Europe being powerless to affect the issue so long as
Austria stands by Germany, as she thoroughly understands that she
has every motive to do.
It is this line of reasoning which shows the power of the German
navy to be a matter of prime importance to the United States. The
power to control Germany does not exist in Europe, except in the
British navy; and if social and political conditions in Great Britain
develop as they now promise, the British navy will probably decline
in relative strength, so that it will not venture to withstand the
German on any broad lines of policy, but only in the narrowest sense
of immediate British interests. Even this condition may disappear,
for it seems as if the national life of Great Britain were waning at the
same time that that of Germany is waxing. The truth is, Germany, by
traditions of two centuries, inherits now a system of state control,
not only highly developed but with a people accustomed to it,—a
great element of force; and this at the time when control of the
individual by the community—that is, by the state—is increasingly
the note of the times. Germany has in this matter a large start. Japan
has much the same.
When it is remembered that the United States, like Great Britain
and like Japan, can be approached only by sea, we can scarcely fail to
see that upon the sea primarily must be found our power to secure
our own borders and to sustain our external policy, of which at the
present moment there are two principal elements; namely, the
Monroe Doctrine and the Open Door. Of the Monroe Doctrine
President Taft, in his first message to Congress, has said that it has
advanced sensibly towards general acceptance; and that
maintenance of its positions in the future need cause less anxiety
than it has in the past. Admitting this, and disregarding the fact that
the respect conceded to it by Europe depends in part at least upon
European rivalries modifying European ability to intervene,—a
condition which may change as suddenly as has the power of Russia
within the decade,—it remains obvious that the policy of the Open
Door requires naval power quite as really and little less directly than
the Monroe Doctrine. For the scene of the Open Door contention is
the Pacific; the gateway to the Pacific for the United States is the
Isthmus; the communications to the Isthmus are by way of the Gulf
of Mexico and the Caribbean Sea. The interest of that maritime
region therefore is even greater now than it was when I first
undertook the strategic study of it, over twenty years ago. Its
importance to the Monroe Doctrine and to general commercial
interests remains, even if modified.
At the date of my first attempt to make this study of the Caribbean,
and to formulate certain principles relative to Naval Strategy, there
scarcely could be said to exist any defined public consciousness of
European and American interest in sea power, and in the methods of
its application which form the study of Strategy. The most striking
illustration of this insensibility to the sea was to be found in
Bismarck, who in a constructive sense was the greatest European
statesman of that day. After the war with France and the acquisition
of Alsace and Lorraine, he spoke of Germany as a state satiated with
territorial expansion. In the matter of external policy she had
reached the limits of his ambitions for her; and his mind thenceforth
was set on internal development, which should harmonize the body
politic and ensure Germany the unity and power which he had won
for her. His scheme of external relations did not stretch beyond
Europe. He was then too old to change to different conceptions,
although he did not neglect to follow the demand of the people as
their industry and commerce developed.
The contrast between the condition of indifference to the sea
which he illustrated and that which now exists is striking; and the
German Empire, which owes to him above all men its modern
greatness, offers the most conspicuous illustration of the change. The
new great navies of the world since 1887 are the German, the
Japanese, and the American. Every state in Europe is now awake to
the fact that the immediate coming interests of the world, which are
therefore its own national interest, must be in the other continents.
Europe in its relatively settled conditions offers really the base of
operations for enterprises and decisive events, the scene of which
will be in countries where political or economical backwardness must
give place to advances which will be almost revolutionary in kind.
This can scarcely be accomplished without unsettlements, the
composing of which will depend upon force. Such force by a
European state—with the single exception of Russia, and possibly, in
a less degree, of Austria—can be exerted only through a navy.