INSTITUTE FOR NONLINEAR SCIENCE {| aoe ae 1GreKee SmBCO Tl ie The Transition to Chaos Conservative Classical Systems and Quantum Manifestations Second Edition ; NSyebittagThis book provides a thorough and comprehensive discussion of clas- sical and quantum chaos theory for bounded systems and for scatter- ing processes. Specific discussions include: * Noether's theorem, integrability, KAM theory, and a definition of chaotic behavior. * Area-preserving maps, quantum billiards, semiclassical quantization, chaotic scattering, scaling in classical and quantum dynamics, dynamic localization, dynamic tunneling, effects of chaos in periodi- cally driven systems and stochastic systems. « Random matrix theory and supersymmetry. The book is divided into several parts. Chapters 2 through 4 deal with the dynamics of nonlinear conservative classical systems. Chapter 5 and several appendices give a thorough grounding in random matrix theory and supersymmetry techniques. Chapters 6 and 7 discuss the manifestations of chaos in bounded quantum systems and open quan- tum systems, respectively. Chapter 8 focuses on the semiclassical description of quantum systems with underlying classical chaos, and Chapter 9 discusses the quantum mechanics of systems driven by time-periodic forces. Chapter 10 reviews some recent work on the sto- eee CC ei Relea Miao meee Ce ule eet lal erm-19)3\-ulello Moco) ol) much of the needed mathematical background, and there are extensive references to the current literature. End-of-chapter problems help stu- eee Uma Mme tele ME colo Mk CUA) has been brought up-to-date throughout, and a new chapter on open quantum systems has been added. Cte a ELLs Linda E. Reichl, Ph.D., is Professor of Physics at the University of Texas at Austin and has served as Acting Director of the llya Prigogine Center for Statistical Mechanics and Complex Systems since 1974. She is a fellow of the American Physical Society and currently is U.S. Editor of the journal Chaos, Solitons, and Fractals. BN 0-387-98788 ll SSS NOI Y es -1 fos oo) www.springer-ny.com 7Linda E. Reichl The Transition to Chaos Conservative Classical Systems and Quantum Manifestations Second Edition With 180 Illustrations 6 SpringerLinda E. Reich! Department of Physics and Center for Statistical Mechanics and Complex Systems University of Texas at Austin Austin, TX 78712 USA reich @physics.utexas.edu Editorial Board Institute for Nonlinear Science, University of California—San Diego Henry D.I. Abarbanel, Physics (Scripps Institution of Oceanography) Morteza Gharib, Applied Mechanics and Engineering Sciences Michael E. Gilpin, Biology Walter Heller, Economics Katja Lindenberg, Chemistry Manuel Rotenberg, Electrical and Computer Engineering John D. Simon, Chemistry Library of Congress Cataloging-in-Publication Data Reichi, L. E. The transition to chaos: conservative classical systems and quantum manifestations / Linda E. Reichl.—[New ed.]. p. cm.—(Institute for nonlinear science) Includes bibliographical references and index. ISBN 0-387-98788-6 (alk. paper) 1. Chaotic behavior in systems. 1. Title. Il. Institute for nonlinear science (Springer-Verlag) Q172.5.C45R45 2004 003.857—de22 2003062211 ISBN 0-387-98788-6 Printed on acid-free paper. © 2004 Springer-Verlag New York, Inc. All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer-Verlag New York, Inc., 175 Fifth Avenue, New York, NY 10010, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. ‘The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights Printed in the United States of America. (EB) 987654321 SPIN 10715217 Springer-Verlag is a part of Springer Science+Business Media springeronline.comDedication This book is dedicated to Dr. Byron E. Cohn, ‘A wonderful teacher at the University of Denver who started me along the beautiful path of physicsAcknowledgements To begin, I wish to thank Katja Lindenberg for inviting me to write this book for the Springer INLS Series. The first edition of this book, which appeared in 1992, was the result of lectures on classical and quantum chaos theory that I gave at the Institute for Nonlinear Science, University of California, San Diego, in 1987, and later at Guangxi Normal University in Guilin, China, and at the University of Texas in Austin, Texas. The first edition focused on classical chaos theory and the manifestations of chaos in bounded quantum systems. This new edition contains selected material from the first edition but also discusses the manifestations of chaos in open systems, which has been a major focus of the field in recent years. The new edition also contains a thorough grounding in random mairix theory and supersymmetry techniques, which have become essential for analyzing properties of quantum systems whose classical counterpart is chaotic. As before, I have attempted to write the book both as a textbook and as a research resource. Because it was necessary to keep the book a reasonable length, I have made a judgment about the material that I use to illustrate ideas, but at the same time I have tried to reference all other relevant work that I know about. This book has benefited from discussions with many colleagues and stu- dents in the fields of classical and quantum chaos over the years. I hope I have done justice to the contributions they have all made to this important field of dynamics. Linda Reichl The University of Texas at Austin March 2003Contents Dedication Acknowledgements 1 Overview 11 Introduction... 2... 2. eee 1.2 Historical Overview 1.3 Plan of the Book 1.4 References .... . 2 Fundamental Concepts 2.1 Introduction... 2... 2.0.20. 2.2 Conventional Perturbation Theory . 2.3. Integrability ........... 2.3.1 Noether’s Theorem . 2.3.2 Hidden Symmetries... . . 2.3.3 Poincaré Surface of Section 2.4 Nonlinear Resonance and Chaos . 2.4.1 Single-Resonance Hamiltonians 2.4.2 Two-Resonance Hamiltonian 2.5 KAMTheory ... 2.6 The Definition of Chaos... . . 2.6.1 Lyapounov Exponent . . . To 2.6.2 KS Metric Entropy and K-Flowsx Conte 3.3 3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 ents Time-Dependent Hamiltonians Conclusions ...........-. Problems... ....-..-5-% References ...-.-......- Preserving Maps Introduction»... . Twist Maps 3.2.1 Derivation of a Twist Map from a Torus . 3.2.2 Generating Functions . - 3.2.3. Birkhoff Fixed Point Theorem 3.2.4 The Tangent Map............ 3.2.5 Homoclinic and Heteroclinic Points . . Melnikov Distance . Whisker Maps . . . The Standard Map o 3.5.1 Rational and Irrational Orbits... . . 3.5.2 Accelerator Modes ........... Scaling Behavior of Noble KAM Tori 3.6.1 Rational Approximates......... 3.6.2 Scaling Properties of Twist Maps . . . Renormalization in Twist Maps 3.7.1 Integrable Twist Map. . . . 3.7.2 Nonintegrable Twist Map 3.7.3. The Universal Map Bifurcation of M-Cycles 3.8.1 Some General Properties 3.8.2 The Quadratic Map... . 3.8.3. Scaling in the Quadratic de Vogelaere Map Cantoti eee eee Diffusion in Two-Dimensional Twist Maps. S00 mfectolCantaig 9s ee |e 3.10.2. Diffusion in the Standard Map. . . . Conclusions Lee Problems... ... References... . . Global Properties 41 4.2 4.3 Introduction... ... Important Models . . 4.2.1 Delta-Kicked Rotor (Standard Map). 4.2.2 The Duffing Oscillator... . 423 Driven Particle in Infinite Square Well Potential - 4.2.4 Driven One-Dimensional Hydrogen ........ Renormalization Map... 00 eee eee eee 134 136 139 140 141Contents xi 4.3.1 The Paradigm Hamiltonian 4.3.2. The Renormalization Map ene 43.3 Fixed Points of the Renormalization Map .... 153 4.4 Application of Renormalization Predictions . 158 4.4.1 Driven Square-Well System .... . 158 4.4.2 Duffing Oscillator oe 161 4.5 Scattering Chaos ........ 162 46 Stochastic Tiling ........ = 167 4.6.1 Delta-Kicked Harmonic Oscillator we 167 4.6.2 Two Primary Resonances Model . . . 168 4.7 Arnol’d Diffusion ........ . 170 4.7.1 Resonance Networks 171 4.7.2. Numerical Observations - 174 4.7.3 Diffusion Along Separatrix Layers . . . 175 4.7.4 Diffusion Coefficient Bo 180 4.7.5 Some Applications . . . 182 4.8 Conclusions bob 50 186 4.9 Problems. ..... 186 4.10 References 187 Random Matrix Theory 189 5.1 Introduction. ........- 189 5.2 Ensembles ............ 192 5.2.1 Gaussian Ensembles 193, 5.2.2 Circular Ensembles . . . 196 5.3 Cluster Functions... 2... ES 19t, 5.3.1 Cluster Expansion of the Probability Densities... 197 5.3.2 Probability Densities and Quaternion Determinants 199 5.3.3. Cluster Functions for Gaussian Ensembles .... 200 5.3.4 Cluster Functions for Circular Ensembles... .. 202 5.4 Eigenvalue Number Density ............ 204 5.4.1 Eigenvalue Number Density for Gaussian Ensembles 204 5.4.2 Bigenvalue Number Density for Circular Ensembles 210 5.5 Eigenvalue Correlations - As-Statistic .......... 2u1 5.5.1 Ag-Statistic - General Expressions . . 211 5.5.2. Ag-Statistics for Gaussian Ensembles . 214 5.5.3 Ag3-Statistics for Circular Ensembles 220 5.6 Eigenvalue Nearest Neighbor Spacing Distribution (GOE) 223 5.6.1 Bigenvalue Spacing Distributions (N=2) ... . . 223 5.6.2 Nearest Neighbor Spacing Distribution (N—+00) . 225 5.6.3 Approximate Nearest Neighbor Spacing Distribu- tions for GOE (N00)... . : 5.7 Eigenvector Statistics ~ Gaussian Ensembles . ee Give General Droperticgs 0 ee 229 5.7.2. Distribution of Eigenvector Components (GOE) . 231xii Contents 5.7.3 Distribution of Eigenvector Components ee 5.7.4 Gaussian Symplectic Ensemble .... . . 7 5.8 Conclusions se 5.9 Problems...... 5.10 References Bounded Quantum Systems Gil ntroduction i ee 6.2 Quantum a 6.3 Symmetries and Degeneracy . 6.4 Quantum Billiards... ..... 6.4.1 The Rectangular Billiard 6.4.2 The Stadium ...... 6.4.3. The Sinai Billiard... . 6.4.4 The Ripple Billiard . . . 6.5 The Quantized Baker’s Map . . 6.6 Time Average as an Invariant a 6.7 Integrable and Nonintegrable Spin Systems... ... . . 6.7.1 Classical Spin Models ere 6.7.2 Quantum Spin Models ............00. 6.7.3 Two-Dimensional Clusters with N Spin 4 2 Objects 6.8 Anharmonic Oscillators... 6... 2.0 eee eee 6.8.1 Polynomial Anharmonicity 6.8.2 Coupled Morse Oscillators . 6.9 Conclusions 6.10 Problems... . 6.11 References . . . Manifestations of Chaos in Quantum Scattering Pro- cesses 7.1 Introduction .......... 7.2 Scattering Theory . 7.2.1 Hamiltonian . 7 72.2 Energy Figenstates 7.2.3 The Reaction Matrix . . 7.2.4 The Scattering Matrix . ae 7.3. Wigner-Smith and Partial Delay Times |. . Soon 7.3.1 Delay Time of a Wave Packet ....... ee 7.3.2 Delay Times for Multichannel Scattering . .. . . 7.3.3 Delay Times and Complex Poles 7.4 Scattering Theory andGOE......... 74.1 The Average S-Matrix ........ 74.2 When Does a GOE Hamiltonian Yield a COE S- Matrix? 7.4.3 S-Matrix Correlation Function (GOE)Contents xiii 244s Delay Time) Density eee 326 7.5 §-Matrix Correlation Functions (COE) . 327 7.6 Green’s Function and S-Matrix . . 330 7.6.1 The Green’s Function . rn) 7.6.2 Green’s Function for Quantum Waveguide... 331 7.6.3 Transmission Amplitudes and the Green’s Function 332 7.7 Absorption Spectrum and Green’s Function ....... 336 7.8 Experimental Observation of RMT Predictions. ..... 339 7.8.1 Experimental Nuclear Spectral Statistics ..... 339 7.8.2 Experimental Molecular Spectral Statistics... 341 7.9 Conclusions nooo 342 7.10 Problems. ..... 344 7.11 References 345 Semiclassical Theory—Path Integrals 348 8.1 Introduction... ee eee 348 8.2 Green’s Function and Density of States . . . . - 2. 350 8.3 The Path Integral... . . 351 8.3.1 The General Case, # =T7+V 351 8.4 Semiclassical Approximation .......... te 854 8.4.1 Method of Stationary Phase 354 8.4.2. The Semiclassical Green’s Function . . 355, 8.4.3 Conjugate Points ... . 360 8.5 Energy Green’s Function . . . . 362 8.5.1 General Expression... 2.0.0... 362 8.5.2 Density of States .... 368 8.6 Ag-Statistic for a Rectangular Billiard 372 8.6.1 Energy Green’s Function for a Rectangular Billiard 372 8.6.2 Density of States for the Rectangular Billiard .. 374 8.6.3 Semiclassical Expression for the Ag-Statistic ... 377 8.7 Gutzwiller Trace Formula. . . . : 379 8.7.1 Response Function for Chaotic Systems 380 8.8 Anisotropic Kepler System . . . bono 385 8.9 Diamagnetic Hydrogen .... . 389 8.9.1 The Model . a 390 8.9.2 Absorption Cross Section . 392 8.9.3 Experiment 393 8.9.4 Semiclassical Cross Section 395 8.10 Conclusions ...........- 398 8.11 Problems . 399 8.12 References 399 Time-Periodic Systems 401 9.1 Introduction... . 401 9.2 Floquet Theory . . 403xiv Contents: 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 9.11 9.12 9.13 9.2.1 Floquet Matrix ....... 9.2.2 Floquet Hamiltonian .. . . Nonlinear Quantum Resonances . . 9.3.1 Two Primary Resonance Model . . . . 9.3.2 Floquet Eigenstates . 9.3.3 Quantum Resonance Overlap 9.3.4 Floquet Eigenvalue Nearest Neighbor Spacing . . Eigenvalue Avoided Crossings and Wave Function Delocal- ization . . 9.4.1 Classical Driven Square-Well System . 9.4.2 Quantum Driven Square-Well System . 9.4.3 Avoided Crossings and Delocalization . 9.4.4 High Harmonic Radiation Lee Dynamical Tunneling in Atom Optics Experiments... 9.5.1 Hamiltonian for Atomic Center-of-Mass . 9.5.2 Average Momentum of Cesium Atoms ...... 9.5.3 Floquet Analysis of Tunneling Oscillations . . . . Quantum Renormalization... . . oe 9.6.1 Paradigm Schrédinger Equation.......... 9.6.2 Higher-Order Resonance: 9.6.3 Quantum Renormalization Map .. . . 9.6.4 Stable Manifold... ..... 9.6.5 Scaling Functions ...... 9.6.6 Scaling of Localization Lengths . . . Quantum Delta-Kicked Rotor .. . 9.7.1 The Schrédinger Equation for the Delta-| Kicked Rotor 9.7.2 KAM: ike Behavior of the Quantum Delta-Kicked Rotor 9.7.3 The Floquet Map . 9.7.4 Spectral Statistics Dynamic Anderson Localization: Delta-Kicked Rotor . . 9.8.1 Tight-Binding Model for the Delta-Kicked Rotor 9.8.2 Diffusion Coefficient and Localization Length 9.8.3 Atom Optics Realization of the Delta-Kicked Rotor Microwave-Driven Hydrogen . . Lene 9.9.1 Experimental Apparatus 9.9.2 One-Dimensional Approximation Dynamic Anderson Localization - Microwave-Driven Hy- drogen 9.10.1 Diffusion in Microwave-Driven Hydrogen... . . 9.10.2. Experimental Observation of Dynamic Localization Conclusions Problems References ... . 403 405 406 406 408 442 443, Add 446 AAT 448 452 453 457 457 460 465 465, 467 469 471 471Contents 10 Stochastic Manifestations of Chaos 10.1 Introduction... 2... 2 ee 10.2 Brownian Motion in Two Space Dimensions . . . 10.3 Random Walk in Two Space Dimensions . 10.4 Time Periodically Driven Brownian Motion in One Space Dimension 10.4.1 Schrédinger-like Bauation 10.5 Conclusion... .. 10.6 References... . . A. Classical Mechanics Al Newton’s Equations . . AQ Lagrange’s Equations... . . A3 Hamilton’s Equations... . . AA The Poisson Bracket... ... A.5 Phase Space Volume Conservation . 6 Action-Angle Coordinates A.7 Hamilton’s Principal Function . A8 References a B Simple Models Bl The Pendulum... .. 2... ee eee eee B.1.1 Libration—Trapped Orbits (Zp < 9) B.1.2_ Rotation—Untrapped Orbits (Ep > g) B.2 Double-Well Potential B.2.1 Trapped Motion—(Ey < 0). B.2.2 Untrapped Motion —(E, > 0) . - B.3 Infinite Square-Well Potential B.4 One-Dimensional Hydrogen B.4.1 Zero Stark Field... . . B.4.2 Nonzero Stark Field... 2.2... 020 C Renormalization Integral C1 v=N= Integer . . C2 v=HAlnteger .. C3. References... . . D Moyal Bracket D.1 The Wigner Function... ........0-. D.2_ Ordering of Operators D.3. Moyal Bracket . . . D4 References... . « E Symmetries and the Hamiltonian Matrix E.1 Space-Time Symmetries ............00004- 505 505 507 509 510 514xvi E.2 E3 Contents E.1.1 Continuous Symmetries E.1.2 Discrete Symmetries a Structure of the Hamiltonian Matrix ...... . E.2.1 Space-Time Homogeneity and isotropy B.2.2. Time Reversal Invariance . References ...........5- Invariant Measures Fl F.2 Hermitian Matrices ........ F.2.1 Real Symmetric Matrix . . F.2.2 Complex Hermitian Matrices . F.2.3 Quaternion Real Matrices . . F.24 General Formula for Invariant Measure of Hermitian Matrices . . F.3. Unitary Matrices F.3.1 Symmetric Unitary Matrices . F.3.2 General Unitary Matrices F.3.3. Symplectic Unitary Matrices . F.34 General Formula for Invariant Measure of Unitary Maric) pee) 7 ree F.3.5 Orthogonal Matrices . . . FA References... 2... eevee ee Quaternions G.1 References Gaussian Ensembles H.1 Vandermonde Determinant .............-..% H.2 Gaussian Unitary Ensemble (GUE) .... . H.3 Gaussian Orthogonal Ensemble (GOE) H.4 Gaussian Symplectic Ensemble (GSE)... - H.5 References... cee eee eee Circular Ensembles Il Vandermonde Determinant... . . 12 Circular Unitary Ensemble (CUE)... . . . 13 Circular Orthogonal Ensemble (COE) . 14 Circular Symplectic Ensemble (COE) . 15 References ....« beeoecccese General Definition of Invariant Measure. . . . F.1.1 Invariant Metric (Length) . . F.1.2 Invariant Measure (Volume) . Volume of Invariant Measure for Unitary Matrices JA Wc] og eobd0b00G0cp00GoDdco000dContents K Lorentzian Ensembles K.1 Normalization of AOE... 2.2... 2. ee ee eee K.2_ Relation Between COE and AOE .,..... K.3 Equivalence of COE and AOE When N-oo. . K.4__ Invariance of AOE under Inversion ..... . K.4.1 Robustness of AOE under Integration . K.5 References .........- L Grassmann Variables and Supermatrices L.l Grassmann Variables ..... 0.2... 5 L.2 Supermatrices 6... 0... 1.2.1 Transpose of a Supermatrix . . . L.22 Hermitian Adjoint of a Supermatrix . 1.2.3. Supertrace of a Supermatrix L.2.4 Determinant of a Supermatrix ... . . ae L3 References 60... eee eee M Average Response Function (GOE) M.1 Gaussian Integral for (Det[ely — Aw])~? M.2 Gaussian Integral for Det{ely — Hy] M3 Gaussian Integral for Response Function Generating Fune- ON ee M.4_ Expectation Value of the Generating Function (Part 1) M.5 The Hubbard-Stratonovitch Transformation ...... . M.6 Expectation Value of the Generating Function (Part 2) . M.7 Average Response Function Density... . . . . M.7.1 Saddle Points for the Integration over a M.7.2 Saddle Points for the Integration over w . . . M.7.3 Asymptotic Expression (N—0o) for the Average Response Function Density... .........5 M.7.4 Wigner Semicircle Law .....-.... eee M8 References... . . See eee eee N Average S-Matrix (GOE) N.1 S-Matrix Generating Function ......... N.2 Average S-Matrix Generating Function... . N.3_ Saddle Point Approximation . . . N.3.1 Case I: £ IN'3'2¢ Cane cee l oe neeg NA Integration over Grassmann Variables NS References... 2.0... eee eee eee O Maxwell’s Equations for 2-d Billiards O.1 References .... . bobuucoUOOOdd5 xvii 583 583 584 585 586 587 587 588 588 590 590 591 591 592 594 595 596 597 598, 600 601 604 608 610 612 614 616 617 618 618 619 622 623 624 624 627 628 631xviii Contents P Lloyd’s Model 632 P.1 Localization Length... 0... eee eee eee 632 P.2 References... 2... 637 Q Hydrogen in a Constant Electric Field 638 Q.1 The Schrédinger Equation . . . . 638 Q.1.1 Equation for Relative Motion 639 Q.1.2. Solution for Ay = 0 641 Q.2 One-Dimensional Hydrogen bees vee 642 Q.3 References. ........... bene vee. 644 Subject Index 645 Author Index 6671 Overview 1.1 Introduction This book is about the nonlinear dynamics of conservative classical and quantum systems. Although classical and quantum mechanics are now rather old subjects (classical mechanics is over 300 years old and quantum mechanics is over 90 years old), the mechanisms affecting their dynamical evolution have only recently been understood. In this book we will focus on the transition to chaos in classical systems and the manifestations of chaos in quantum systems. One of the important discoveries in quantum physics in recent years is that the information content of quantum systems is extremized when the underlying classical system undergoes a transition to chaos. The information content approaches that of a system whose dy- namics is governed by a random Hamiltonian matrix chosen to extremize information. For this reason, random matrix theory has become essential to quantum chaos theory. In this book we include a self-contained discussion of the random matrix theory and supersymmetry techniques necessary for the study of the statistical properties of quantum systems. Toward the end of the book, we include a short chapter showing that the manifestations of chaos can also appear in stochastic systems. The book is divided into several parts. Chapters 2 through 4 deal with the theory of nonlinear classical conservative systems. Chapter 5 and several appendices contain an overview and summary of random matrix theory and supersymmetry techniques necessary to understand many of the results of quantum chaos theory. Chapters 6 and 7 discuss the manifestations of2 1. Overview chaos in bounded quantum systems and open quantum systems (scattering systems), respectively. Chapter 8 focuses on the semiclassical description of quantum systems with underlying classical chaos, and Chapter 9 discusses the quantum mechanics of systems driven by time-periodic forces. Finally in Chapter 10, we review some recent work on the stochastic manifestations ‘of chaos. In the remainder of this chapter, we give a brief overview of the material contained in this book. 1.2 Historical Overview On April 28, 1686 the first of the three books that comprise Newton’s Principia was formally presented to the Royal Society, and by July 1687 the complete first edition (consisting of perhaps 300 copies) was published. The publication of this work was probably the most important single event in the history of science because it formulated the science of mechanics in terms of just three basic laws: 1. A body maintains its state of rest or uniform velocity unless a net force acts on it. 2. The time rate of change of momentum, p, of a body is equal to the net force, F, acting on it (F = $ or, if mass, m, is constant, F = ma, where a is the acceleration). 3. To every action there is an equal and opposite reaction. In the Principia, Newton not only wrote the three laws but also gave a systematic mathematical framework for exploring the implications of these laws. In addition, in the Principia Newton proposed his universal inverse- square law of gravitation. He then used it to derive Kepler’s empirical laws of planetary motion, to account for the motion of the moon and the phenomenon of tides, to explain the precession of the equinoxes, and to account for the behavior of falling bodies in Earth’s gravitation field. The success and power of Newton’s laws led to a great optimism about our ability to predict the behavior of mechanical objects and, as a conse- quence, led to the huge growth in science that we see today. In addition it was accompanied by a deterministic view of nature that is perhaps best exemplified in the writings of Laplace. In his Philosophical Essay on Proba- bilities he states [Laplace 1951]: Given for one instant an intelligence which could comprehend all the forces by which nature is animated and the respec- tive situation of the beings who compose it—an intelligence sufficiently vast to submit these data to analysis—it would embrace in the same formula the movements of the greatest bodies of the universe and those of the lightest atom; for it, nothing would be uncertain and the future, as the past, would be present before its eyes.”1.2. Historical Overview ' 3 This deterministic view of nature was completely natural given the suc- cess of Newtonian mechanics and persists up until the present day. Newton’s three laws of motion led to a description of the motion of point masses in terms of a set of coupled second-order differential equations. The theory of extended objects can be derived from Newton’s laws by treating them as collections of point masses. If we can specify the initial velocities and posi- tions of the point particles, then Newton’s equations for the point particles (obtained from the second law) should determine all past and future mo- tion. However, we now know that the assumption that Newton’s equations can predict the future is a fallacy. Newton’s equations are, of course, the correct starting point of mechanics, but in general they only allow us to determine the long-time behavior of integrable mechanical systems, few of which can be found in nature. Newton’s laws, for most systems, describe inherently random behavior and cannot determine the future evolution of any real system (except for very short times) in more than a probabilistic sense. The belief that Newtonian mechanics is a basis for determinism was for- mally laid to rest by Sir James Lighthill (Lighthill 1986] in a lecture to the Royal Society on the three hundredth anniversary of Newton’s Principia. In his lecture Lighthill says “... speak... once again on behalf of the broad global fraternity of practitioners of mechanics. We are all deeply conscious today that the enthusiasm of our forebears for the marvelous achievements of Newtonian mechanics led them to make generalizations in this area of predictability which, indeed, we may have generally tended to believe before 1960, but which we now recognize were false. We collectively wish to apolo- gize for having misled the general educated public by spreading ideas about the determinism of systems satisfying Newton’s laws of motion that, after 1960, were to be proved incorrect ....” Tn a sense, Newton (and Western science) were fortunate because the solar system has amazingly regular behavior considering its complexity, and one can predict its short-time behavior with fairly good accuracy. Part of the reason for this is the weakness of the gravitational force and the fact that the two-body Kepler system is integrable even though a three- body gravitational system is not integrable. Newton’s derivation of Kepler’s laws was based on the properties of the two-body system. However, the dynamical interactions of the many bodies that comprise the solar system lead to deviations from the predictions of Kepler’s laws, and lead one to ask why the solar system is, in fact, so regular. Is the solar system stable [Moser 1975]? Will it maintain its present configuration into the future? These questions have not yet been fully answered. Questions concerning the stability and the future evolution of the so- lar system have occupied scientists and mathematicians for the past 300 years. Until computers were invented, all mathematical theories used per- turbation expansions of various types. In the eighteenth century, important contributions were made by Buler, Lagrange, and Laplace on predicting41. Overview the change in the geometry of orbits due to small perturbations and on the overall stability of orbits. In addition, Lagrange [Lagrange 1889] re- formulated Newtonian mechanics in terms of a variational principle that vastly extended our ability to analyze the behavior of dynamical systems and allowed a straightforward extension to continuum mechanics. In the nineteenth century, there were two very important pieces of work that laid the groundwork for our current view of mechanics. Hamilton refor- mulated mechanics [Hamilton 1940] so that the dynamics of a mechanical system could be described in terms of a momentum-position phase space rather than a velocity-position phase space as is the case for the Lagrangian formulation. This step is extremely important because in the Hamiltonian formulation (which describes the evolution of mechanical systems in terms of coupled first-order differential equations) the flow of trajectories in phase space is volume-preserving. Furthermore, if symmetries exist (such as the space-time symmetries), then some of the generalized momenta of the sys- tem may be conserved, thus reducing the dimension of the phase space in which we must work. The relation between the symmetries of a system and conservation laws was first clarified in work by Noether (Noether 1918]. It provides one of the most important tools of twentieth century science and is extremely impor- tant to everything we shall discuss in this book. Indeed, the key to much of what we are able to predict in science is symmetry because symmetries imply conservation laws, and conservation laws give conservative classical mechanics and quantum mechanics whatever predictive power they have. Conservation laws are even responsible for the existence of thermodynamics and hydrodynamics. The other extremely important piece of work in the nineteenth century was due to Poincaré and not only closed the door on an era but created the first crack in the facade of determinism. Much of the work subsequent to Newton involved computation of deviations, after long time, from Ke- pler type orbits for two massive bodies perturbed by a third less massive body. The idea was to take a Kepler orbit as a first approximation and then compute successive corrections to it using perturbation theory. One must then show that the perturbation expansions thus obtained converge. The problem of whether or not such perturbation series converge was so important that it was the subject of a prize question posed by King Os- car II of Sweden in 1885. The question read as follows: “For an arbitrary system of mass points which attract each other according to Newton’s laws, assuming that no two points ever collide, give the coordinates of the indi- vidual points for all time as the sum of a uniformly convergent series whose terms are made up of known functions” [Moser 1975]. Poincaré entered the contest and won the prize by showing that such series could be expected to diverge because of small denominators caused by internal resonances. We now know that the resonances that give rise to these small divisors are associated with the onset of chaos. Because of these divergences, it appears1.2. Historical Overview 5 to be impossible to make long-time predictions concerning the evolution of mechanical systems (with a few exceptions such as the two-body Kepler system) using perturbation expansions. No further progress was made on the problem of long-time prediction in mechanics until 1954 when Kolmogorov [Kolmogorov 1954] outlined a proof that, for systems of the type proposed in King Oscar's question, a majority of the trajectories (for certain values of the parameters) are quasiperiodic and can be described in terms of a special type of perturbation expansion. In 1962, Arnol’d [Arnol’d 1963] constructed a formal proof of Kolmogorov’s results for a three-body system with an analytic Hamiltonian, and Moser [Moser 1968] obtained a similar result for twist maps. The result of the work of Kolmogorov, Arnol’d, and Moser (KAM) is that series expansions describing the motion of some orbits in many-body systems are convergent provided the natural frequencies associated with these orbits are not close to resonance. : Conservative systems are either integrable or nonintegrable. Integrable systems have as many independent isolating constants of the motion as they have degrees of freedom. For an integrable classical mechanical system with N degrees of freedom, each isolating integral of motion constrains the flow of trajectories to a 2N —1-dimensional surface in the 2N-dimensional phase space. The actual flow of trajectories in phase space lies on the intersection of these N surfaces. Thus, for integrable systems, a given trajectory lies on an N-dimensional surface (the intersection of N surfaces) in the 2N- dimensional phase space, and every trajectory is either quasiperiodic or periodic and every trajectory is stable. The only example of an integrable mechanical system with N degrees of freedom (N > 2) is the Toda lattice {Toda 1981], and it is integrable both classically and quantum mechanically. Nonintegrable systems may themselves be divided into two classes. One class contains the completely chaotic systems such as the Sinai billiard and the noncircular stadium billiard. Such systems generally have infinitely hard convex surfaces or hard surfaces and irregular shape. The hard surface makes the Hamiltonian nonsmooth. Nonintegrable systems with smooth Hamiltonians comprise the second class of nonintegrable system. The vast majority of mechanical systems belong to this second class. They generally contain a mixture of quasiperiodic (KAM) orbits and chaotic orbits and a mixture of stable and unstable periodic orbits. In such systems, nonlinear resonances can occur between various degrees of freedom. Quasiperiodic orbits lie on lower-dimensional surfaces in regions of phase space where degrees of freedom are not strongly coupled. Resonances cause enhanced transfer of energy between various degrees of freedom and change the topo- logical structure of local regions of the phase space. Overlapping resonance regions destroy quasiperiodic orbits and create chaos. The mechanism by which resonances destroy KAM tori in classical sys- tems was clarified by Greene [Greene 1979]. For systems with two degrees of freedom, each KAM torus can be defined uniquely by its irrational winding6 1. Overview number. Each irrational winding number can be approximated uniquely by a sequence of rational fractions given by the continued fraction for the irrational winding number. Each of these rational fractions is associated with a resonance zone, called a rational approximate to the KAM torus. Kadanoff and Shenker [Shenker and Kadanoff 1982] and MacKay [MacKay 1983] were able to show that at the parameter value at which a given KAM torus (with quadratic irrational winding number) is destroyed, the ratio- nal approximates have self-similar structure and the areas in phase space that they occupy are related by scaling laws. They also showed that the rational approximates play a dominant role in the destruction of KAM tori. The work of Greene, Kadanoff and Shenker, and MacKay focused on area- preserving maps. Escande and Doveil [Escande and Doveil 1981] developed a renormalization theory for destruction of KAM tori directly from the Hamiltonian for systems with two degrees of freedom. Thus, Hamiltonian systems, much like equilibrium systems near a phase transition, can exhibit self-similar structure. Much of the behavior that occurs in classical systems also occurs in their quantum counterpart. However, because of the Heisenberg uncertainty re- lations, we are forced to describe classical and quantum systems from quite different perspectives. In classical systems, we can examine the evolution of individual orbits in phase space, and we can see directly the chaotic flow of trajectories in phase space. If we were to describe the evolution of the classical system in terms of the probability distribution in phase space, using the Liouville equation, we would have to search for the signatures of chaos in the behavior of the probability distributions and eigenvalues of the Liouville operator. This has been done for very simple chaotic maps [Driebe 1999], but it is a formidable task when dealing with Newtonian mechanical systems with two or more degrees of freedom. When we study quantum systems, we have no phase space in which to describe the evolution of individual orbits because of the Heisenberg uncer- tainty relations. A single quantum state occupies volume of order A in the classical phase space, where h is Planck’s constant and N is the number of degrees of freedom. We are forced from the outset to study quantum sys- tems at the level of a linear probability (probability amplitude to be more precise) equation, namely the Schrédinger equation. As we shall show in this book, most of the mechanisms at work in nonlinear classical systems are also at work in their quantum counterparts. For example, nonlinear resonances exist in quantum systems and can destroy constants of the mo- tion (good quantum numbers) in local regions of the Hilbert space. They form self-similar structures, but only down to scales of order hY and not to infinitely small scales as they do in classical systems. However, because the Schrddinger equation is an equation for probability amplitudes rather than probabilities, we will find some new phenomena that can occur in quantum. systems but not in classical systems.1.2. Historical Overview 7 One of the most important discoveries of quantum chaos theory is that the statistical properties of energy spectra and scattering delay times indi- cate that the information content of a quantum system is extremized as its classical counterpart undergoes a transition to chaos. The idea of study- ing the spectral statistics of quantum systems is largely due to Wigner, who in the 1950’s analyzed the statistical properties of nuclear scattering resonances. It was found that the nearest neighbor spacing of scattering resonances, for some nuclear scattering processes, has a distribution that agrees with the distribution of spacings of eigenvalues of ensembles of ran- dom Hermitian matrices (the Gaussian ensemble) whose matrix elements extremize information. The work of Wigner led Dyson [Dyson 1962] to study the statistical properties of ensembles of random unitary matri- ces (the circular ensembles) that extremize information. The connection between chaos theory and random matrix theory was made in 1979 by McDonald and Kaufman [McDonald and Kaufman 1979], who found that classically chaotic quantum billiards have spectral spacing distributions given by the Gaussian ensembles. Comparison between statistical proper- ties of deterministic quantum systems with underlying classical chaos and predictions of random matrix theories that extremize information is now a standard tool of quantum mechanics. For this reason, in this book we give a systematic and complete grounding in those aspects of random matrix theory and supersymmetry theory necessary to analyze quantum chaotic systems, and we use them to analyze the manifestations of chaos in both bounded systems and open quantum systems (scattering systems). In quantum systems, destruction of constants of the motion (good quan- tum numbers) due to resonance overlap occurs in local regions of Hilbert space and can cause the wave function of a particle to spread throughout the chaotic regions of the phase space. This can cause a change in the physics of the quantum system from insulator to conductor, or if we are discussing the behavior of an atom in a microwave field, from a nonionized to ionized atom. However, in quantum systems an additional phenomenon, called dy- namic Anderson localization, can occur that can restrict the spread of the wave function in chaotic regions. In the early days of quantum mechanics, before the work of Heisenberg and Schrédinger, the quantum version of a classical system was obtained by quantizing the action variables. This is straightforward if the classical system is integrable and one can find the action variables. However, Ein- stein, who knew of the work of Poincaré, as early as 1917 [Einstein 1917] pointed out that there may be difficulties with this method of quantization if invariant tori do not exist in the classical phase space, as is the case with chaotic systems. Indeed, until the work of Gutzwiller in the early 1980s {Gutzwiller 1982], there was no way to link classically chaotic systems to their quantum counterparts. However, Gutzwiller showed that Feynman path integrals, in the semiclassical limit, provide such a link, and the spec- tral properties of a quantum system, whose classical counterpart is chaotic,8 1. Overview are determined largely in terms of an infinite sum over the unstable peri- odic orbits of the classical system. Berry [Berry 1985] showed that these same periodic orbits also influence the range of fluctuations that can occur in the energy spectrum of the quantum system. Probably the most widely studied systems, as regards the transition to chaos, are systems driven by time-periodic external fields. With a time- periodic force, one can cause a nonlinear system, with only one degree of freedom, to undergo a transition to chaos. In such systems, energy is not conserved but due to a discrete time translation invariance, the Floquet energy is conserved. Thus all the techniques used in energy-conserving sys- tems can be applied to these driven systems. For classical time-periodic systems it is often possible to derive maps that greatly facilitate the study of the transition to chaos. For both classical and quantum systems it is possible to derive a renormalization theory based on self-similar networks of nonlinear resonances that exist in such systems (there is now some ex- perimental evidence that nonlinear resonances in quantum systems form a self-similar structure). In classical systems, this renormalization theory facilitates the study of the breakdown of KAM tori and scaling behavior of phase space flow through Cantor set, structures (cantori). In quantum systems, it facilitates the study of scaling of localization lengths in regions of underlying mixed phase space. 1.3. Plan of the Book This book provides a thorough discussion of classical chaos theory and quantum chaos theory, and those aspects of random matrix theory useful for the analysis of quantum systems. We begin in Chapter 2 with an introduction to basic concepts. We will prove Noether’s theorem, which relates symmetries to constants of the mo- tion, and we will define the concept of integrability. We will also give examples of nonlinear resonances and show that they cause topological changes in the structure of phase space flow and cause perturbation ex- pansions to diverge in those regions. Chaotic orbits appear in regions of phase space where resonance zones overlap. In chaotic regions, neighbor- ing trajectories move apart exponentially in time with a rate given by the Lyapounov exponents. In Chapter 2, we define Lyapounov exponents and define chaos in terms of them. We conclude Chapter 2 by generalizing our discussion to include systems with time-periodic Hamiltonians. Chapter 3 is devoted entirely to area-preserving maps called twist maps. ‘These represent surfaces of section of the classical phase space for systems with N = 2 degrees of freedom. In Chapter 3, we describe the difference be- tween integrable and nonintegrable twist maps in terms of the Birkhoff fixed point theorem. We distinguish between stable and unstable fixed points and1.3. Plan of the Book 9 show how to determine the stability properties and points of bifurcation of fixed points in terms of the tangent map, which is a linear mapping of the neighborhood of the fixed point. We show the very complex behavior of the stable and unstable manifolds of unstable fixed points as chaos sets in in their neighborhood. In Chapter 3, we also describe the self-similar structure that exists in the neighborhood of certain KAM tori (those with quadratic irrational winding number) just as they break and form a Can- tor set structure (cantorus) in phase space. Also in Chapter 3, we describe the diffusion processes that occur in a fully chaotic sea, in the presence of accelerator modes, and in the neighborhood of stable islands. When we work with Hamiltonian systems, it is usually not possible to construct twist maps that represent surfaces of section for those systems. However, there is still a great deal we can determine about the system if we know the Hamiltonian. We can often locate the positions of primary resonances in phase space. Once this is done, we can then begin to locate the whole infinite hierarchy of higher-order nonlinear resonances and derive conditions for the destruction of KAM tori between them. Self-similarity in the classical phase space allows us to construct a renormelization map that can give very accurate estimates for parametric values at which certain KAM tori are destroyed. In open classical systems, when a particle scatters from a chaotic region of the phase space, we often observe delay times for the scattered particle that have a fractal distribution. There are many systems that do not satisfy the conditions of the KAM theorem. When that happens in systems with two degrees of freedom, a stochastic web can form that dynamically “tiles” the phase space. In Chapter 4, we also describe the resonance structure of systems with three or more degrees of freedom. We shall see that in such systems resonance zones form a web (Arnol’d web) or interconnected network in the phase space. For systems with more than two degrees of freedom, a trajectory can reach any neighborhood of the phase space by diffusing along this web. However, the rate of diffusion might be very slow. Chapter 5 is devoted to the discussion of random matrix theory. Ran- dom matrix theory is based on the assumption that the matrix elements of a hermitian or unitary matrix are independent random variables. One can define an invariant measure on the space of matrix elements and then determine the probability distribution of the matrix elements by requir- ing that the information contained in the matrix elements be extremized. This leads to a Gaussian distribution for Hermitian matrix elements (the Gaussian ensemble) and a uniform distribution for unitary matrix elements (the circular ensemble). The eigenvalues of these ensembles of random ma- trices are correlated and repel. Once the joint probability distribution of energy eigenvalues is known, we can determine some statistical properties of energy-level sequences. In Chapter 5, we focus on the two most widely used statistics, the nearest neighbor eigenvalue spacing distribution and the Asg-statistic, the latter of which determines the mean square deviation of10 1. Overview the spectral staircase function from a straight line. These two statistics can be quite different for integrable and nonintegrable quantum systems. We also look at the distribution of elements of the eigenvectors of the random matrices. In Chapter 6, we discuss the behavior of some bounded quantum systems whose classical counterparts undergo a transition to chaos. The Schrédinger equation for these systems is linear. Nonlinearities appear in the Hamil- tonian. We describe how the appearance of level repulsion in the energy spectrum may be associated with the destruction of underlying symmetries. We analyze the statistical properties of several different types of chaotic bil- liards, some of which have been realized in microwave cavity experiments. We also show how chaos manifests itself in quantum spin systems and anharmonic oscillator systems. ‘A research area that has grown considerably since the first edition of this book is the study of the behavior of quantum particles that scatter from chaotic reaction regions. In Chapter 7, we focus on the behavior of quantum waveguides and derive an expression for the scattering matrix (S-matrix) based on the reaction matrix theory of scattering first developed by Wigner and Eisenbud [Wigner and Eisenbud 1949]. This approach to scattering theory is especially useful for the analysis of chaotic quantum scattering because it also allows straightforward comparison between deterministic scattering processes and scattering processes governed by random Hamil- tonian matrices. In Chapter 7, we discuss properties of scattering delay times, and we discuss properties of the averages of the S-matrix and S- matrix correlation functions using both the Gaussian orthogonal ensemble and the circular orthogonal ensemble. We also derive a relation between the S-matrix and the energy Green’s function for a scattering system, and show that poles of the S-matrix give valuable information about the absorption spectrum of a system. Finally, at the end of Chapter 7, we show a variety of experimental and numerically obtained data on nuclear and molecular energy-level sequences obtained from scattering data. We show that for “pure sequences,” which are sequences with a fixed spin and parity (the known good quantum numbers), there is a lack of close spacings. This is one indicator that the internal dynamics of these systems is nonintegrable and likely is showing the manifestations of chaos. ‘The connection between classically chaotic systems and their quantum counterpart is shown in Chapter 8, where we use semiclassical path in- tegrals to obtain the spectral properties of quantum systems. We begin with a derivation of Feynman path integrals and then obtain the semiclas- sical limit by a series of approximations that involve neglecting terms of order h/? and smaller (h is Planck’s constant) in the path integral. The semiclassical path integral can then be expressed in terms of an infinite sum over classical paths (paths that extremize the action). We apply the path integral to several problems. We first obtain the spectrum for a quan- tum particle in a smooth potential well and we obtain the expected WKB1.3. Plan of the Book i (Wentzel-Kramers-Brillouin) energies for the particle. We then compute the staircase function for an integrable billiard and show that the periodic orbit with the shortest period determines the longest wavelength fluctua- tion contained in the staircase function. We also derive the Gutzwiller trace formula, which expresses the trace of the Green’s function of a quantum system in terms of periodic orbits of the classical system. We show that the trace formula gives very good results for the energy levels of the anisotropic Kepler system, a classically chaotic system. Finally, we conclude Chapter 8 with a numerical and experimental study of the influence of periodic orbits on the absorption spectrum of diamagnetic hydrogen. Chapter 9 is devoted to periodically driven quantum systems. We first derive the Floquet theory for such systems and apply it to studies of quan- tum nonlinear resonance overlap and the effect of level repulsion on the spread of wave functions in chaotic systems. We also use Floquet analy- sis to interpret the results of recent dynamic tunneling experiments using cold atom optics. It is possible to develop a quantum renormalization map based on the self-similarity of nonlinear resonance in quantum systems, We use this map to investigate the scaling properties of localization lengths in quantum systems with underlying chaos. We next focus on the quantum delta-kicked rotor, which was the first system in which dynamic Anderson localization was observed numerically. We describe dynamical Anderson localization in an atom optics experiment that approximates the delta- kicked rotor. We also describe extensive experiments on microwave-driven hydrogen that confirm the existence of higher-order nonlinear resonances in quantum systems and also show dynamic Anderson localization. Finally, in Chapter 10, we show by means of several examples that many of the ideas discussed in the previous chapters may also extend to stochastic systems. This book contains seventeen appendices that give additional background on subjects of importance to this book. For example, there is a review of the effect of symmetries on the structure of Hamiltonian matrices. There is a derivation of the measures for Hermitian and unitary matrices used in random matrix theory. There is a derivation of the normalization constants and expressions for probability distributions of the Gaussian and circular ensembles in terms of quaternion matrices. There is a review of properties of Grassmann variables and the application of supersymmetry techniques in deriving averages of the Green’s function and S-matrices. There are other appendices as well that will aid the reader with some of the theory concepts in this book. We do not have room in this book to discuss in detail all of the interest- ing applications of conservative chaos theory, so in the concluding section of each chapter we have given references to additional topics of interest. Also at the end of most chapters, we have given some problems that may illustrate and clarify concepts.12 1. Overview 1.4 References Arnol’d, V.1. (1963): Russ. Math. Surv. 18 9; 18 85. Berry, M.V. (1985): Proc. Roy. Soc. London A 400 229. Driebe, D.J. (1999): Fully Chaotic Maps and Broken Time Symmetry (Kluwer Academic Publishers, Dordrecht). Dyson, F.J. (1962): J. Math. Phys. 3 140. Einstein, A. (1917): Verh. Dtsch. Phys. Ges. 19 82. Escande, D.F. and Doveil, F. (1981): J. Stat. Phys. 26 257. Greene, J. (1979): J. Math. Phys. 20 1183. Gutzwiller, M.C. (1982): Physica D 5 183. Hamilton, W.R. (1940): The Mathematical Papers of Sir William Rowan Hamil- ton; Vol. II, Dynamics, edited by A.W. Conway and J.L. Synge (Cambridge University Press, Cambridge U.K.). Kolmogorov, A.N. (1954): Dokl. Akad. Nauk. SSSR 98 527 (1954) (An English version appears in R. Abraham, Foundations of Mechanics (W.A. Benjamin, New York, 1967, Appendix D). Lagrange, J.L. (1889): Mechanique Analytique (Gauthier-Villars, Paris). Laplace, P.S. (1951): A Philosophical Essay on Probabilities, translated by F.W. ‘Truscott and F.L. Emory (Dover, New York) Lighthill, J. (1986): Proc. Roy. Soc. London A 407 35. MacKay, RS. (1983): Physica D 7 283. McDonald, S.W. and Kaufman, A.N. (1979): Phys. Rev. Lett. 42 1189. Moser, J. (1968): Nachr. Akad. Wiss. Goettingen II, Math. Phys. Kd 1 1. Moser, J. (1975): Is the Solar System Stable? Neue Zurcher Zeitung, May 14, 1975. Noether, B. (1918): Nach. Ges. Wiss. Goettingen 2 235. Shenker, $.J. and Kadanoff, L.P. (1982): J. Stat. Phys. 27 631. Toda, M. (1981): Theory of Nonlinear Lattices (Springer-Verlag, Berlin). Wigner, E.P. and Bisenbud, L.B. (1949): Phys. Rev. 72 29.2 Fundamental Concepts 2.1 Introduction There are three basic concepts that are essential for understanding the dynamical behavior of nonlinear conservative systems. The first is the con- cept of global symmetries, which serve to constrain the dynamical flow of the system to lower-dimensional surfaces in the phase space. Some of these global symmetries are obvious and are related to the space-time symmetries of the system. Others are not obvious and have been called hidden symme- tries by Moser [Moser 1979]. When there are as many global symmetries as degrees of freedom, the dynamical system is said to be integrable. The second important concept is that of nonlinear resonance. As Kolmogorov [Kolmogorov 1954], Arnol’d [Arnol’d 1963], and Moser [Moser 1962] have shown, when a small symmetry-breaking term is added to the Hamilto- nian, most of the phase space continues to behave as if the symmetries still exist. However, in regions where the symmetry-breaking term allows resonance to occur between otherwise uncoupled degrees of freedom, the dynamics begins to change its character. When resonances do occur, they generally occur on all scales in the phase space and give rise to an incred- ibly complex structure, as we shall see. The third important concept is that of chaos or sensitive dependence on initial conditions. For the class of systems in which symmetries can be broken by adding small symmetry- breaking terms, chaos first appears in the neighborhood of the nonlinear resonances. As the strength of the symmetry-breaking term increases and142. Fundamental Concepts the size of the resonance regions increases, ever larger regions of the phase space become chaotic. ‘As we shall show in Sect. 2.2, the dynamical evolution of systems with broken symmetry cannot be determined using conventional perturba- tion theory, because of the existence of nonlinear resonances. This occurs because nonlinear resonances cause a topological change locally in the structure of the phase space, and conventional perturbation theory is not adequate to deal with such topological changes. In Sect. 2.3, we introduce the concept of integrability. A system is inte- grable if it has as many global constants of the motion as degrees of freedom. ‘The connection between global symmetries and global constants of motion was first proven for dynamical systems by Noether [Noether 1918]. We will give a simple derivation of Noether’s theorem in Sect. 2.3. In Sect. 2.3, we illustrate these methods for the simple three-body Toda lattice. It is usu- ally impossible to tell if a system is integrable or not just by looking at the equations of motion. The Poincaré surface of section provides a very useful numerical tool for testing for integrability and will be used throughout the remainder of this book. We will illustrate the use of the Poincaré surface of section for the classic model of Henon and Heiles [Henon and Heiles 1964]. In Sect. 2.4, we introduce the concept of nonlinear resonances and il- lustrate their behavior for some simple models originally introduced by Walker and Ford {Walker and Ford 1969]. These models are interesting be- cause they show that resonances may appear or disappear as parameters of the system are varied and the overlap of nonlinear resonances leads to the onset of chaos. Conventional perturbation theory does not work when nonlinear resonances are present. But Kolmogorov, Arnol’d, and Moser (collectively called KAM) have developed a rapidly converging perturba- tion theory that can be used to describe nonresonant regions of the phase space, precisely because it is constructed to avoid the resonance regions. KAM perturbation theory will be described in Sect. 2.5. In practice, chaos is defined in terms of the dynamical behavior of pairs of orbits that initially are close together in the phase space. If the orbits move apart exponentially in any direction in the phase space, the flow is said to be chaotic. The rate of exponential divergence of pairs of orbits is measured by the so-called Lyapounov exponents. There will be one such exponent for each dimension in the phase space. If all the Lyapounov exponents are zero, the dynamical flow is regular. If even one exponent is positive, the flow will be chaotic. A detailed discussion of the behavior of Lyapounov exponents for conservative systems is given in Sect. 2.6 and is illustrated in terms of the Henon-Heiles system. Systems with positive Lyapounov exponents also have positive KS metric entropy. The KS metric entropy is defined in Sect. 2.6 and computed for the baker’s transformation, one of the simplest known chaotic dynamical systems. Much of the work done on the transition to chaos in conservative sys- tems has been done on one degree of freedom conservative systems driven2.2. Conventional Perturbation Theory ‘ 15 by a time-periodic external field. Such systems are conservative in a higher- dimensional phase space and are volume-preserving. They are particularly easy to study both analytically and numerically because the location of resonances is largely determined by the structure of the unperturbed sys- tem, and Poincaré surfaces of section are strobe plots. In Sect. 2.7, we describe the mechanism by which chaos occurs in the conservative Duff- ing system, which consists of a particle in a double well potential driven by a monochromatic external time-periodic field. We will see clearly why a chaotic region (stochastic layer) always forms at the separatrix of nonlinear resonance zones in nonintegrable systems. 2.2 Conventional Perturbation Theory Historically, the difficulties in obtaining long-time predictions for the evo- lution of mechanical systems was brought into focus with Poincare’s proof that conventional perturbation expansions generally diverge and cannot be used as a tool to provide long-time prediction. In order to build some in- tuition concerning the origin of these divergences, let us consider a system with two degrees of freedom having a Hamiltonian Ho(p1,p2, 41, 42), which after a canonical transformation can be written in terms of action-angle variables (Ji, Jo, 1,02) in the form Ho(Ji, Jz). For example, the Hamilto- nian for the relative motion of a moon of mass mi, orbiting a planet of mass mp (the Kepler system), can be written =F, (2.2.1) where (p,,p¢) and (r,@) are the relative momentum and positions, respec- tively, of the two bodies in polar coordinates, E is the total energy of the system, p= 24" is the reduced mass, and k = Gmima (G is the grav- itational constant). The total angular momentum, L, is conserved for this problem so the plane of motion, (r, ), is taken to lie in the plane perpendic- ular to L. After a canonical transformation from coordinates (p,,p3,7,¢) to action-angle coordinates (.J,, J2,01,02), the Hamiltonian takes the form [Goldstein 1980] = pk? Mo= + bP (2.2.2) The motion is fairly complicated (elliptic or hyperbolic orbits) in terms of coordinates (p,,p4,7,¢), but in terms of action-angle coordinates it is simple. Hamilton’s equations of motion yield dj, _ _OHo “=~ 30, = (2.2.3)16 2. Fundamental Concepts Figure 2.2.1. For integrable systems with two degrees of freedom, each trajectory lies on a torus constructed from the action-angle variables (Ji, J2,01,62). The radii of the torus are p) = /2J; for i = (1,2). If the frequencies wj = (i = 1,2) are commensurate, the trajectory will be periodic. If the frequencies are incommensurate, the trajectory will never repeat. and ef = ae = wi(Sh, Je), (2.2.4) where i = (1,2) and t is the time. Thus, we find that K=q (2.2.5) and 0; = wit +d;, (2.2.6) where c; and d; are constants determined by the initial conditions. We see immediately that the energy of this system is constant. It is useful to picture the motion of this system as lying on a torus as shown in Fig. 2.2.1. The torus will have two constant radii, which we define as pj = /2J; for i = (1,2), and two angular variables (6;,02). A single orbit of the Kepler system will evolve on this torus according to Eqs. (2.2.5) and (2.2.6). Notice that there are two frequencies associated with this system, w; and wy. If these two frequencies are commensurate (that is, if mw; = nw, where m and n are integers), then the trajectory will be periodic and the orbit: will repeat itself. If the two frequencies are incommensurate (irrational multiples of one another), then the trajectory will never repeat itself as it moves around the torus and eventually will cover the entire surface of the torus. Note also that the frequencies themselves depend on the action variables and therefore on the energy of the system. This is a characteristic feature of a nonlinear system. Let us now assume that a perturbation acts in the plane of motion due to the presence of another planet. We shall treat this perturbation as an2.2. Conventional Perturbation Theory os external field. In the presence of this perturbation, the Hamiltonian will take the form H = Ho(Ai, Jo) + V (Ji, Ja, 01,02), (2.2.7) where € is a small parameter, ¢ < 1. We wish to find corrections to the unperturbed trajectories, J; = c;, due to the perturbation. Since we cannot solve the new equations of motion exactly, we can hope to obtain approx- imate solutions using perturbation expansions in the small parameter ¢. Let’s try it. First we note that since we are dealing with periodic bound state motion, we can expand the perturbation in a Fourier series. We then write the Hamiltonian in Eq. (2.2.7) in the form H = Ho(sisJ2)+€ D> > Vassna (dis Je) cos(ni01 +262). (2.2.8) my =—o0ng=—00 Next, we introduce a generating function, G(Ji, Je, 91,02), which we define as Gi, Fa, 01,82) = SiO + Jobo +€ SYS Gnine(Ay Je) sin(m191 + 2262), (2.2.9) ae ee where gn,n. Will be determined below. The generating function in Eq. (2.2.9) generates a canonical transformation from the set of action-angle variables, (Ji, J2,61,62), to a new set of canonical action-angle variables, (Si, Ja, 01, 2), via the following equations [Goldstein 1980]: Bs 5g, 7 te > y NaGny.ny CO8(N181 + n282) (2.2.10) my=—o0 nes and oG a SBawre mn: = 55 =H +e > 7 Sm ‘Gra sna Ryesin math + mabe). (2.2.11) The new Hamiltonian, H’(%1,J2, 91,92), is obtained from Eq. (2.2.8) by solving Eqs. (2.2.10) and (2.2.11) for (Jj,0;) as a function of (J;,0;) and plugging into Eq. (2.2.8). If we do that and then expand H'(.Ji, J2, 1,2) in a Taylor series in the small parameter ¢, we find H(i, Ja, 01,02) =Hi(Iirh) + Ss yi naw + 2W2) Gn, nz C08(M1O1 + n2O2) m=—00 m= rr rr Lr Llc Err —— mj=—00 ng=—0018 2, Fundamental Concepts where _ oH, ~ OT: Now remove terms of order € by choosing Evie nd) Grima = meas) (2.2.14) wi (2.2.13) Then H' (Si, Ja, 1, O2) = Hi( Ai, Je) + O() (2.2.15) and k=5-- OD niVnsina Co9(1©1 +7282) § O(2) (9.0.16) mic toons= (raw + nawe) To lowest order in ¢, this is the solution to the problem. New actions, J, have been obtained that contain corrections due to the perturbation. If, for example, € = 0.01, then by retaining only first-order corrections we neglect terms of order e? = 0.0001. To first order in ¢, J is a constant and 0; varies linearly in time. This is the hope. However, there is a catch. For any of this to have meaning, we must have [mw + nga] >> €Vins na (2.2.17) But the condition in Eq. (2.2.17) breaks down when internal nonlinear res- ‘onances occur and cause the perturbation expansion to diverge. Poincaré showed that it is a general property of perturbation expansions of this type that they can be expected to diverge. 2.3. Integrability A concept that is essential to the remainder of this book is that of integra- bility. Let us consider a system with N degrees of freedom. Its phase space has 2N dimensions. Such a system is integrable if there exist N independent isolating integrals of motion, J;, such that Fi(pi,--+s PN, Gs-+-5dw) = Cis (2.3.1) for i =1,..., N, where C; is a constant and p; and q are the canonical momentum and position associated with the ith degree of freedom. The functions J; are independent if their differentials, dj, are linearly inde- pendent. It is important to distinguish between isolating and nonisolating integrals [Wintner 1947]. Nonisolating integrals (an example is the initial coordinates of a trajectory) generally vary from trajectory to trajectory and usually do not provide useful information about a system. On the other hand, isolating integrals of motion, by Noether’s theorem, are due to2.3. Integrability . 19 symmetries (some “hidden”) of the dynamical system and define surfaces in phase space. The condition for integrability may be put in another form. A classical system with N degrees of freedom is integrable if there exist N indepen- dent globally defined functions, 1;(p1,-..,PNv,4i,---)4v); for i=1,..., N, whose mutual Poisson brackets (see Appendix A.4) vanish, {i,j} Poisson = 0, (2.3.2) fori = 1,..., N and j = 1,..., .N. Then the quantities J; form a set of N phase space coordinates. In conservative systems, the Hamil- tonian, H(pi,..-,pn.41,---,4n), Will be one of the constants of the motion. In general, the equation of motion of a phase function, f = f(P1,---;PNs41y +++ .4N; 8), is given by df _ of dt Ot Thus Eqs. (2.3.2) and (2.3.3) imply that 44 = 0. If a system is integrable, there are no internal nonlinear resonances leading to chaos. All orbits lie on N-dimensional surfaces in the 2V-dimensional phase space. + {H, f} Poisson- (2.3.3) 2.3.1 Noether’s Theorem As was shown by Noether [Noether 1918], isolating integrals result from symmetries. For example, the total energy is an isolating integral (is a constant of the motion) for systems that are homogeneous in time (invari- ant under a translation in time). Total angular momentum is an isolating integral for systems that are isotropic in space. Noether’s theorem is generally formulated in terms of the Lagrangian (see [Goldstein 1980] and Appendix A). Let us consider a dynamical system with N degrees of freedom whose state is given by the set of generalized velocities and positions ({4;},{q:}). Let us consider a system whose La- grangian, L = L({q:}, {q:}), is known. For simplicity, we consider a system with a time-independent Lagrangian. The equations of motion are given by the Lagrange equations OL d (aL oq dt Cs For such systems, Noether’s theorem may be stated as follows. ) =o (i=1,...,N). (2.3.4) Noether’s theorem If a transformation tot =t+6t, gilt) a) =alt)+ Salt), and 4G > Gt) = Galt) + 54:(t)20 2. Fundamental Concepts (for i La}, (ail) > LAE AGED = LAO} GED), (2.3.5) and leaves the action integral invariant NN) leaves the Lagrangian form invariant, [raraae {ae [axtao)tao)=0 236) then there exists an isolating integral of motion associated with this symmetry transformation. ¢ Before we proceed to show this, we must distinguish between variations of the coordinates at a fixed time, gi(t) > gf jee = alt aes (t) and variations at a later time (as we indicated above) qi(t) — 4{(t!) = gi(t) + 6q:(t). 54: (t) is a convective variation and differs pa 6Qi(t) by a convective term, 6q:(t) = 6Qi(t) + i6t [Reichl 1998). Proof of Noether’s theorem Let us write Eq. (2.3.6) in the form ta+5ta rt [ dt LLG}, {4} a dt L({ai(t)}, {ai()}) =0, (2.3.7) t+6ty th where on the leftmost integral we have let the dummy variable t/ + ¢. Next let {4(t)} = {ai(4) + 6Qi(t)} and {a} = {ai(t) + 6Qi()}, and expand the integral to first order in the variations. We then find [ra eta toto +3 [ (5p) + (ae) 82] } . £ aL ({ai(0)}. {a(0)}) = 0. (238) If we next keep only first-order contributions in the variations in the limits of integration, we find £af5 [a+ (@el} + dt2L(te) — 6t,L(t) = 0, : (2.3.9) where L(t.) = L({qi(tk)}; {gi(tx)}). Equation (2.3.9) can now be rewritten in the form L dt {a om) 4d (FF x) 80+ (=) cal}. (23.10)2.3, Integrability + 21 Let us now make use of Lagrange’s Eqs. (2.3.4) and note that 6Q; = £6Q;. Then, after some rearrangement of terms, we find [od ue (G0)00}- Let us now rewrite Eq. (2.3.11) in terms of our convective variations. We then find Coil So @lek Ds} 4s @JoEs}- and we have obtained an isolating integral as a result of our symmetry transformation. e (2.3.11) (2.3.12) . (2.3.13) To illustrate the use of Eq. (2.3.13), let us consider some examples. As- sume that we translate the system in time by a constant amount, ét = «, but let 64; = 0. Then we have $f - De (5)\-¢ 0 (2.3.14) since the quantity in curly brackets is the Hamiltonian (see Appendix A). Thus homogeneity in time gives rise to the Hamiltonian as an isolating integral and to energy conservation. Suppose that we let 6t = 0 but trans- late one coordinate, q,, by a constant amount, 64; = ¢6;,;, where 6;,; is the Kronecker delta. Then we find ea a (Fe) =F =0. (2.3.15) Thus the generalized momentum associated with the degree of freedom, q, is an isolating integral, and the component of the momentum, pj, is conserved. The variations could, in general, be functions of space or time. Then the isolating integrals resulting from the symmetry transformation would be much more complicated. However, few such isolating integrals are known aside from the ones due to the space-time symmetries. 2.3.2 Hidden Symmetries In order for a system to be integrable, it must have as many conserved quantities as there are degrees of freedom. In general, not all of these can come from the space-time symmetries but may come from what Moser has called hidden symmetries [Moser 1979]. One notable example of such22 2. Fundamental Concepts a hidden symmetry occurs for the two-body Kepler problem. Because of the homogeneity of this system in time and space, the total energy and the center-of-mass momentum are conserved. In addition, the gravitational force is a central force and therefore this system exhibits isotropy in space, which means that the total angular momentum is also conserved. These space-time symmetries are sufficient to make this system integrable since they provide six conservation laws for the six degrees of freedom. However, there is still another conserved quantity, the Laplace-Runge-Lenz vector r A=pxL-pk trl (2.3.16) [Moser 1970], [Abarbanel 1976], [Goldstein 1980], where p is the relative momentum, L is the total angular momentum, : and k are as defined in Sect. 2.2, and r is the relative displacement of the two bodies. This additional symmetry is responsible for the fact that there is no precession of the perihelion (the point of closest approach of the two bodies) for the two-body Kepler system. This conservation law does not hold for any other central force problem. Hidden symmetries underlie the relatively new field of soliton physics. One type of soliton, the nontopological soliton, occurs in integrable dy- namical systems and is most commonly found in continuous media and on length scales where the underlying discreteness of matter plays no role. There is one mechanical system with a finite number of degrees of freedom, however, that is now known to support solitons. That is the N-body Toda lattice [Toda 1967], [Toda 1981]. The Toda lattice is a collection of equal- mass particles coupled in one dimension by exponentially varying forces. It is integrable and therefore has N isolating integrals of the motion. The Toda lattice is one of the few discrete lattices for which soliton solutions are exact. The continuum limit of the Toda lattice yields the Korteweg-de Vries equation, which is the classic equation describing nontopological soli- tons in continuum mechanics. The first real indication that the Toda lattice was integrable came from numerical experiments by Ford et al. [Ford et al. 1973]. This prompted theoretical work by Henon [Henon 1974] and Flaschka [Flaschka 1974], who found expressions for the N isolating integrals of the motion. The actual solution of the equations of motion was due to Date and Tanaka (Date and Tanaka 1976], although significant contributions were made by Kac and van Moerbeke [Kac and van Moerbeke 1975]. If we use techniques from soliton physics, it is fairly easy to show that the Toda lattice is integrable. Let us demonstrate this for the three-body Toda lattice. For a periodic one-dimensional lattice, the Hamiltonian can be written Bp +B B ey (lore pew) penn) 9), (2.8.17)2.3. Integrability + 23 This system has three degrees of freedom since the three masses move in one spatial dimension. The equations of motion are HT R= a = (e7 (4-441) — @(Gi-2-48)) (2.3.18) and . OH d= 5, =P (2.3.19) where i = 1, 2, 3, and gis = qi, Divs = pi due to the periodicity of the lattice. Following Flaschka (Flaschka 1974], let us make a noncanonical transformation to new variables ({a;}, {b;}) for (i = 1,2,3), where ay = Fer Borman) and b= (2.3.20) Let us now introduce the symmetric matrix Z br a1 as A) =[( a be a (2.3.21) a3 a2 by and the antisymmetric matrix 7 0 a as Bi)=|{ -a 0 a |}. (2.3.22) ag —a, 0 ‘The equations of motion can then be written in the form AQ tas aap = BAW - ADB. (2.3.23) The matrices A(t) and B(t) are called Lax pairs [Lax 1968]. They are functions of the canonical coordinates, ({pi,qi}), and, therefore, will vary in time. The Hamiltonian, H, is related to the trace of A?(t), a at ‘Tr A?(t) = BF +03 + 83 + 2(a? + a3 +2) = 5(H +3). (2.3.24) Since this is a conservative system, ‘Tr A(t) is independent of time. Toda type lattices are the only known three-body mechanical systems for which Lax pairs can be constructed. The fact that Eq. (2.3.23) holds automatically means that the three-body Toda lattice is integrable. We can see this as follows. Let us introduce yet another matrix, O(t), which is a solution of the equation dt) _ Bt)O(t). (2.3.25) dt Since B(t) is antisymmetric, O(t) is orthogonal. That is, OT (t) = O-1(t), where O7(t) is the transpose and O-1(t) is the inverse of O(t). We also24 2. Fundamental Concepts can write ao =-O71(t)B(t). (2.3.26) Using Eqs. (2.3.25) and (2.3.26), we can write Bq. (2.3.23) as ee) i £08 6-*@) + oo 2230, (2.3.27) where the matrix Z is defined as £L=07\(t)A(t)O(t). (2.3.28) Note that Eqs. (2.3.27) and (2.3.28) indicate that £ is independent of time. O(t) may be thought of as an evolution operator that propagates A(t) in time so that £ = A(0). Let us now write AO) = AHHH), (2.3.29) where A(t) and (2) are the eigenvalues and eigenvectors, respectively, of A(t). Then, from Eq, (2.3.28) we can write LO*(t) A(t) = A(t)O~* (S(t). (2.3.30) ‘Thus X(t) is an eigenvalue of £ and A(t) and therefore must be independent of time (i.e., A(t) = , where is a constant). If we let ; (i = 1,2,3) denote the three time-independent eigenvalues of the time-dependent matrix A(t), then from Eq. (2.3.24) we can write the Hamiltonian in the form 3 H=23°N -3. (2.3.31) rat The eigenvalues of A(t) constitute the three independent integrals of the motion for the Toda lattice. 2.3.3 Poincaré Surface of Section How can we tell if a system is integrable or not? There is no simple way in general. For systems with two degrees of freedom, we can check numerically by constructing a Poincaré surface of section. To see how this works, let us consider a conservative system (a system with a Hamiltonian independent of time). For such systems, the energy is conserved. The Hamiltonian is then an isolating integral of the motion and can be written H(p1,P2,% 42) = E, (2.3.32) where the energy, F, is constant and restricts trajectories to lie on a three- dimensional surface in the four-dimensional phase space. jFrom Eq, (2.3.32) we can write p2 = po(pi, 1,42, £). If the system has a second isolating integral, Ta (p1, P25 M592) = Cas (2.3.33)