This action might not be possible to undo. Are you sure you want to continue?

Mathematical Models

SIMM's Classics in Applied Mathematics series consists of books that were previously allowed to go out of print. These books are re-published by SIAM as a professional service because they continue to be important resources for mathematical scientists. Editor-in-Chief Robert E. O'Malley, Jr., University of Washington Editorial Board Richard A. Brualdi, University of Wisconsin-Madison Herbert B. Keller, California Institute of Technology Andrzej Z. Manitius, George Mason University Ingram Olkin, Stanford University Stanley Richardson, University of Edinburgh Ferdinand Verhulst, Mathematisch Instituut, University of Utrecht Classics in Applied Mathematics C. C. Lin and L. A. Segel, Mathematics Applied to Deterministic Problems in the Natural Sciences Johan G. F. Belinfante and Bernard Kolman, A Survey of Lie Groups and Lie Algebras with Applications and Computational Methods James M. Ortega, Numerical Analysis: A Second Course Anthony V. Fiacco and Garth P. McCormick, Nonlinear Programming: Sequential Unconstrained Minimization Techniques F. H. Clarke, Optimization and Nonsmooth Analysis George F. Carrier and Carl E. Pearson, Ordinary Differential Equations Leo Breiman, Probability R. Bellman and G. M. Wing, An Introduction to Invariant Imbedding Abraham Berman and Robert J. Plemmons, Nonnegative Matrices in the Mathematical Sciences Olvi L. Mangasarian, Nonlinear Programming *Carl Friedrich Gauss, Theory of the Combination of Observations Least Subject to Errors: Part One, Part Two, Supplement. Translated by G. W. Stewart Richard Bellman, Introduction to Matrix Analysis U. M. Ascher, R. M. M. Mattheij, and R. D. Russell, Numerical Solution of Boundary Value Problems for Ordinary Differential Equations K. E. Brenan, S. L. Campbell, and L. R. Petzold, Numerical Solution of Initial-Value Problems in Differential-Algebraic Equations Charles L. Lawson and Richard J. Hanson, Solving Least Squares Problems J. E. Dennis, Jr. and Robert B. Schnabel, Numerical Methods for Unconstrained Optimization and Nonlinear Equations Richard E. Barlow and Frank Proschan, Mathematical Theory of Reliability Cornelius Lanczos, Linear Differential Operators Richard Bellman, Introduction to Matrix Analysis, Second Edition Beresford N. Parlett, The Symmetric Eigenvalue Problem *First time in print.

ii

Classics in Applied Mathematics (continued) Richard Haberman, Mathematical Models: Mechanical Vibrations, Population Dynamics, and Traffic Flow Peter W. M. John, Statistical Design and Analysis of Experiments Tamer Basar and Geert Jan Olsder, Dynamic Noncooperative Game Theory, Second Edition Emanuel Parzen, Stochastic Processes Petar Kokotovic, Hassan K. Khalil, and John O'Reilly, Singular Perturbation Methods in Control: Analysis and Design Jean Dickinson Gibbons, Ingram Olkin, and Milton Sobel, Selecting and Ordering Populations: A New Statistical Methodology James A. Murdock, Perturbations: Theory and Methods Ivar Ekeland and Roger Temam, Convex Analysis and Variational Problems Ivar Stakgold, Boundary Value Problems of Mathematical Physics, Volumes I and II J. M. Ortega and W. C. Rheinboldt, Iterative Solution of Nonlinear Equations in Several Variables David Kinderlehrer and Guido Stampacchia, An Introduction to Variational Inequalities and Their Applications F. Natterer, The Mathematics of Computerized Tomography Avinash C. Kak and Malcolm Slaney, Principles of Computerized Tomographic Imaging R. Wong, Asymptotic Approximations of Integrals O. Axelsson and V. A. Barker, Finite Element Solution of Boundary Value Problems: Theory and Computation David R. Brillinger, Time Series: Data Analysis and Theory Joel N. Franklin, Methods of Mathematical Economics: Linear and Nonlinear Programming, Fixed-Point Theorems Philip Hartman, Ordinary Differential Equations, Second Edition Michael D. Intriligator, Mathematical Optimisation and Economic Theory Philippe G. Ciarlet, The Finite Element Method for Elliptic Problems Jane K. Cullum and Ralph A. Willoughby, Lanczos Algorithms for Large Symmetric Eigenvalue Computations, Vol. I: Theory M. Vidyasagar, Nonlinear Systems Analysis, Second Edition Robert Mattheij and Jaap Molenaar, Ordinary Differential Equations in Theory and

Practice

Shanti S. Gupta and S. Panchapakesan, Multiple Decision Procedures: Theory and Methodology of Selecting and Ranking Populations Eugene L. Allgower and Kurt Georg, Introduction to Numerical Continuation Methods Heinz-Otto Kreiss and Jens Lorenz, Initial-Boundary Value Problems and the NavierStokes Equations J. L. Hodges, Jr. and E. L. Lehmann, Basic Concepts of Probability and Statistics, Second Edition

iii

This page intentionally left blank

Mathematical Models

Mechanical Vibrations, Population Dynamics, and Traffic Flow

An Introduction to Applied Mathematics

Richard Haberman

Department of Mathematics Southern Methodist University Dallas, Texas

Society for Industrial and Applied Mathematics Philadelphia

Copyright ©1998 by the Society for Industrial and Applied Mathematics. This SIAM edition is an unabridged republication of the work first published by Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1977.

10 9 8 7 6 5

All rights reserved. Printed in the United States of America. No part of this book may be reproduced, stored, or transmitted in any manner without the written permission of the publisher. For infor-mation, write to the Society for Industrial and Applied Mathematics, 3600 University City Science Center, Philadelphia, PA 19104-2688. Library of Congress Catalog Card Number: 97-62401 ISBN 0-89871408-7 is a registered trademark.

vi

Contents

(Starred sections may be omitted without loss of continuity)

FOREWORD PREFACE TO THE CLASSICS EDITION PREFACE

xi xiii xv

**Part 1: Mechanical Vibrations
**

1. 2. 3. 4. 5. 6. 7. 9. Introduction to Mathematical Models in the Physical Sciences Newton's Law Newton's Law as Applied to a Spring-Mass System Gravity Oscillation of a Spring-Mass System Dimensions and Units Qualitative and Quantitative Behavior of a Spring-Mass System A Two-Mass Oscillator

1

3 4 6 9 12 16 18 20 23 29 33 34 40 42 51

vii

8. Initial Value Problem 10. Friction 11. Oscillations of a Damped System 12. Underdamped Oscillations 13. Overdamped and Critically Damped Oscillations 14. A Pendulum 15. How Small is Small?

A Dimensionless Time Variable 17. Exponential Growth 35. . Linearized Stability Analysis of an Equilibrium Solution 19. Population Models 32. Growth Models with Time Delays 41. . . Phase Plane Solution of the Logistic Equation 39. Introduction to Two-Species Models 117 119 120 122 129 131 138 * 143 * 151 155 159 162 171 178 185 . Phase Plane of a Nonlinear Pendulum 23. Period o f a Nonlinear P e n d u l u m . Constant Coefficient First-Order Difference Equations 34. Phase Plane of a Linear Oscillator 22. Discrete One-Species Models with an Age Distribution 36. Conservation of Energy 20. What Happens if a Pendulum is Pushed Too Hard? 2 5 . 26. Equilibrium Positions and Linearized Stability 28. Nonlinear Frictionless Systems 18.viii Contents 16. Explicit Solution of the Logistic Equation 40. Introduction to Mathematical Models in Biology 31. Nonlinear Oscillations with Damping 27. Can a Pendulum Stop? 24. Density-Dependent Growth 38. A Discrete One-Species Model 33. Linear Constant Coefficient Difference Equations 42. Energy Curves 21. Further Readings in Mechanical Vibrations 53 * 54 56 61 67 70 76 82 84 87 * 91 100 104 114 Part 2: Populations Dynamics — Mathematical Ecology 30. Nonlinear Pendulum with Damping 29. Stochastic Birth Processes 37. Destabilizing Influence of Delays 43.

Traffic Flow and Traffic Density 59. Limitations of the Lotka-Volterra Equation 54. Further Reading in Mathematical Ecology 187 191 192 193 199 203 203 205 212 216 223 224 225 228 242 244 245 247 255 Part 3: Traffic Flow 56. A Velocity-Density Relationship 62. Method of Elimination B. Predator-Prey Models 49. Experimental Observations 63. Saddle Points C. Traffic Flow 257 259 260 265 273 275 282 286 289 . and Linearization 45. Nodes D. Flow Equals Density Times Velocity 60. Stability of Two-Species Equilibrium Populations 47. Conservation of the Number of Cars 61. Derivation of the Lotka-Volterra Equations 50. System of Two Constant Coefficient First-Order Differential Equations A. Phase Plane. Phase Plane of Linear Systems A.Contents ix 44. Summary 48. Introduction to Traffic Flow 57. Equilibrium. Systems Method (using Matrix Theory) 46. Automobile Velocities and a Velocity Field 58. Average Populations of Predators and Prey 52. Two Competing Species * 55. Man's Influence on Predator-Prey Ecosystems 53. General Remarks B. Qualitative Solution of the Lotka-Volterra Equations 51. Spirals E.

Steady-State Car-Following Models 65. A Highway Entrance 86. Linearization 67. The Earliest Shock 81. Exits and Entrances 84. Constantly Entering Cars 85. Effect of a Red Light or an Accident 83. Traffic Density Waves 69. A Nearly Uniform Traffic Flow Example 71. A Linear-Velocity Density Relationship 74. Discontinuous Traffic 78. An Interpretation of Traffic Waves 70. Congestion Ahead 77. Further Reading In Traffic Flow Index 293 298 301 303 309 314 315 319 323 331 339 344 345 347 354 360 363 370 372 385 387 389 394 395 . Uniform Traffic Stopped By a Red Light 79.x Contents 64. After a Traffic Light Turns Green 73. Partial Differential Equations 66. An Example 75. Nonuniform Traffic — The Method of Characteristics 72. A Stationary Shock Wave 80. A Linear Partial Differential Equation 68. Validity of Linearization 82. Wave Propagation of Automobile Brake Lights 76.

it is a special pleasure to welcome his classic Mathematical Models to a book series that was initiated with the related classic by Lin and Segel. O' Malley. Editor-in-Chief Classics in Applied Mathematics xi .Foreword Before courses in math modeling became de rigueur. Robert E. and traffic flow. My students and I can attest that this carefully crafted book is perfect for both self-study and classroom use. Since this is the very kind of applied math SIAM encourages. Richard Haberman had already demonstrated that mathematical techniques could be unusually effective in understanding elementary mechanical vibrations. population dynamics. as well as how such intriguing applications could motivate the further study of nonlinear ordinary and partial differential equations. Jr.

This page intentionally left blank .

Conservation of cars is introduced to derive the nonlinear partial differential equations for the traffic density. researchers debate the advantages of the continuum model of traffic presented here. In the future. or group projects. population dynamics. Here problems in mechanical vibrations. It is well written and hence suitable for selfstudy. The method of characteristics is carefully developed for these nonlinear partial differential equations and applied to various typical traffic situations. Discrete models for population growth are also presented. To use mathematics. Although it was written over 20 years ago. When developing strategies to deal with congested transportation systems today. this book is still relevant. and traffic flow are developed from first principles. and when teaching in recent years I have supplemented the book with a discussion of iterations of the logistic map and the period doubling route to chaos. one needs to understand the physical context. The phase plane is introduced to discuss nonlinear phenomena. it will be more common for highways to have sensors imbedded in the road and cameras focused above. In the independent presentations of mechanical vibrations and population dynamics. nonlinear ordinary differential equations are analyzed by investigating equilibria solutions and their linearized stability. as opposed to direct numerical simulation of large numbers of individually interacting cars. reading. The mathematical models and techniques developed continue to be of fundamental interest and hence provide excellent background and motivation to the reader for further studies.Preface to the Classics Edition Mathematics is a grand subject in the way it can be applied to various problems in science and engineering. but can be used for undergraduate courses in mathematical modeling or nonlinear dynamical systems or to supplement courses in ordinary or partial differential equations. It is intended as an introduction to applied mathematics. Richard Haberman xiii .

This page intentionally left blank .

not just the mathematical techniques. Mechanical vibrations (more specifically the motion of spring-mass systems and pendulums) is naturally followed by a study of other topics from physics. In addition. Often the mathematical model must be modified and the process of formulation. and interpretation continued. there is sufficient familiarity with traffic situations to enable the reader to thoroughly understand the traffic models that are developed. and the validity of the mathematical model is questioned. A previous exposure to physics will aid the reader in the part on mechanical vibrations. By studying three diverse areas in which mathematics has been applied this text attempts to introduce to the reader some of the fundamental concepts and techniques of applied mathematics. Thus we will be illustrating the relationships between each area and the appropriate mathematics. The material on population dynamics requires no background in biology. population dynamics. The topics discussed supplement rather than substitute for an introductory physics course. In this way a mathematical model is carefully formulated. In each area. Since one area at a time is investigated in depth in this way. xv . requiring at times the introduction of new mathematical methods. mathematical ecology (involving the population growth of species interacting with their environment) is a possible first topic in biomathematics. Similarly. but the text is readily accessible to those without this background. We will not find it necessary to refer to exceedingly technical research results. In each. Mechanical vibrations.Preface I believe the primary reason for studying mathematics lies in its applications. experimental motivation is self-contained. relevant observations and experiments are discussed. The solution is then interpreted. the experiments and common observations necessary to formulate and understand the mathematical models are relatively well known to the average reader. solution. The resulting mathematical problem is solved. Here we attempt only to introduce these various topics and leave the reader to pursue those of most interest. it is hoped that the reader will find these three topics as interesting as the author does. Furthermore these three topics were chosen for inclusion in this text because each serves as an introduction to more specialized investigations. the reader has the opportunity to understand each topic. such as heat transfer and fluid dynamics. and traffic flow (investigating the fluctuations of traffic density along a highway) introduces the reader in a simpler context to many mathematical and physical concepts common in various areas of engineering. and traffic flow are chosen areas to investigate in an introduction to applied mathematics for similar reasons.

and second-order constant coefficient equations. In this way the book has been designed to be substantially covered in one semester. On the other hand. Throughout.xvi This text has been written with the assumption that the reader has had the equivalent of the usual first two years of college mathematics (calculus and some elementary ordinary differential equations). the divergence theorem) is never used (nor is it needed). mathematical techniques are developed. a longer treatment of these subjects will be beneficial for some. these sections are not of equal length. Each major part is divided into many subsections. More specifically. the mathematical models of traffic flow involve first-order (nonlinear) partial differential equations. many) of the sections can be covered in an amount of time equal to that of a single lecture. I believe. while others are quite difficult Most are word problems. Usually more than one (and occasionally. This is best accomplished by attempting a significant number of the included exercises. The exercises have been designed such that their difficulty varies. considered by many to be one of the fundamental unifying themes of applied mathematics. the reader must take an active part. Few correspond to as much as a single lecture. as they are fully explained where needed. and hence are relatively independent of the previous material. A background in more advanced techniques is not necessary. Linear algebra and probability are also not required (although they are briefly utilized in a few sections which the reader may skip). The method of characteristics is slowly and carefully explained. Many critical aspects of these prerequisites are briefly reviewed. The concepts of equilibrium solutions and their stability are developed. However. resulting in the concept of traffic density wave propagation. enabling the reader to consider the relationships between the mathematics and the models. Many more problems are included than are reasonable for the average reader to do. depending on the background of the reader. but equal emphasis is placed on the mathematical formulation of the problem and the interpretation of the results. However. it is mostly restricted to first. emphasis is placed on the nonlinear aspects of ordinary differential equations. a knowledge of calculus including partial derivatives is required. the discussion of mechanical vibrations and population dynamics proceed in simikr ways. Phase plane methods are introduced and linearization procedures are explained in both parts. . in order to learn mathematics. but vector integral calculus (for example. Almost all readers will probably find some too easy. Although some knowledge of differential equations is required. Mathematically. In both.

A course has been offered for a few years based on preliminary versions of this text. with material added by individual instructors. Mark Ablowitz for his many thoughtful and useful suggestions. Terry Butler. I wish sincerely to thank Dr. For others. this text may be used as the basis of a full years's introduction to applied mathematics. For my own education. the applied mathematics group at the Massachusetts Institute of Technology must be sincerely thanked. Having no interest or knowledge in mathematics. Eugene Speer and Dr. Richard Falk who have co-taught the material with me. This text is a reflection of my own philosophy of applied mathematics. Annette Roselli whose accurate work was second only to her patience with the numerous revisions. for example. wave and Laplace's equation (and the mathematics of Fourier series as motivated by separation of variables of these partial differential equations. Besides the usual gratitude to one's wife. anyone's philosophy is strongly influenced by one's exposure. the heat. However. a second semester of applied mathematics could consist of. RICHARD HABERMAN . this was an exceptionally difficult effort. For the opportunity and encouragement to develop an applied mathematics course for which this text was written. Student comments have been most helpful as have been the insights given to me by Dr. especially Mrs. My appreciation to the typists of the manuscript (originally class notes). Furthermore his interest in the needs of students reinforced my own attitudes and resulted in this text. Any credit for much of this book must be shared with them in some ill-defined way. Also I would like to express my appreciation to Dr. my thanks to Liz for the thankless task of helping in rewriting the many drafts.Prefacexv Furthermore.

This page intentionally left blank .

Mechanical Vibrations .

This page intentionally left blank .

that are necessary to develop. Applied mathematics can be said to involve three steps: 1. The nonlinear frictionless pendulum and spring-mass systems are briefly studied. 3. However. A pendulum is then analyzed (Secs. we will attempt to give equal emphasis to all three aspects. the formulation of a problem—the approximations and assumptions. It may be impossible in an experiment to entirely eliminate certain undesirable effects. 10–13). One cannot underestimate the importance of good experiments in developing mathematical models. 3 . stressing the concepts of equilibrium and stability (Secs. simplify. the object being to see which effects account for given observations and which effects are immaterial. when theory and experiment quantitatively agree. As new experiments or observations are made. aside from an attempt to mimic nature. 14–16) since its properties are similar to those of a spring-mass system. the mathematical model is continually revised and improved. To illustrate this we first study some problems from physics involving mechanical vibrations. simplified by many approximations including linearization (Sees. However.Experimental observations necessitate the consideration of frictional forces (Sees. 2. This occurs because the real world consists of many interacting processes. In this manner mathematics becomes an integral part of the scientific method. Many different explanations can sometimes be given that agree qualitatively with experiments or observations.1. A spring-mass system is analyzed. then we can usually be more confident in the validity of the theory. mathematical models are important in their own right. A mathematical model has an advantage in that we are able to consider only certain effects. Furthermore one is never sure which effects may be negligible in nature. and understand the mathematical model. In this text. 2–9). Introduction to Mathematical Models in the Physical Sciences Science attempts to establish an understanding of all types of phenomena. the solving of realistic problems (including relevant computations). 17-18). the interpretation of the mathematical results in the context of the nonmathematical problem. The process of applying mathematics never ends. based on experiments or observations.

Newton's Law To begin our investigations of mathematical models. linear systems (frictionless). 26–28). Furthermore. Nonlinear systems which are damped are then discussed (Secs. 2. 19–20). moves back and forth (oscillates). Physical problems cannot be analyzed by mathematics alone. Newton discovered that the motion of a point mass is well described by the now famous formula . A spring-mass system cannot be solved without formulating an equation which describes its motion. this simple spring-mass system exhibits behavior of more complex systems. For example. we proceed in the following manner: 1. 3. 2. a problem with which most of you are somewhat familiar will be considered. once set in motion.4 Mechanical Vibrations before energy principles and phase plane analysis are used (Secs. Fortunately many experimental observations culminated in Newton's second law of motion describing how a particle reacts to a force. 4. This should be the first fundamental principle of an applied mathematician (although apparently some mathematicians would frequently wish it were not so). linear systems with friction. nonlinear systems with friction. 21-25). also aid in the understanding of the up-and-down motion of the ocean surface. in a sense. Observations of this kind of apparatus show that the mass. historically this problem played an important part in the development of physics. nonlinear systems (frictionless). Although few people today have any intrinsic interest in such a spring-mass system. We will discuss the motion of a mass attached to a spring as shown in Fig. 2-1: Figure 2-1 Spring-mass system. Mathematical models of increasing difficulty are formulated. the oscillations of a spring-mass system resemble the motions of clock-like mechanisms and. Examples of nonlinear frictionless oscillators are worked out in detail (Secs.

then where a is the vector acceleration of the mass Newton's second law of motion (often referred to as just Newton's law). Consider Fig. where v is the velocity of the mass and x its position: If the mass is constant (which we assume throughout this text). At least two assumptions are necessary for the validity of Newton's law.* For our purposes. the position of the center of mass (see exercise 2. for although mathematics is frequently treated as a science of exactness.1. this formula is valid only to the extent in which the finite size of a mass can be ignored. states that the force on a particle equals its mass times its acceleration. Thus. easily remembered as "F equals ma" The resulting acceleration of a point mass is proportional to the total force acting on the mass. A second approximation has its origins in work by twentieth century physicists in which Newton's law is shown to be invalid as the velocities involved approach the speed of light. . which shows two masses (m1 and m2) attached to the opposite ends of a rigid (and massless) bar: Figure 2-2.3.1). We emphasize the word approximation. *Newton's second law can be applied to finite sized rigid bodies if x. the position of the point mass. we will be satisfied with discussing only point masses. is replaced by xcm. equation 2. as long as the velocity of a mass is significantly less than the speed of light. EXERCISES 2. However. There are no point masses in nature. 2 Newton's Law where f is the vector sum of all forces applied to a point mass of mass m. 2-2. The forces f equal the rate of change of the momentum mv. Newton's law remains a good approximation.5 Sec. mathematics is applied to models which only approximate the real world.

if x is replaced by the center of mass xcmji. show m(dzxcm/dt2) =JF. 2. F = Gl + G2]. The motion of the center of mass of the rigid body is thus determined. m = m\ +J"2. Newton's third law of motion. then what is the total mass ml (b) Using the result of exercise 2. (b) Show that xcm lies at a point on the rigid bar connecting m± to m2. where is the center of mass xcm ? [Hint: Divide the bar up into N equal pieces and take the limit as N —» oo. (a) Suppose that an external force GI is applied to /HI. Figure 2-4 shows a rigid bar of length L: Figure 2-4. xcm ~ (m^x1 + m2x2)/(mi + w2).] (c) If the total force on the mass is F.2. Newton's Law as Applied to a Spring-Mass System We will attempt to apply Newton's law to a spring-mass system. its rotation remains unknown. and G2 to mi. However. (a) If the mass density p(x) (mass per unit length) depends on the position along the bar. Generalize the result of exercise 2. implies that PI = — Fi.e. The bar applies a force FI to mass ml and also a force FZ to w 2 as seen in Fig.3. stating that the forces of action and reaction are equal and opposite. and F is the sum of forces applied. The bar is free to move and rotate due to imposed forces..2. By applying Newton's second law to each mass. 2-3: Figure 2-3. where m is the total mass.1 to a rigid body consisting of N masses. call it the x direction. It is assumed that the mass moves only in one direction. 2. show the law can be applied to the rigid body consisting of both masses. show that 3.6 Mechanical Vibrations mi is located at 3ct and m 2 is located at x2. in which case the mass is governed by . xcm is the center of mass.

This place at which we center our coordinate axis. the mass could move only at a constant velocity. Furthermore. (This statement. is easily verifiable—see exercise 3.) Thus the observed variability of the velocity must be due to forces probably exerted by the spring. then the spring pushes the mass again towards the equilibrium position (F > 0). *Throughout this text. then the spring exerts a force pulling the mass back towards the equilibrium position (that is F < 0). Let us suppose a series of experiments were run in an attempt to measure the spring force. Thus we might obtain the results shown in Fig. we assume that the width of the mass is negligible. Thus. for example. is called the equilibrium or unstretched position of the spring*: Figure 3-1 spring. x = 0. . We have assumed that the force only depends on the amount of stretching of the spring.1. 3 Newton's Law as Applied to a Spring-Mass System If there were no forces F. where a curve is smoothly drawn connecting the experimental data points marked with an "*": Figure 3-2 Experimental spring force. known as Newton's first law. as we see in Fig. If we stretch the spring (that is let x > 0). the force exerted by the spring would increase. 3-2. Such a force is called a restoring force. the force does not depend on any other quantities. Equilibrium : no force exerted by the The distance x is then referred to as the displacement from equilibrium or the amount of stretching of the spring. if the spring is contracted (x < 0). 3-1. the force is assumed to be the same no matter what speed the mass is moving at. we would observe that as we increase the stretching of the spring. one should study the motions of spring-mass systems under different circumstances. At some position the mass could be placed and it would not move. there the spring exerts no force on the mass. Similarly. To develop an appropriate model of the spring force.7 Sec.

in a complex manner. However. 3-3 shows that this curve can be approximated by a straight line: Figure 3-3 Hooke's Law: approximation of experimental spring force. Doubling the displacement. Newton's second law of motion yields the simplest mathematical model of a spring-mass system. It depends on the elasticity of the spring. k is called the spring constant. for stretching of the spring which is not too large (corresponding to at most a moderate force). on the stretching. Using Hooke's law. prove Newton's first law. EXERCISES 3. .1. This linear relationship between the force and the position of the mass was discovered by the seventeenth century physicist Hooke and is thus known as Hooke's law. then using Newton's second law (m(d2x/dt2) = F). Fig.8 Mechanical Vibrations A careful examination of the experimental data shows that the force depends. doubles the force. Thus is a good approximation for the spring-force as long as the mass is not very far from its equilibrium position. (a) If the motion is only in the jc direction. Newton's first law states that with no external forces a mass will move along a straight line at constant velocity.

9

Sec. 4

Gravity

(b) If the motion is in three dimensions, then using Newton's second law (m(d2x/dt2) = F), prove Newton's first law. (Why is the motion in a straight line ?) 3.2. If a spring is permanently deformed due to a large force, then do our assumptions fail ?

4.

Gravity

In Sec. 3, we showed that the differential equation

describes the motion of a spring-mass system. Some of you may object to this model, since you may find it difficult to imagine a horizontally oscillating spring-mass system such as that shown in Fig. 4-1:

Figure 4-1.

It may seem more reasonable to consider a vertical spring-mass system as illustrated in Fig. 4-2:

Figure 4-2.

10

Mechanical Vibrations

The derivation of the equation governing a horizontal spring-mass system does not apply to the vertical system. There is another force—gravity. We approximate the gravitational force as a constant* — mg, the mass m times the acceleration due to gravity —g. The two forces add vectorially and hence Newton's law becomes

where y is the vertical coordinate, y — 0 is the position at which the spring exerts no force. Is there a position at which we could place the mass and it would not move, what we have called an equilibrium position ? If there is, then it follows that dy/dt = dzyjdt2 = 0, and the two forces must balance: Thus we see

is the equilibrium position of this spring-mass-gravity system (represented by Fig. 4-3), not y — 0:

Figure 4-3

Gravitational effect on spring-mass equilibrium.

*Actually, a gravitational force of attraction F exists directed between any two point masses mi and m2. Its magnitude is inversely proportional to the square of the distance between them, r, the so-called inverse-square law, where G is a universal constant determined experimentally. If the earth is spherically symmetric, then the force due to the earth's mass acting on any point mass is directed towards the center of the earth (or downwards). Thus the radial component of the gravitational force on a mass m is

where M is the mass of the earth. If the displacement of the spring is small as compared to the radius of the earth r 0 (not a very restrictive assumption!), then the gravitational force

//

Sec. 4 Gravity

Only at that position will the force due to gravity balance the upward force of the spring. The spring sags downwards a distance mg/k when the mass is added, a result that should not be surprising. For a larger mass, the spring sags more. The stiffer the spring (k larger), the smaller the sag of the spring (also quite reasonable). It is frequently advantageous to translate coordinate systems from one with an origin at y = 0 (the position of the unstretched spring) to one with an origin at y = —mg/k (the equilibrium position with the mass). Let Z equal the displacement from this equilibrium position:

Upon this substitution, equation 4.2 becomes

This is the same as equation 4.1. Thus the mass will move vertically around the new vertical equilibrium position in the same manner as the mass would move horizontally around its horizontal equilibrium position. For this reason we may continue to study the horizontal spring-mass system even though vertical systems are more commonplace.

EXERCISES

4.1. A mass m is thrown upward with initial speed v0. Assume that gravity is constant. How high does the mass go before it begins to fall ? Does this height depend in a reasonable way on m, v0, and g! A mass m is rolled off a table (at height h above the floor) with horizontal speed v0. Where does the mass land? What trajectory did the mass take? A mass m is thrown with initial speed v0 at an angle 8 with respect to the horizon. Where does the mass land? What trajectory did the mass take? For what angle does the mass land the farthest away from where it was thrown (assuming the same initial speed) ?

4.2. 4.3.

can be approximated by —GmM/r$, a constant. Thus the universal constant G is related tog by The rotation of the earth only causes very small modifications of this result. In addition, since the earth is not spherically symmetric, there are local variations to this formula. Furthermore, inhomogeneities in the earth's internal structure cause measurable variations, (which are useful in mineral and oil exploration).

12

Mechanical Vibrations

5.

Oscillation of a Spring-Mass System

We now proceed to analyze the differential equation describing a springmass system,

The restoring force is proportional to the stretching of the spring. Although this equation has been derived using many approximations and assumptions, it is hoped that the understanding of its solution will aid in more exact investigations (some of which we will pursue). Equation 5.1 is a second-order linear differential equation with constant coefficients. As you should recall from a course in differential equations, the general solution of this differential equation is where

and where Cj and c2 are arbitrary constants. However, for those readers who did not recognize that equation 5.2 is the general solution of equation 5.1, a brief review of the standard technique to solve constant coefficient linear differential equations is given. The general solution of a second-order linear homogeneous differential equation is a linear combination of two homogeneous solutions. For constant coefficient differential equations, the homogeneous solutions are usually in the form of simple exponentials, ert. The specific exponential(s) are obtained by directly substituting the assumed form ert into the differential equation. If en is substituted into equation 5.1, then a quadratic equation for r results, The two roots are imaginary, where CD = ^/kjm. Thus the general solution is a linear combination of e"" and e~tntt

13

Sec. 5 Oscillation of a Spring-Mass System

where a and b are arbitrary constants. However, the above solution involves the exponential function of an imaginary argument. The displacement x must be real. To show how equation 5.3 can be expressed in terms of real functions, we must recall that

as is derived in exercise 5.6 using the Taylor series of sines, cosines, and exponentials. A similar expression for e~tf0t, can be derived from equation 5.4a by replacing CD by -co. This results in

where the evenness of the cosine function [cos (—y) = cos y] and the oddness of the sine function [sin (—y) = —siny] has been used. Equations 5.4a and 5.4b are called Euler's formulas, which when applied to equation 5.3 yield The desired result follows, if the constants c^ and c2 are defined by

The constants cl and cz are arbitrary siice given any value of c1 and c2, there exists values of a and b, namely

Since the algebra is a bit involved, it is useful to memorize the result we have just derived: An arbitrary linear combination of eltat and e

ifat

,

is equivalent to an arbitrary linear combination of cos cot and sin cot,

14

Mechanical Vibrations

In the above manner you should now be able to state without any hesitation that the general solution of

IS

where co — */k/m and c1 and c2 are arbitrary constants. The general solution is a linear combination of two oscillatory functions, a cosine and a sine. An equivalent expression for the solution is

This is shown by noting in which case

If you are given cl and c 2 , it is seen that both A and 00 can be determined. Dividing the two equations yields an expression for tan 00, and using sin2 0o + c°s2 0o — 1 results in an equation for A2:

The expression, x = A sin (cot + 00), is especially convenient for sketching the displacement as a function of time. It shows that the sum of any multiple of cos cot plus any multiple of sin cot is itself a sinusoidal function as sketched in Fig. 5-1:

Figure 5-1

Period and amplitude of oscillation.

A is called the amplitude of the oscillation; it is easily computed from the above equation if c, and c2 are known. The phase of oscillation is cot + 00,

15

Sec. 5 Oscillation of a Spring-Mass System

00 being the phase at / = 0. In many situations, this agrees with the observed motion of a spring-mass system. This motion is referred to as simple harmonic motion. The mass oscillates sinusoidally around the equilibrium position x = 0. The solution is periodic in time. As illustrated in Fig. 5-1, the mass after reaching its maximum displacement (x largest), again returns to the same position T units of time later. The entire oscillation repeats itself every T units of time, called the period of oscillation. Mathematically a function /(/) is said to be periodic with period T if To determine the period T, we recall that the trigonometric functions are periodic with period 2n. Thus for a complete oscillation, as t increases to t + T, from equation 5.5 cot + 00 must change by 2n: Consequently the period T is

co, called the circular frequency (as is explained in exercise 5.7), is the number of periods in 2n units of time:

The number of oscillations in one unit of time is the frequency/,

measured in cycles per second (sometimes known as a Hertz). Since a springmass system normally oscillates with frequency l/2n+/kfm, this value is referred to as the natural frequency of a spring-mass system of mass m and spring constant k. Other physical systems have natural frequencies of oscillation. Perhaps in a subsequent course you will determine the natural frequencies of oscillation of a vibrating string or a vibrating drum head!

EXERCISES

5.1. Sketch x = 2 sin (3/ - »/2). 5.2. If x = —cos t + 3 sin t, what is the amplitude and phase of the oscillation? Sketch this function.

16

Mechanical Vibrations

5.3. 5.4. 5.5. 5.6.

If x — —cos t + 3 sin (/ — n/6), what is the amplitude of the oscillation? If x =. —sin 2t, what is the frequency, circular frequency, period, and amplitude of the oscillation? Sketch this function. It was shown that x = aeieot -f be~ieot is equivalent to x = Ci cos cot + C2 sin cot. Show that if c\ and 02 are real (that is if x is real), then b is the complex conjugate of a. The Taylor series for sin x, cos x, and e* are well known for real x:

5.7.

5.8.

The above Taylor expansions are also valid for complex x. (a) Show that eteot = cos cot + i sin cot, for co real. (b) Show that e~tcat = cos cot — i sin cot, for CO real. Consider a particle moving around a circle, with its position designated by the polar angle 0. Assume its angular velocity dBjdt is constant, d0/dt = co. Show that the x component of the particle's position executes simple harmonic motion (and also the y component), co is measured in radians per unit of time or revolutions per 2n units of time, the circular frequency. (a) Show that x = c\ cos cot + c^ sin cot is the general solution ofm(d2x/dt2) = — kx. What is the value of col (b) Show that an equivalent expression for the general solution is x = B cos (cot + 0o). How do B and 60 depend on c\ and c± 1

**6. Dimensions and Units
**

In the previous section, the formula for the circular frequency of a simple spring-mass system was derived,

As a check on our calculations we claim that the dimensions of both sides of this equation agree. Checking formulas by dimensional analysis is an important general procedure you should follow. Frequently this type of check will detect embarrassing algebraic errors. In dimensional analysis, brackets indicate the dimension of a quantity. For example, the notation [x] designates the dimension of x, which is a length

Thus. kilogram for mass. and seconds for time. However. The dimensions of the two sides must agree. inches. Meters (and the other metric units) are familiar to all of the world except the general public in the United States. This result can be used to determine the dimensions of an acceleration: Note that This is obvious from a physical point of view. Local folklore says that this unit was the length of a slightly inebriated student as he was rolled across the bridge by some "friends". However.e. This is shown in general directly from the definition of a derivative. The right-hand side is the difference between two values of y (hence having the dimension of y) divided by a small value of z (having the dimension of z). the velocity? Clearly. .. At the other extreme. as an aid in conversion to those familiar with the British-American system the equivalent length in feet or miles and the equivalent mass in pounds will appear afterwards in parentheses.* In any calculation to eliminate possible confusion only one unit of length should be used. or smoots. meters for lejjgth.e. miles. i. Mathematically we note that The dimensions of a derivative of any quantity is always the ratio of the dimension of that quantity divided by the dimension of the variable which we are differentiating with respect to. i. a smoot is a unit of length only used to measure the distance across the Charles River in Boston on the frequently walked Harvard Bridge. this result is shown below (since sometimes the dimension of a quantity might not be as obvious * Units are chosen to facilitate communication and understanding of the magnitudes of quantities. a length L divided by a time T. [x] = L. measured in units of feet. who seem reluctant to change. 6 Dimensions and Units L.. meters. In this text we will use metric units in the m-k-s system. What is the dimension of dx/dt.17 Sec.

(a) What is the dimension of col (b) Show that CO has the same dimensions as */k/m. the effect of varying the different parameters is investigated. then where M is a unit of mass. the gravitational acceleration is within 1 percent of this value.8 meters/sec2 (32 feet/sec2).? Since F — —kx. As discussed in Sec. What is the dimension of A. (a) Give an example of a possible dependence of y on k and m. Qualitative and Quantitative Behavior of a Spring-Mass System To understand the predictions of the mathematical model of a spring-mass system.2 Suppose a quantity y having dimensions of time is only a function of k and m. . You will probably need to note that a radian has no dimension.1). EXERCISES 6. the value we use is g — 9. In this way co can be shown to have the same dimension as */k/m (see exercise 6. (b) Can you describe the most general dependence that y can have on k and ml 7. Since g is an acceleration. 6. the units will be a length per unit of time squared.18 Mechanical Vibrations as it is in this problem): Thus.1. The formula (d/d0) sin 6 = cos 9 shows this to be true. g is only an approximation. 4. Everywhere at the surface of the earth. An important formula is the one derived for the period of oscillation.

with the same mass. Although each by itself is qualitatively and quantitatively understood. then the formula shows that the period increases. but it is doubtful that we could have known that quadrupling the weight results in an increase in the period by a factor of two! In mathematical models. then mathematical models are desirable in order to discover which mechanisms best account for the known data. Thus we suspect that the larger k is. what differences in the motion should occur? Let us compare two different springs represented in Fig.. On the other hand.e. When quantitative results are known (perhaps due to precise experiments). that is one whose spring constant k is larger. In complex problems sometimes two or more effects interact. Equation 7. their interaction may need mathematical analysis in order to be understood even qualitatively. The one which is firmer has a larger restoring force and hence it returns more quickly to its equilibrium position. The system oscillates more slowly (is this reasonable?).19 Sec.1 also predicts this qualitative feature. then we expect that our formula gives us the quantitative effects for the given problem—one of the major purposes for using mathematics. which quantities are important and which can be ignored. if the mass is increased using the same spring. In particular for a spring-mass system. the shorter the period. Without relying on the mathematical formula. 7 Qualitative and Quantitative Behavior of a Spring-Mass System Suppose that we use a firmer spring. If our intuition about a problem does not correspond to what a mathemat- . 7-1 : Figure 7-1. If the two agree. i. usually the qualitative effects are at least partially understood. we might have suspected without using any mathematics that increasing the mass increases the period. Quantitative results are often unknown. In any problem we should compare as much as possible our intuition about what should happen with what the formula predicts.

2. is zero. dx/dt. where c{ and c2 are arbitrary constants and <o — ^/kfrn. On the other hand.2 which satisfies the initial conditions that the mass is at x0 at t — 0.20 Mechanical Vibrations ical formula predicts.227 kilogram (£ pound) weight is observed to oscillate 12 times per second. in which case the mathematical formulation and solution has aided in directly improving one's qualitative understanding. What is the spring constant? If a . One way to initiate motion in a spring-mass system is to strike the mass. What is the natural frequency of oscillation of this spring-mass system ? 8. Initial Value Problem In the previous sections. it may occur that the intuition is correct and consequently that either there was a mathematical error in the derivation of the formula or the model upon which the analysis is based may need improvement. the velocity of the mass.5 centimeters (one inch). A weight (of unknown mass) is placed on a vertical spring (of unknown spring constant) compressing it by 2. what would be the resulting frequency ? 7. .1. A simpler method is to pull (or push) the mass to some position (say x0) and then let go. A . we have shown that is the general solution of the differential equation describing a spring-mass system.91 kilogram (two pound) weight was placed on the same spring. and at t — 0. The constants c\ and c2 will be determined from the initial conditions of the spring-mass system. then further investigations of the problem are necessary. Mathematically we wish to determine the solution of equation 8. EXERCISES 7. Perhaps the intuition is incorrect.

is often best for simple sketches. Thus In general the velocity is obtained by differentiating the formula for displacement : In this problem the mass is at rest at t = 0. To solve this initial value problem. as shown in Fig. Other initial value problems are posed in the exercises. 8-1. Figure 8-1 a Oscillation of a spring-mass system.1 satisfies the initial conditions. The mass executes simple harmonic motion. Setting the initial velocity equal to zero yields In this manner satisfies this initial value problem. at t = 0. This process continues indefinitely. It should be noted that the form of the general solution is more suited for the initial value problem than although the latter. For example. the arbitrary constants c.21 Sec. and c2 are determined so that equation 8. x = x0. 8 Initial Value Problem In the above example of an initial value problem the mass is initially at rest. The spring is initially stretched and hence the spring pulls the mass to the left. . as previously mentioned. Two initial conditions are necessary since the differential equation involves the second derivative in time.

Suppose that a mass is intially at x — x0 with an initial velocity v0. one corresponding to the mass initially at rest at x = x0 and the other corresponding to the mass initially at the equilibrium position with velocity v0.1. and m.2. Suppose that a mass is initially at the equilibrium position of a spring. EXERCISES 8. but is initially moving with velocity v0. k. 8. What is the amplitude of oscillation? Show that the amplitude depends in a reasonable way on v0. What is the amplitude of the total oscillation ? 22 .Figure 8-1 b Oscillation of a spring-mass system. Show that the resulting motion is the sum of two oscillations.

We analyzed the resulting oscillations which occur when the force can be approximated as being simply proportional to the stretching of the spring. Suppose instead of attaching a spring and a mass to a rigid wall. ml and ra2. we must formulate Newton's law of motion for each mass.23 Sec. Before discussing modifications to this model. To analyze that question. move. we must return to fundamental principles. The force on each mass equals its mass times its acceleration. Let us assume that the two masses are ml and m2. 9-2. 9. xl and x2 are the distances each mass is from a fixed origin): Figure 9-2. 9-1: Figure 9-1. we introduce the position of each mass (for example. as an example let us consider a more complicated spring-mass system.3. A Two-Mass Oscillator In the previous sections we carefully developed a mathematical model of a spring-mass system. 9 A Two-Mass Oscillator 8. We insist that the system is constrained to move only horizontally (as might occur if both masses slide along a table) as shown in Fig. while the connecting spring is known to have an unstretched length / and spring constant k. we attach a spring and a mass to another mass which is also free to move. in Fig. Consider a spring-mass system initially at rest with initial displacement XQ. . We wish to know the manner in which the two masses. Show that the maximum and minimum displacements occur halfway between times at which the mass passes its equilibrium position. In order to obtain expressions for the accelerations.

Thus Having derived the equations describing the motion of this system of two masses coupled by a spring. for example. Each force is an application of Hooke's law. Ft and F2. one which in fact yields a solution more easily interpreted. The only force on each mass is due to the spring. The stretching of the spring is the length of the spring x2 — x t minus the unstretched length /: x2 — x± — L The magnitude of the force is just the spring constant k times the stretching. In . The direct analysis of such a system is not difficult. although the magnitude of the force on mass mi is the same. An error may be made if we are not careful. We must now proceed to solve these equations. we prefer to approach the solution of this problem in a different way. then and To complete the derivation of the equations of motion. it is not necessary that 2 — xi — l> f°r i*1 many circumstances the spring may be stretched or compressed. for example. k(x2 — x^ — /).1 is a linear equation for xlt it contains x2 which is at present unknown! A similar predicament appears in equation 9. By inspection we note that the equations simplify if they are added together.2. we may impose initial conditions such that initially x2 — *i ^ /. we have only begun the process of applying mathematics to this physical situation.24 Mechanical Vibrations X Although the unstretched length of the spring is /. Now from Newton's law of motion. At first glance. we have a difficult task in front of us for although equation 9. Certainly.2 form a linear system of two coupled second order ordinary differential equations involving the two unknowns xt and x2.1 and 9. we must determine the two forces. the spring is stretched (x2 — xt — I > 0). In any situation the force on mass ml is k(x2 — Xi — /) and thus However. then the mass m1 is being pulled to the right. but the direction of the force is quite important. the spring force acts in the opposite direction. However. Equations 9. The force is in the positive xl direction. If. it follows that if jpj is the force on mass ml and if F2 is the force on mass m2. the force is proportional to the stretching of the spring (it is not proportional to the length of the spring).

1 by ml and equation 9. you will soon discover that although adding the two equations results in a significant simplification. Through this subtle trick* we discover that Letting z be the stretching of the spring we see that By this rather lucky manipulation. we see the remarkable result that the stretching of the spring executes simple harmonic motion. we have shown that we mean that the center of mass of the system of two masses moves at a constant velocity. the solution of the mathematical equation of interest simplifies if we use insight gained from the physical problem. does not accelerate. equation 9. We see this can be accomplished by dividing each equation by the mass (i. Instead we note that the force only depends on the stretching of the spring. but hardly aids in understanding the possibly complex behavior of each individual mass. Perhaps we can directly determine a differential equation for the stretching of the spring..1. but moves at a constant velocity (determined from initial conditions). By saying there is no acceleration. Instead as often occurs in applied mathematics. -f /w2). subtracting the two is not helpful.2 from equation 9. x2 — *i — /.25 Sec. The simple expression for the motion of the center of mass is quite interesting. The circular frequency is ^/k(\jml + l/m2). (/M.XI + /w2x2)/(w. . 9 A Two-Mass Oscillator that manner The sum of the two forces vanishes (since one force is minus the other). Try subtracting equation 9.3 can be re-expressed as Thus the center of mass of the system. the system obeys Newton's first law and will not accelerate. (as discussed in exercise 2.la). If this system of two masses attached to a spring is viewed as a single entity.2 by /w2) and then subtracting rather than by subtracting right away. then since there are no external forces on it. Equation 9.e. *You should note that we did not use any mathematical motivation for these steps.

the center of mass moves at a constant velocity! We could discuss more complex spring-mass systems.1. we will return in the next sections to the study of a single mass attached to a spring. one spring with spring constant ki (and unstretched length /i) and the other with spring constant £2 (and unstretched length /2) as shown in Fig. Suppose that the initial positions and initial velocities were given for a springmass system of the type discussed in this section: Determine the position of each mass at future times. In the following exercises consider two springs. The mass m necessary for this analogy is such that This mass m is less than either Wj or m2 (since 1/m > \jml and 1/m > l/ra2).26 Mechanical Vibrations We also note that this is the same type of oscillation that would result if a certain mass m were placed on the same spring.3. 9. that is one with spring constant k but fixed at the other end. and 9.4..6. it is thus called the reduced mass: Attaching a spring-mass system to a movable mass reduces the effective mass.2.e. each obeying Hooke's law. Instead. suppose the system is aligned vertically (rather than horizontally) and that consequently a constant gravitational force is present. 9.-9.] 9. the stretching of the spring executes simple harmonic motion as though a smaller mass was attached. Consider two masses (of mass m\ and m 2 ) attached to a spring (of unstretched length / and spring constant k) in a manner similar to that discussed in this section. [Hint: See equations 9. But don't forget that the entire system may move. i. However. EXERCISES 9. 9-3: .6.5. Analyze this system and compare the results to those of this section.

9-3 and 9-5): Figure 9-5. 9.4. . (c) Show that the mass executes simple harmonic motion about its equilibrium position. (b) What position of the mass would be called the equilibrium position of the mass ? If both springs are identical. where should the equilibrium position be ? Show that your formula is in agreement.3.5. 9-3 and 9-6): Figure 9-6. (a) What position of the mass would be called the equilibrium position of the mass? (b) Show that the mass executes simple harmonic motion about its equilibrium position. (d) What is the period of oscillation ? (e) How does the period of oscillation depend on dl Suppose that a mass m were attached to two springs in parallel (refer to Figs.27 Sec. Suppose that a mass m were attached between two walls a distance d apart (refer to Figures 9-3 and 9-4): Figure 9-4. 9 A Two-Mass Oscillator Figure 9-3. (a) Briefly explain why it is not necessary for d = l\ + 72. (c) What is the period of oscillation ? (d) If the two springs were to be replaced by one spring. 9. what would be the unstretched length and spring constant of the new spring such that the motion would be equivalent ? Suppose a mass m were attached to two springs in series (refer to Figs. 9.

] Consider two masses each of mass m attached between two walls a distance d apart (refer to Figs. [Hint: Apply Newton's law also to the massless point at which the two springs are connected. What position of each mass would be called the equilibrium position of the system of masses ? (b) Show that the distance between the masses oscillates. describe the motion that could take place both qualitatively and quantitatively.4a-d. as illustrated in Fig. (a) Suppose that the left mass is a distance x from the left wall and the right mass a distance y from the right wall. (e) If x = y. 9-8.6.7. Figure 9-8. 9-3 and 9-7): Figure 9-7. Answer the same questions as in exercise 9. by cutting the string connecting /MI and mz). 9. (a) Suppose that the system is in equilibrium when x = L. Consider a mass mt attached to a spring (of unstretched length d) and pulled by a constant force F2. describe the motion that could take place both qualitatively and quantitatively. then what is the period and amplitude of oscillation of mi ? . (d) If the distance between the two masses remained constant. Fz = m2g. Is L > d or is L < dl If L and rfare known. what is the spring constant kl (b) If the system is at rest in the position x = L and the mass mi is suddenly removed (for example. Assume that all three springs have the same spring constant and unstretched length.28 Mechanical Vibrations 9. What is the period of that oscillation ? (c) Show that x-y executes simple harmonic motion with a period of oscillation different from (b).

. as shown in Fig. 10-2. The amplitude of oscillation remains constant. Friction Our mathematical model shows that the displacement of a simple spring-mass system continues to oscillate for all time. Figure 10-1. In this case the mathematical model of a spring-mass system. We suspect that the mass oscillates around its equilibrium position with smaller and smaller magnitude until it stops.29 Sec. for we can at least imagine a spring-mass system that exhibits many oscillations before it finally appears to significantly decay. 10-1. the mass never stops completely nor does the amplitude even decay! Does this correspond to our experience? If we displaced the mass to the right. then we would probably expect the mass to oscillate in the manner sketched in Fig. 10 Friction 10. Must we reject a mathematical model that yields perfect periodic motion ? Absolutely not. Figure 10-9.

For example. we must include other forces. Figure 10-4. a damped oscillation meaning the same as an oscillation which decays. representing any restoring force. This force-velocity relationship is called a linear damping force. When the velocity is negative. we can imagine this kind of force resulting from the "friction" between the mass and the surrounding air media. How can we improve our model to account for the experimental observation that the amplitude of the mass decays ? Perhaps when the restoring force was approximated by Hooke's law. When the spring is moving to the right. resisting the motion of the mass. In order to account for the observed decay. Ff > 0. Figure 10-3. 10-3. The simplest mathematical way this can be accomplished is to assume that the frictional force is linearly dependent on the velocity: where c is a positive constant (c > 0) referred to as the friction coefficient. in later sections we will show that this is not the case as the equation m(d2x/dt2) = —f(x).30 Mechanical Vibrations is a good approximation for times that are not particularly long. a force preventing motion. then this frictional force Ff must be negative. The accuracy of this assumed form of the force-velocity relationship should be verified . However. the importance of simple harmonic motion is in its aid in understanding more complicated periodic motion. dx/dt > 0. Ff < 0. then there is a force exerted to the right. dxfdt < 0. the frictional force must be positive. When the velocity is positive. that is. Figure 10-4 shows that when the spring is moving to the left. never has oscillatory solutions that decay in time. What causes the amplitude of the oscillation to diminish ? Let us conjecture that there is a resistive force. as shown in Fig. Furthermore. the possibility of decay was eliminated. then there is a force exerted to the left.

y depends on the roughness of the surface (and the weight of the mass). 10-5. Ff ^ —c(dxfdt). hence.31 Sec. illustrated in Fig. 10-6 (see exercise 10. Instead. Here we claim that in some situations (but certainly not all) this is a good approximate expression for the force resulting from the resistance between an object and its surrounding fluid (liquid or gas) media (especially if the velocities are not too large). Independent forces add vectorially. does not act in the way previously described. experiments indicate that once the mass is moving the friction force is resistive but has a magnitude which is approximately constant independent of the velocity. . 10 Friction experimentally.4). We model this experimental result by stating as sketched in Fig. In later sections exercises will discuss the mathematical solution of problems involving this type of friction. called Figure 10-6 Coulomb friction. the differential equation describing a spring-mass system with a linear damping force and a linear restoring force is or equivalently The force corresponding to the friction between a spring-mass system and a table.

10.3.. moving in a straight line. Reading. . However. then Ff could be any value such that I Ff | <. meteorites. In some problems both linear damping and Coulomb friction occur. Differential Equations.7. known as Newtonian damping. Suppose that an experimentally observed frictional force is approximated by Ff = 01(4x1 dt)3. the particle will come to rest in a finite time. In certain physical situations the damping force is proportional to the velocity squared. and when will it stop ? 10. 10. What is the dimension of the constant c defined for linear damping? 10. Mass. Theory and Use in Time and Motion.5. An excellent brief discussion is given by Dickinson.9. with /? > 0. sketch the total frictional force as a function of the velocity. no restoring force) approaches a terminal velocity. In this case. 10. (a) What is the sign of y ? (b) What is the dimension of y ? (c) Assume that if dxjdt = 0. (a) What is the sign of a ? (b) What is the dimension of a ? (c) Show that the resulting differential equation is nonlinear. How far will the particle travel. What values of x are then equilibrium positions of the springmass system ? 10. 10.2. then at what time does the mass of a spring-mass system first stop moving to the right ? Will the mass continue to move after that time? 10. y.] A particle not connected to a spring. If initially x = x0 > 0 and the velocity is «0 > 0.e.32 Mechanical Vibrations Coulomb friction.6.8.4. In this case show that where a > 0. show that x = 0 is the only equilibrium position of the mass. equation 10. Consider Coulomb friction. If gravity is approximated by a constant and if the frictional force is proportional to the velocity.4.: Addison-Wesley. From the differential equation for a spring-mass system with linear damping. where a is a constant. [You may wish to think about this effect for raindrops. we will limit our discussion to linear damping. in this text for the most part. or parachutes.3. show that a free-falling body (i.1. 1972. EXERCISES 10. is subject to a retardation force of magnitude fi(dx/dt)n. (a) Show that if 0 < n < 1. Assume the same form of friction as in exercise 10.

11 Oscillations of a Damped System (b) What happens if n = 1 ? (c) What happens if 1 < n < 2? (d) What happens if n > 2 ? 10. (c) Compare the results of parts (a) and (b). If this is not true then perhaps a spring-mass system decays as a result offerees other than a linear damping force. 11. equation 10. which shows a glass of mass m starting at x = 0 and sliding on a table of length H: Figure 10-7. where r satisfies the characteris- . 10-7.33 Sec. Answer the same question as in part (a).10. Oscillations of a Damped System A spring-mass system with no forces other than a spring force and friction is governed by It must be verified that solutions to this equation behave in a manner consistent with our observations. In order to solve a constant coefficient homogeneous ordinary differential equation recall that the two linearly independent solutions "almost always" can be written in the form of exponentials ert.3 ? (b) Suppose that the frictional force instead is Describe physically what Ff represents. Consider Fig. (a) For what initial velocities v0 will the glass fall off the table on the right if the only force is Coulomb friction.

34 Mechanical Vibrations tic equation obtained by direct substitution. Underdamped Oscillations If c2 < 4mk. we see . EXERCISES 11. We call this the underdamped case. the roots of the characteristic equation are complex conjugates of each other (see equation 11. this shows that c2 and mk must have the same dimensions (as you can easily verify). Show that c2 has the same dimension as mk. In this case. the damping force is not particularly large. The three cases c2 ig 4mk must be distinguished.1) is By factoring e~c</2m.1. real and equal. then the coefficient of friction is small.2. A different form of the general solution corresponds to each case. 11. and complex. since the roots are respectively real and unequal. What dimension should the roots of the characteristic equation have? Verify that the roots have this dimension. The two roots of this equation are From a dimensional point of view.3) where Thus the general solution of the differential equation (11. 12.

while at the minimum value of the sinusoidal function. The exponential Ae~ct/2m is called the amplitude of the oscillation. It varies smoothly throughout. Thus we first sketch the exponential Ae~ct/2m and its negative — Ae~ct/2m in dashed lines. 12 Underdamped Oscillations Recalling that an arbitrary linear combination of e{0" and e~'°" is equivalent to an arbitrary linear combination of cos cot and sin cot. x equals minus the exponential. 12-2 (exactly where depends on the phase 0 0 ) : Figure 12-2 Underdamped oscillation. At the maximum value of the sinusoidal function. x equals the exponential alone. the maxima and minima are not halfway between the appropriate zeroes. Indeed this seems quite similar to what we expect to observe in the case of a *Although the zeroes are easy to locate.35 Sec. Periodically at the "jt's" the solution lies on the two exponential curves drawn in Fig. Figure 12-1. The solution is the product of an exponential and a sinusoidal function (each sketched in Fig. Thus in the figure we sketch in a solid line the motion of a spring-mass system in the slightly damped case*. 12-1). the function is zero. . Thus the amplitude exponentially decays. we observe that the motion of a linearly damped spring-mass system in the underdamped case (c2 < 4mk) is described by or A sketch of the solution is most easily accomplished using the latter form. Halfway between the marks.

.36 Mechanical Vibrations spring-mass system with sufficiently small friction. The period of oscillation follows from equations 12. c = 0. Figure 12-3 Oscillation with negligible damping. Thus we wish to determine the amplitude of oscillation after one period. it cannot be ignored as the amplitude of the oscillation will diminish in time only with friction. In this case though the solution is not exactly periodic. We will show that for some time the solution behaves as though there were no friction if (read much less than The friction of a spring-mass system is said to be negligible if during many oscillations the amplitude remains approximately the same as is represented in Fig.2 and is However if c2 <C 4mk. 12-3. we can approximate the period by its frictionless value: (read "approximately equals"). « = ^kjm. we can speak of an approximate circular frequency Notice this expression for the frequency reduces to the frictionless value if there is no friction. It is equivalent to say the friction is negligible if after one "period" the amplitude of the oscillation has remained approximately constant. However. As long as there is friction (c > 0). no matter how small.1 and 12. the mass never absolutely stops.

However..] *Improvemer*«5 to this approximation of the period are suggested in the exercises.37 Sec. damping is negligible (with a 95 percent criteria) if This calculation has been simplified using the rough nmeriecal aproxiamtion tionsince +9. Since e~* approximately equals 1 if x is small.. the natural logarithm near 1 may be accurately approximated using the Taylor series formula Using this formula. If friction is sufficiently small (c2 < 4mA:).* In this manner we can roughly estimate the amplitude of oscillation after one period. . EXERCISES 12. then the exponential has not decayed much in one "period." Using a numerical criteria. yields or The natural logarithm of .1.95 can be obtained from a mathematical table. 12 Underdamped Oscillations This is the first term in a Taylor series expansion of the period. it has been shown that where (b) Determine the first few terms of the Taylor series for small z of (1 — z)1/2. [Hint: Use the binomial expansion (see page 89). the damping might be said to be negligible if after one "period" the mass returns to at least 95 percent of its original position.8696. it follows that if c2 <C 4mk. that is if Taking the natural logarithm of both sides.

when the forcing function is a maximum. Consider the effect of striking the mass of a spring-mass system that is 12..3.. B cos (O0t. then determine the general solution of the differential equation.6. . However. the stretching of the spring is a maximum. then roughly sketch the solution. Show that the local maximum or minimum for the displacement of an underdamped oscillation does not occur halfway between the times at which the mass passes its equilibrium position. 12.e. if the forcing frequency is greater than the natural frequency (ct)0 > \/k/m). when the forcing function is a maximum.38 Mechanical Vibrations (c) Using the result of (b).2. Show that the ratio of two consecutive local maximum displacements is a constant. What is that amplitude? Note it is independent of initial conditions. c = 0). then the mass oscillates "180 degrees out of phase" with the forcing function (i. Show that the solution is oscillatory if CO0 =£ ^kjm. show that x —» oo as t —> oo.7. What is the amplitude of the oscillation ? 12. (b) If c2 < 4mk. then the mass oscillates "in phase" with the forcing function (i. 12. then at what frequency is the response largest ? 12.e. show that the time period between successive local maxima (or minima) is constant. Suppose that there is an additional force. If c < 0. Show that the solution algebraically grows in time if CO0 = */k/m (this is called resonance and occurs if the forcing frequency CO0 is the same as the natural frequency \/k/m). (a) If the coefficient of friction is zero (i. (c) If c2 <C 4mk.5. the force is called a negative friction force.4. Assume that B and (O0 are known.e. Consider a spring-mass system with linear friction (but without gravity). Determine the motion of an underdamped spring-mass system which is initially at its equilibrium position (x — 0) with velocity v0. (b) If in addition c2 < 4mk. Thus for sufficiently large times the motion is accurately approximated by an oscillation of constant amplitude. show that for large times the oscillation approximately obeys the following statements: If the forcing frequency is less than the natural frequency (C00 < */k/rri). What is that period ? 12.8. show that the general solution consists of the sum of oscillatory terms (of frequency 000) and terms which exponentially decay in time. improve the approximation At what time has the amplitude of oscillation of a spring-mass system with negligible friction decayed to \\e of its original value? Does this time depend in a reasonable way on k. If c2 <C 4mA:. m. On the other hand. and vice versa). the compression of the spring is a maximum and vice versa). (a) In this case. a periodic forcing function. and c ? 12. (d) In part (b) the ratio of the amplitude of oscillation of the mass to the amplitude of the forcing function is called the response..

] (c) If (as before) At —> 0 and /0 —> oo such that /0 At —> I. Calculate the limit of part (b) as At —> 0.8 for an alternate derivation. as A/1 —> 0.. and show that as At —* 0. and then zero thereafter: (a) Show that if c2 < 4mk. then for 0 < t < A/. This explains a method by which a nonzero initial velocity occurs. show that the new initial conditions after the impulse are x(G) = x0 and (dx/dtm = Urn. (b) Calculate the position and velocity of the mass at / = At. (b) If At is small. show that [Hint: Use Taylor expansions.9. fQAt = / (called the impulse). Reconsider exercise 12.39 Sec. .e. (d) Briefly explain the following conclusion: the effect of an impulsive force is only to instantaneously increase the velocity by Ijm. Assume that /(/) is approximately constant for some short length of time At. i. 12 Underdamped Oscillations initially at rest. Further assume that as A? —> 0. 12. (a) Show that for 0 < t < Ar. The problem is to solve subject to the initial conditions where /(/) is the force due to the striking. (c) Assume that the force is large for the short length of time A/. /0 —»-oo.

we should expect that the mass decays to its equilibrium position quite quickly.40 Mechanical Vibrations 13. . Instead. the mass either decays to its equilibrium position as seen in Fig. Overdamped and Critically Damped Oscillations On the other hand. Exercise 13.1 is where r t and r 2 are real and both negative. The motion of the mass is no longer a decaying oscillation. 13-1 (a) or (b). The solution of equation 11.1 shows that it does not oscillate. 13-l(c): Figure 13-1 Overdamped oscillations. or it shoots past the equilibrium position exactly once before returning monotonically towards the equilibrium position as seen in Fig. if the friction is sufficiently large. then and we call the system overdamped. If the friction is sufficiently large.

What is the approximate maximum amplitude ? 13. m. then show that the mass crosses its equilibrium position only if the initial velocity is sufficiently negative. Assume that friction is extremely large (c2 > 4mA:). In particular. A sketch would indicate no qualitative difference between this case and the overdamped case.3. from a physical point of view this case is insignificant.41 Sec.2.3b. (a) Show that the mass either decays to its equilibrium position (without passing through it). m. sketch the approximate solution.1 for the critically damped case c2 = 4mA:. (b) Estimate the characteristic roots if c2 > 4mk.4.] (b) Solve the differential equation governing the spring-mass system (with friction) if the mass term can be neglected. (c) Based on part (b). then roughly sketch the solution you expect by physical reasoning. However. then approximate the solution. EXERCISES 13. What is the value of this critical velocity ? Does it depend in a reasonable way on x0. The mass crosses its equilibrium position only if the initial velocity is sufficiently negative (assuming that the initial position is positive). (b) If the initial position x0 of the mass is positive. Assume that friction is extremely large (c2 >> 4mA:). When c2 = 4mk. or that the mass shoots past its equilibrium position exactly once before returning monotonically towards its equilibrium position. c. Mathematically c2 = 4mk requires a separate discussion since both exponential solutions become the same. 13 Overdamped and Critically Damped Oscillations It cannot cross the equilibrium position more than once. Show that the two initial . (a) If the mass is initially at x = 0 with a positive velocity. the solution may go through the equilibrium position once at most. 13. and kl 13. [Hint: Use the result of exercise 13. Do exercise 13. Any small deviation from equality will result in either of the previous two cases. For mathematical completeness.1. Assume that friction is sufficiently large (c2 > 4mA:). the solution in this case is Although there is an algebraic growth t. the spring-mass system is said to be critically damped. This is because the quantities c. and k are all experimentally measured quantities—there is no possibility that these measurements could be such that c2 = 4mk exactly. (a) If the mass is initially at x = XQ with velocity v0. the solution still returns to its equilibrium position as t —* oo since the exponential decay is much stronger than an algebraic growth.

Consider a vertical spring-mass system with linear friction. x = XQ and dx/dt = v0. (d) Compare your result to exercise 13. 13. We may now be wondering what effect the neglected nonlinear terms may have.] (a) Using a computer. To give us additional motivation to analyze nonlinear problems. Show that this solution does not approximate very well the solution to (a) for all time. shown on the next page in Fig. that is. (c) Solve the differential equation governing the spring-mass system (with friction) if the restoring force term can be neglected. 13.42 Mechanical Vibrations 13. Show that the difference between the exact solution and the particular solution tends towards zero as t —> oo (if c > 0) independent of the initial conditions. [Hint: Let ri —> rz. Why not ? Applying which one of the two initial conditions yields a result consistent with (a) ? Show that the solutions obtained in (a) and (b) are quite similar except for small times. Show that the two initial conditions can be satisfied. satisfying the initial conditions x(G) = 1 and (dx/dt)(0) = VQ for various negative initial velocities.8.5. determine the motion of an overdamped spring-mass system (let m = l.c = 3). 14-1. at / = 0.7. (c) Estimate this critical velocity. conditions cannot be satisfied. but with a spring of greater length with the same spring constant. Consider a pendulum of length L. (b) Show that the mass crosses its equilibrium position only if the initial velocity is sufficiently negative. Assume and also assume a particular solution is known. Show that the limiting solution is the same as for the case c2 = 4mk. We know from observations that a pendulum oscillates in a manner at least qualitatively . At one end the pendulum is attached to a fixed point and is free to rotate about it. Suppose two additional forces are present. Show that the motion is the same as that which would occur without gravity.1b. 14. gravity and any other force/}(r) only depending on time. we now discuss a common physical system whose mathematical formulation results in a specific nonlinear equation. Assume that c2 > 4mk. A mass m is attached at the other end as illustrated below.fc = l. 13.6. (b) Take the limit of the solution obtained in (a) as c2 —> 4mk. A Pendulum We have investigated in some depth a spring-mass system which is governed by a linear differential equation. (a) Solve the initial value problem.

43

Sec. 14 A Pendulum

Figure 14-1

A pendulum.

similar to a spring-mass system. To make the problem easier, we assume the mass m is large enough so that, as an approximation, we state that all the mass is contained at the bob of the pendulum (that is the mass of the rigid shaft of the pendulum is assumed negligible). Again we apply Newton's second law of motion, F = ma. The pendulum moves in two dimensions (unlike the spring-mass system which was constrained to move in one dimension). However, a pendulum also involves only one degree of freedom as it is constrained to move along the circumference of a circle of radius L. This is represented in Fig. 14-2.

Figure 14-2.

Consequently, we will now develop the form Newton's law takes in polar coordinates. In two or three dimensions, Newton's law for a mass m is

where x is the position vector of the mass (the vector from the origin to the mass). (In 3-dimensional rectangular coordinates, x = xi + yj -f zk, the acceleration is given by

44

Mechanical Vibrations

since /, j, k are unit vectors which not only have fixed magnitude but also have fixed directions.) In polar coordinates (centered at the fixed vertex of the pendulum), the position vector is pointed outward with length L,

where f is the radial unit vector. The polar angle 6 is introduced such that 0 = 0 corresponds to the pendulum in its "natural" position.* (See Fig. 14-3.)

Figure 14-3.

L is constant since the pendulum does not vary in length. Thus

However, although the magnitude of t is constant (\r\ = 1), its direction varies in space. To determine the change in the radial unit vector r, we express it in terms of the Cartesian unit vectors. From Fig. 14-5

Note the minus sign. You can easily verify that f is a unit vector. The 0-unit vector is perpendicular to f and of unit length (and is in the direction of

*Be careful—this definition of the polar angle differs from the standard one shown in Fig. 14-4.

Figure 14-4.

45

Sec. 14

A Pendulum

Figure 14-5

Radial and angular unit vectors.

increasing 0). Thus also from Fig. 14-5:

^ In order to calculate the acceleration vector d2x/dt2, the velocity vector dx/dt must first be calculated:

Since L is a constant for a pendulum, dLjdt = 0, and hence

From equation 14.3a,

which we note in general is more simply written as

Similarly,

Thus the velocity vector is in the direction of 0,

The magnitude of the velocity is L(dOjdi), if motion lies along the circumference of a circle. Why is it obvious that if L is constant, then the velocity is in the 0 direction ? [Answer: If L is constant, then in a short length of time the position vector has changed only a little, but the change must be in the 0 direction (see Fig. 14-6). In fact we see geometrically that

46

Mechanical Vibrations

Figure 14-6

Rate of change of position vector.

The 6 component of the velocity is the distance L times the angular velocity dO/dt. The acceleration vector is obtained as the derivative of the velocity vector:

The angular component of the acceleration is L(d20/dt2). It exists only if the angle is accelerating. If L is constant, the radial component of the acceleration, —L(dO/dt)2, is always directed inwards. It is called the centripetal acceleration and will occur even if the angle is only steadily increasing (i.e., even if the angular velocity dO/dt is constant). For any forces F, when a mass is constrained to move in a circle, Newton's law implies

For a pendulum, what are the forces ? Clearly, there is a gravitational force —mg], which should be expressed in terms of polar coordinates. From the definitions of f and 0, equation 14.3,

Thus the gravitional force, mg = —mgj = mg cos Of — mg sin 0 9. Perhaps this is more readily seen by breaking the direction —/ into its polar components, as is done in Fig. 14-7. Are there any other forces on the mass ? If there were no other forces, then the mass would not move along the circle. The mass is held by the rigid shaft of the pendulum, which exerts a force — Tf towards the origin of as yet

47

Sec. 14 A Pendulum

Figure 14-7.

Figure 14-8

Forces on pendulum.

unknown magnitude T (and, as will be shown, of nonconstant magnitude). The forces on the bob of the pendulum are illustrated in Fig. 14-8. This results in motion along the circle. Thus

Each component of this vector force equation yields an ordinary differential

equation

(A two-dimensional vector equation is equivalent to two scalar equations.) T could be obtained from the second equation (if desired, which it frequently isn't), after determining 0 from the first equation. Equation 14.7a implies that the mass times the 8 component of the acceleration must balance the 0 component of the gravitational force. The mass m can be cancelled from both sides of equation 14.7a. Thus, the motion of the pendulum does not depend on the magnitude of the mass m attached to the pendulum. Only varying the length L (or g) will affect the motion. This qualitative fact has been determined even though we have not as yet solved the differential equation. Furthermore, only the ratio g(L is important as from equation 14.7a

(This is an advantageous procedure whenever possible in applied mathematics

48

Mechanical Vibrations

—the determination of the important parameters. In fact for a spring-mass system, m(d2x/dt2) = — kx, it is not the two parameters k and m that are important, but only their ratio kjm, since d'1xldt* = —(kjm)x. Again a conclusion is reached without solving the equation.) The pendulum is governed by a nonlinear differential equation, equation 14.8, and hence is called a nonlinear pendulum. The restoring force — mg sin 6 does not depend linearly on 0, the unknown! Nonlinear problems are usually considerably more difficult to solve than linear ones. Before pursuing this nonlinear problem (and others), we recall from calculus that for 0 small, Geometrically the functions sin 0 and 0 are nearly identical for small 0 (0 being the linearization of sin 0 around the origin as seen in Fig. 14-9):

Figure 14-9

For 6 small, 6 approximates sin 6.

Using that approximation, the differential equation becomes

called the equation of a linearized pendulum. This is the same type of differential equation as the one governing a linearized spring-mass system without friction. Hence a linearized pendulum also executes simple harmonic motion. The pendulum oscillates with circular frequency

49

Sec. 14 A Pendulum

and period T = In.^L/g, as long as 0 is small. This result can be checked as to its dimensional consistency. The effect of changing the length of the pendulum (or changing the magnitude of gravity) can be qualitatively and quantitatively determined immediately. Again the period of oscillation is independent of amplitude (as an approximation for small amplitude oscillations). Apparently this was first realized by Galileo, who observed the swinging of lamps suspended from long cords (i.e., a pendulum) in churches. You must remember that these observations were made before accurate clocks existed, and thus Galileo used his pulse to measure time! In later sections we will investigate what happens if 0 is not small.

EXERCISES

14.1. Consider the differential equation of a pendulum. Show that the period of small amplitude oscillations (around its natural position) is T = 2n\/L/g. Briefly discuss the dependence of the period on L, g, and m. 14.2. Consider a mass m located at x = xi + yj, where x and y are unknown functions of time. The mass is free to move in the x-y plane without gravity (i.e., it is not connected to the origin via the shaft of a pendulum) and hence the distance L from the origin may vary with time. (a) Using polar coordinates as introduced in Sec. 14, what is the velocity vector ? (b) Show that the acceleration vector a = (d2/dt2)x is

(c) If L is independent of / and dO/dt is constant, sketch the trajectories. What direction is a. Is this reasonable ? (d) If 0 is independent of t, sketch possible trajectories. From part (b), show that the acceleration is in the correct direction. 14.3. Consider a mass m located at x — xi + yj and only acted upon by a force in the direction of f with magnitude —g(L) depending only on L = |5c|, called a central force. (a) Derive the differential equations governing the angle 0 and the distance L [Hint: Use the equation for a determined in exercise 14.2b], (b) Show that L2(d0/dt) is constant [Hint: Differentiate L2(d0/dt) with respect to /]. This law is called Conservation of Angular Momentum. 14.4. A particle's angular momentum around a fixed point is defined as a vector, the cross-product of the position vector and the momentum:

where * is the position vector relative to the fixed point. If all forces are

50

Mechanical Vibrations

in the direction of the fixed point (called central forces, see exericse 14.3), then show that angular momentum is conserved, i.e., show

14.5. Find an approximate expression for the radial force exerted by the shaft of the pendulum in the case of small oscillations. Show that the tension T «= mg. Is this reasonable ? Improve that approximation, and show that the radial force is not constant in time. 14.6. Consider a swinging spring of unstretched length L0 and of spring constant k attached to a mass m as illustrated in Fig. 14-10.

Figure 14-10

A swinging spring.

14.7.

14.8.

14.9.

14.10. 14.11.

Show that the only force in addition to gravity acting on the mass is — k(L — Lo)f. Derive the differential equations governing the motion of the mass. [Hint: Use the result of exercise 14.2]. One of Kepler's laws of planetary motion states that the radius vector drawn from the sun to a planet describes equal areas in equal times, that is, the rate of change of the area is a constant. Prove this using the result of exercise 14.3. [Hint: The differential area subtended is dA = %L2 dO (Why?), and thus what is dA/dtt]. Newton knew an experimentally determined value of g, the radius of the earth, and the distance the moon is from the center of the earth. Using this information and assuming the moon moves in a circular orbit, estimate how long it takes the moon to go around the earth. [Hint: Recall g — GM/rl]. Newton in the seventeenth century essentially used the preceeding ideas to help verify the inverse-square universal law of gravitation. Consider an object resting on the equator. It obviously moves in a circle with a radius of approximately 4000 miles. (a) Estimate the velocity of the object. (b) There are two forces holding the object from moving radially: the force exerted by the surface of the earth and the gravitational force. Show that the radial acceleration of the object is much less than the gravitational acceleration. Refer to exercise 14.9. Since most of you do not live at the equator, approximately calculate your present acceleration. In^what direction is it? From equation 14.3, show that f is perpendicular to 9.

Show that d0/dt = -f(dQ\dt\ 15.. [Hint: Use the result of exercise 14. i. The above formula is valid as long as/(. L — L(t) is known: Figure 14-11 A variable length pendulum. yields *The formula for the Taylor series (with remainder) of f(x) around x = xo is where x is an intermediate point (xo < x < x if x > XQ).t) is continuous in the interval and the derivatives of f(x) through the Mh derivative are also continuous.51 Sec. What error is introduced by neglecting all the nonlinear terms? An application of an extension of the mean value theorem (more easily remembered as the Taylor series with remainder*). . Show that df/dt is perpendicular to / 14. Consider any nonconstant vector/of constant length. as shown in Fig.] 14. How Small is Small? The nonlinear ordinary differential equation describing the motion of a pendulum was simplified to a linear one by using the approximation Is this a good approximation ? The Taylor series of sin 0. 14-11. Derive the differential equation governing the angle 0. The remainder appears like the N + 1st term except that the Mh derivative is evaluated at the unknown intermediate point x rather than at the point XQ.14. 15 How Small is Small? 14. The underlined expression is the first N terms of the Taylor series.e. is valid for all 0.13. but x is otherwise unknown.12. Consider a pendulum whose length is varied in a prescribed manner.2.

(The percentage error is 02/6. what is the actual percent error? (c) Compare part (a) to part (b).2. Figure 15-1. we must still investigate whether the solution to the linear equation.) At 0 equal one radian.1. For what angles is sin 0 «E 0 a valid approximation with an error guaranteed to be less than 10 percent? (a) Using the Taylor series with remainder. (since cos (1000^) = 1 and cos (1000. since | cos 0 \ < 1. but cos (lOOOrc) is not a good approximation to cos (1000. 15-1 we compare 0 to sin 0 using a set of tables. but 0 is otherwise unknown.52 Mechanical Vibrations where 0 is such that 0 < 0 < 0. d20/dt2 = —(g/L) sin0! Approximate equations do not always have solutions that are a good approximation to the solution of the exact equation. the error is at most 16 percent. E = —(03/6) cos 0. As an oversimplified example we know that (1000. It is seen that sin 0 % 0 is a good approximation. The error E in our approximation. yet even £ radian is not a particularly small angle. However. EXERCISES 15.5)71 in some sense approximately equals 10007T. what is the maximum percent error that occurs when sin 0 is approximated by 0 for 0 = 30° ? (b) For 0 = 30°. is a good approximation to the solution to the more difficult nonlinear equation. is bounded by 03/6. 15. in Fig. . even for angles that are not too small. d26jdt2 = —(g/L)0. In fact.571) = 0).57i). At 0 equal ^ radian. (not too bad an approximation considering one radian is about 57° which is not a very small angle). the error is reduced to at most about 4 percent.

To determine all solutions to the differential equation. This suggests three parameters of significance in the calculation. in which case .53 Sec.Q. However. and g/L. For example. Let us scale the time t by any constant time Q. if time is measured in a certain way. it appears we must vary these three parameters. Initial conditions are needed to solve a differential equation on the computer. In general. 16 A Dimensionless Time Variable 16. then we will show that only two parameters are important. let (g/L)Qz = 1 or Q = +/L/g. 00.0. by which we mean let Using the chain rule yields the differential equation to be solved with the modified initial conditions: Q is chosen such that there are less parameters necessary in the problem. A Dimension/ess Time Variable let us suppose that we need to compute solutions to this equation numerically.

we only have to vary these two parameters. Show. Using an appropriate dimensionless time variable. If the equilibrium position is x = 0. Let us describe what this scaling of time represents physically. by independently scaling both time and the angle 0. Show that £l0/cOo is a dimensionless parameter.3. and is thus called a dimensionless time variable! The only important parameters in the nonlinear problem (if we measure time based on the frequency of a small oscillation) is 00. we have shown Here the force. Why doesn't this work for the nonlinear pendulum ? 16. 17. The circular frequency of small amplitude oscillations is co0 — +/~gJL and thus The variable T has no dimensions. Nonlinear Frictionless Systems For a spring-mass system without friction. Consider a linearized pendulum. 00 and Q. To obtain numerical solutions. —/(*). that the three parameters 0 0 > ^o. the initial amplitude of the pendulum.1. depends only on the position of the mass. then the spring exerts no force there. . /(O) = 0.0+/L/g. and the ratio of the initial angular velocity Q0 to the circular frequency of small oscillations! EXERCISES 16. and g/L reduce to one parameter. what equation governs a frictionless spring-mass system with a linear restoring force ? 16.54 Mechanical Vibrations This resulting problem has only two parameters of significance.2.

x) is such that it can be expanded in a Taylor series. We will now consider motions of a spring-mass system such that the amplitudes are not necessarily small. . n. since Before solving these nonlinear ordinary differential equations. The nonlinear pendulum also satisfies a differential equation of that form. However. For larger amplitudes oscillations are still expected. if the pendulum is in that position and at rest it will stay there. 17 Nonlinear Frictionless Systems We assume that/(. Although both are equilibrium positions. what properties do we expect the solution to have? For small amplitudes the solution most likely oscillates periodically. at least for the nonlinear pendulum. Since 6E is a constant dOE(dt = d26Efdt2 = 0. is Hooke's law (the linearized spring-mass equation). as shown in Fig.If the force is a restoring force. Figure 17-2 Inverted equilibrium position of a pendulum. 17-2 is the "inverted" position of a pendulum: Figure 17-1 Natural equilibrium position of a pendulum. (Other mathematical solutions to this equilibrium problem are physically equivalent. It is only at these two positions that the forces will balance. For 9 = 6E to be an equilibrium position. then f(x) is positive for positive x (and vice versa). where k = /'(O).1 is appropriate. while 0E = n as demonstrated in Fig. yielding no motion. and hence k is positive.55 Sec. there are certain equilibrium positions for the nonlinear pendulum. Since /(O) = 0. Thus the result for small amplitudes of oscillation in which the nonlinear terms can be neglected. Futhermore. BE = 0.2 implies Consequently. 17-1. Then equation 17. and thus equation 17.1. there is a fundamental difference between these solutions that is immediately noticeable.) 0E = 0 is the "natural" position of a pendulum. 0 — 0£ must solve the differential equation 17. that is.

W. in some sense the solution does not tend towards the equilibrium. In general if equation 17. DiPrima. Boyce and R. More precise mathematical definitions of different kinds of stability can be given. For example. In the discussion of population dynamics later in the text. On the other hand an equilibrium is not stable just because there exists one initial condition such that the solution stays near the equilibrium. it is said to be unstable. we give a mathematical method to distinguish between stable and unstable equilibrium solutions. for example. but stays the same "distance away". for example. to be stable it must stay near for all initial conditions.56 Mechanical Vibrations QE = 0 is stable and 0E = n is very unstable! Any mathematical solution must illustrate this striking difference between these two equilibrium positions. As illustrated by the two equilibrium positions for a nonlinear pendulum. economics. We repeat. However. we will investigate the stability of equilibrium populations. an equilibrium solution is unstable if solutions tend "away" from the equilibrium and stable if solutions either tend "toward" the equilibrium or stay the same "distance away" (for example.1 is valid then there is an equilibrium position at any value of x such that f(x) = 0. chemistry. (See. for our purposes the abstractness of the rigorous definitions of stability is unnecessary. New York: John Wiley & Sons. When an equilibrium solution is not stable. Basically. If a mass m is acted upon by a force —/(*)» then . what is meant by "near" is carefully defined. Now we are pursuing this subject with respect to the stability of the equilibrium positions of a pendulum. 18. then the equilibrium is unstable. 1969. Elementary Differential Equations and Boundary Value Problems. the concept of stability is not a difficult one. In summary. an equilibrium solution of a time-dependent equation is said to be stable if the (usually timedependent) solution stays "near" the equilibrium solution for all initial conditions "near" the equilibrium.) In this section. and widely diverse fields of engineering and physics. even if only one initial condition exists for which the solution tends "away" from the equilibrium. Linearized Stability Analysis of an Equilibrium Solution The concept of the stability of solutions is considered to be one of the fundamental aspects of applied mathematics. for the natural position of the linearized pendulum the amplitude of oscillation remains the same.) In a precise discussion. Other areas in which stability questions are important include.

the displacement from equilibrium y approximately satisfies the above linear differential equation with constant coefficients. We say the equilibrium solution is stable if for initial conditions sufficiently near the equilibrium solution. As an approximation Although this equation can be explicitly solved. . we investigate positions x of the mass near its equilibrium position. as well as the higher-order terms of the Taylor series (if f'(xE) ^ 0).57 Sec. the function f ( x ) can be approximated using the first few terms of its Taylor series around x = XE: Thus Since x = XE is an equilibrium position. Thus the stability of the equilibrium solution is determined by the time dependence of the displacement from equilibrium. A simple analysis of equation 18. (x — xE)2 is much smaller than x — XE and consequently the quadratic term [(x — x^f'^x^/Il] can be ignored. For x near to XE.3. it is more convenient to introduce the displacement from equilibrium y: Using y as the new dependent variable The coefficient f'(xE) is a constant. the solution stays close to the equilibrium solution. 18 Linearized Stability Analysis of an Equilibrium Solution x — XE is an equilibrium position if To analyze the stability of this equilibrium position. Otherwise the equilibrium solution is said to be unstable. f(xE) = 0 and thus For x sufficiently near to XB.

in this case the force tends to restore the mass. If f'(xE) = 0. If displaced from equilibrium. such that €Xi(t) is now the displacement from equilibrium (also known as the amount the position is perturbed or the perturbation). indicating that the solution does not tend to equilibrium as / —» °°. 3. In this case we say the equilibrium position is stable. while 6E = n is an unstable equilibrium position. As an example. If /'(Xe) > 0> then the mass executes simple harmonic motion (about its equilibrium position). Since the solution consists of a combination of these two effects. shows that: 1. then this linearized stability analysis is inconclusive as additional terms of the Taylor series need to be calculated. the force will push it further away. We note that Thus. If this expression is substituted into the differential equation "'The term neutrally stable is sometimes used. f(6) = g sin 0. Thus in this case the equilibrium point is said to be unstable. An equivalent method known as a perturbation method is sometimes used to investigate the linearized stability of an equilibrium solution. let us investigate the linearized stability of the equilibrium positions of a nonlinear pendulum: The equilibrium positions are 9E = 0 and 0E = n. the displacement from equilibrium will exponentially grow for most initial conditions. 0E — 0 is a stable equilibrium position of a nonlinear pendulum. a small parameter 6 is introduced. For slight departures from equilibrium.58 Mechanical Vibrations known as a linearized stability analysis. as we know by our experience with pendulums. If /'(*E) < 0> then the system has some exponential decay and some exponential growth. 2.* This shows the importance of simple harmonic motion as it will describe motion near a stable equilibrium position. and therefore /'(#) = g cos 0. To facilitate remembering that x is near to XE. . 0 < e < 1.

2 and 14. a circular orbit is stable).] (c) Using a linearized stability analysis.59 then Sec.1. since XE is a constant.] "The symbol O(f 2) is read as "order c2.3 it was shown that (a) Show that m[(d2L/dt2) . for an inverse-square law.e. this reduces to equation 18. Suppose (a) Determine all possible equilibrium solutions.. EXERCISES 18. i." It indicates that the order of magnitude of the most important neglected term is f2.1. Using either method.2. In parts (b) and (c) assume the radial force is an inversepower law. The author believes this is unnecessary in this text. Again using the Taylor series it is seen that By neglecting the O(e2)* terms. In exercises 14. . (b) What is the radius L0 of an allowable circular orbit? [Hint: L0 = {c/[m(rf0A//)2]}1/(n+1>. [Hint: The answer is x = 0 and x = 1. 18.3. Assume that x = XQ is a stable equilibrium point of equation 18.. In this problem we will investigate the stability of circular planetary orbits. 18 Linearized Stability Analysis of an Equilibrium Solution results.e.3.L~ 3H%] = -g(L\ a second-order differential equation. A more careful treatment of this symbol can be developed. show that this circular orbit is stable only if n < 3 (i. the difficult to solve nonlinear differential equation is approximated by an easily analyzed linear differential equation. What is the period of small oscillations around that equilibrium point ? 18.

The gravitational force between any two masses is an attractive force of magnitude G(mim2lr2). (a) If the moonship is located at a distance y from the center of the earth.4.5.e. ql and q^ are the two charges. (a) If the middle particle is located at a distance y from one of the particles. The electrical force between two charged particles is —F(qlq2lrz) (alike charges repel and different charges attract). (b) Calculate the equilibrium position of the positively charged particle. assume that both are fixed in space. where r0 is the constant distance between the earth and moon.. show that 18. (where G is a universal gravitational constant. Consider an isolated positive electrically charged particle of charge qe (and mass me) located directly between two fixed positive charged particles of charge qpl and qpt respectively. where F is a universal electrical constant. and r is the distance between the two charges. (c) Is the equilibrium position stable ? Is your conclusion reasonable ? (d) Compare this problem to exercise 18. where a > 0.7.6. Consider a mass m attached to the exact middle of a stretched string of length /. show that 18.6. mi and m^ are the two masses. Which positions are equilibrium positions? Are they stable? Consider a system which satisfies 18. (c) Is the equilibrium position stable ? Is your conclusion reasonable ? (d) Compare this problem to exercise 18. (a) Show that the force does not always restore the mass towards x = 0. where r0 is the approximately constant distance between the two positively charged particles.) (b) Calculate the equilibrium position of the moonship. where a > 0. i. . (b) Is x = 0 a stable equilibrium solution ? (c) Is x = 1 a stable equilibrium solution ? Consider a spring-mass system with a nonlinear restoring force satisfying 18.60 Mechanical Vibrations 18. (Assume that the earth has no effect on the moon. (b) Which positions are equilibrium positions ? Are they stable ? Consider a stranded moonship of mass ms somewhere directly between the moon (of mass mm) and the earth (of mass me).7. and r is the distance between the masses).8.

we were able to analyze the solution of in the neighborhood of an equilibrium position. 19. Conservation of Energy In the previous section. (The tension is the force exerted by the string. show that the inverted position is unstable. We are especially interested now in determining the behavior of solutions to equation 19. equilibrium and the initial velocity are small. 18. Consider a nonlinear pendulum. (c) Describe the motion in the neighborhood of the equilibrium position. 18. 18-1. Calculate the tension T in the string.1 Vertically vibrating mass. First multiply both sides of equation .3. One of these constants can be obtained in a manner to be described. Figure 18.11. Consider equation 18. What is the exponential behavior of the angle in the neighborhood of this unstable equilibrium position ? 18.1 valid far away from an equilibrium position. Using a linearized stability analysis.10. The general solution to a second-order differential equation (even if nonlinear) contains two arbitrary constants.61 Sec. Here we continue the investigation of this nonlinear equation representing a spring-mass system without friction. show that the amplitude of the oscillation is small if the initial displacement from. Suppose (a) What are the dimensions of a and ft ? (b) Determine all equilibrium positions. Calculate the period of oscillation of the mass.) (b) Assume that the tension remains the same when the mass is displaced vertically a small distance y. If f'(xE) > 0. 19 Conservation of Energy (a) Suppose that the mass when attached forms in equilibrium a 30°30°-120° triangle (due to gravity g) as in Fig.9. x = 0.

Especially in cases in which/(. can be derived by noting . Thus This expression corresponds to the one obtained by indefinite integration if E = %mv\. it is said to be conserved. remains the same throughout the motion. The left-hand side is an exact derivative.t) does not have a simple integral.. since Thus.62 Mechanical Vibrations 19. i. An alternate expression. If there is a function F(x) such that dF/dx =f (i. The quantity ^m(dxjdf)2 + F(x). Both sides of this equation can now be integrated.1 by the velocity dx/dt. it is often more advantageous to do a definite integration from the initial position x0 with initial velocity v0.e. then indefinite integration or where E is a constant of integration. After multiplying by dt. F(x) = yields fX f(x)dx). which we will show is the total energy.e. more easily interpreted..

Then The quantity is again said to be conserved. Thus if the external force is —/(*)» i. Often (but not always) it is convenient to measure potential energy relative to an equilibrium position (i.e. The principle of conservation of energy is frequently quite useful (and always important).e.2 is called the equation of conservation of energy or the energy equation. as a mass speeds up it must gain its kinetic energy from the potential energy already stored in the system. since it is constant throughout the motion. It is thus in essence the work or energy that is stored in the system for "potential" usage. with gravity —mgj. The force necessary to raise the mass is minus the external force [for example.e. and hence is called the potential energy (relative to the position x = Xi). let xt = XE). to x is fx *Xl •"XI f(x) dx. for a variable force g. it is that portion of energy due to the motion of the mass (hence the term kinetic). Part of it is called the kinetic energy. consider an object being thrown vertically subject only to the force of gravity... Newton's law implies *The definition of work W is the force times the distance (i. In two or three spatial dimensions if the force is /. J *\ J Xl g ds). then the work necessary to raise the mass from jc. being initially equal to This constant of the motion is called the total energy..63 Sec. the work done by the . The total energy is shared between kinetic energy and potential energy. w = f* x -" force is W = f \ _ g-ds. 19 Conservation of Energy where xl is any fixed position. Equation 19. For example. m(d2x/dt2) = —/(x). the force necessary to raise a mass is +mgj]. As an example. f X f(x) dx is the work* necessary to raise the mass from x: to x.

Thus conservation of energy implies Consequently the highest point is Other applications are considered in the exercises. because initially the potential energy is zero since initially y = 0. a relative . 19-1. potential energy at an equilibrium position has an extremum point dF/dx = 0 (a relative minimum if f'(xE) > 0. yielding (for example) Fig. The applied force is —/(*)• Thus the derivative of the potential energy is minus the force.Hence at the highest point the kinetic energy must be zero. dy/dt = 0 at y = >W. The potential energy relative to the ground level (a convenient position since there is no equilibrium for this problem) is Conservation of energy implies that the sum of the kinetic energy and the potential energy does not vary in time.64 Mechanical Vibrations where y is the height above the ground. The derivative of the potential energy is/(x). but thrown upward with velocity v0 (the initial velocity dy/dt = v0). conservation of energy more immediately yields the result. Although it is not difficult to directly solve this differential equation (integrate twice!) and then determine the highest point (as was done in exercise 4. the velocity must equal zero. is known and is sketched. Suppose the potential energy relative to a position x = x.1). At that point all of the energy must be potential energy. The constant E equals the initial kinetic energy. A reasonable question is: What is the highest point the object reaches before it falls back to the ground. Since f(xE) = 0. Furthermore at the highest point y = jmax. Suppose the object is initially at y = 0.

Thus the potential energy has a relative minimum at a stable equilibrium position. a relative maximum. and if f'(xE) = 0 either a relative minimum. or a saddle point). an expression for the velocity. can be obtained. this first order differential equation is separable: Integrating from t = t0 (where x = x0). Furthermore. Similarly at an unstable equilibrium position. However. dxfdt. Recall in studying the linearized stability of an equilibrium position. . These facts are noted in Fig. 19 Conservation of Energy Figure 19-1 Potential energy F(x) illustrating equilibrium positions. it was determined that an equilibrium position is stable if f'(xE) > 0 and unstable if/'(jc£) < 0. this equation solves for t as a function of jc. From conservation of energy.65 Sec. Furthermore. It is easy to remember these results as they are the same as would occur if a ball were placed on a mountain shaped like the potential energy curve. 19-1. this formula is not particularly helpful in understanding the qualitative behavior of the solution. We usually are more interested in the position x as a function of t. yields an implicit solution where the appropriate sign again must be chosen in the integrand depending on whether the velocity is positive or negative. the potential energy has a relative maximum.2. equation 19. maximum if /'(*E) < 0. It is positive if the velocity is positive and vice versa. The sign of the square root must be chosen appropriately.

19.. Consider the equation of a nonlinear pendulum: (a) What are the two equilibrium positions? (b) Define the potential energy as Evaluate this potential and formulate conservation of energy.6. F (i.1. Compare this to the result obtained from the exact explicit solution. In general show that the potential energy when a mass is at rest equals the total energy. evaluate the maximum displacement of the mass from its equilibrium position. determine a quantity which is a constant of motion. (d) Show that this potential energy (as a function of 6) has a relative minimum at the stable equilibrium position and a relative maximum at the unstable equilibrium position. Derive an expression for the potential energy if the only force is gravity. 19.66 Mechanical Vibrations EXERCISES 19. (d) What is the velocity of the mass when it passes its equilibrium position ? 19. Suppose a mass m located at (x. A linear spring-mass system (without friction) satisfies m(d2x/dt2) = —kx. y) such that then show that the total energy (kinetic energy plus potential energy) is con- . If there exists a function <f>(x. The potential energy is defined as where Xi is any fixed position.4. (a) Derive that (b) Consider the initial value problem such that at / = 0. (c) Using the expression for conservation of energy.e. x = x0 and dx/dt = v0.3.2. 19. The equation of a linearized pendulum is L(d20/dt2) — —gO. 19. The kinetic energy is defined as it again equals ^mv2. By multiplying by d&fdt and integrating. y) is acted upon by a force field.5. Evaluate E. m(d2x/dt2) = F).

A mass m is thrown upwards at velocity v0 against the inverse-square gravitational force (F = -GmM/y2). 20 Energy Curves served.8. numerically integrate any initial value problem for a frictionless spring-mass system with a linear restoring force. (a) How high does the mass go ? (b) Determine the velocity at which the mass does not return to earth.2. A better understanding of the solution can be obtained by analyzing the energy equation 19. Let us consider x and dxfdt as variables rather than x and t.67 Sec. the so-called escape velocity. Is the energy constant ? 19. Energy Curves In the previous section a complex expression was derived for the motion of a spring-mass system. namely where the potential energy only depends on the position. (c) Estimate this value in kilometers per hour (miles per hour). and the total energy is constant. . representing conservation of energy. Using a computer. Such a force field F is called a conservative force field. 19. In this manner conservation of energy yields a relationship between x and dx/dt.7. In this case show that the potential energy equals 0(x) — $(xi). 20.

.1 as in Fig. the solution x(t) corresponds to one point on this curve since if x(t) is known so is dx/dt. The values of x are restricted. This coordinate system is called the phase plane. Along this curve energy is conserved.2. if f(x) = kx (let x{ — 0). x space. for each value of E: Figure 20-1 Typical energy curve. x space.68 Mechanical Vibrations Graphing equation 20. The actual graph of the curves of constant energy (corresponding to different constant values of £) depends on the particular potential energy function. 20-2. then Sketching the potential energy and the total constant energy yields Fig. E > kx2/2 as sketched in the hatched region. sketching a curve in the dx/dt vs. For these values of x there will be two possible values of dxfdt. greater than the potential energy. As time changes. the total energy is Figure 20-2. since we have expressed the equation in terms of the two variables x and dxfdt. the point corresponding to the solution changes. 20-1 will yield some curve in the dx/dt vs. For each time. referred to as the two phases of the system (position and velocity). determined from conservation of energy. The curve sketching the path of the solution is called the trajectory in the phase plane. In the next sections we will illustrate how to use this information to sketch the phase plane. For example. Since the kinetic energy is positive Qm(dx/dt)z > 0). equation 20.

69

Sec. 20 Energy Curves

Suppose part of one such energy curve in the phase plane relating x and dx/dt is known, and looks as sketched in Fig. 20-3. Although x and dx/dt are as yet unknown functions off, they satisfy a relation indicated by the curve in

Figure 20-3.

the phase plane. This curve is quite significant because we can determine certain qualitative features of the solution directly from it. For example for the curve in Fig. 20-3, since the solution is in the upper half plane, dx/dt > 0, it follows that x increases as / increases. Arrows are added to the phase plane diagram to indicate the direction the solution changes with time. In the phase plane shown in Fig. 20-4, since jc increases, the solution x(t) moves to the right as time increases. As another example, suppose that the curve shown in Fig. 20-5 corresponds to the solution in the phase plane. Again in the upper half plane dxfdt > 0 (and hence jc increases). However, in the lower half plane, x decreases.

Figure 20-4.

Figure 20-5.

Although the explicit solution of nonlinear equations only occasionally is easily interpreted, the solution in the phase plane often quickly suggests the qualitative behavior.

70

Mechanical Vibrations

**21. Phase Plane of a Linear Oscillator
**

As a simple example of the analysis of a problem using a phase plane, consider the linear spring-mass system

Although we already know the explicit solution (including the solution of the initial value problem), let us ignore it. Instead, let us suppose that we do not know the solution nor any of its properties. We will show how the energy integral determines the qualitative features of the solution. The energy integral is formed by multiplying the above equation by dxjdt and then integrating:

where the constant E can be determined by the initial conditions of the mass. Thus,

As a check, differentiating equation 21.2 yields equation 21.1. The potential energy is kx2/2. Since the kinetic energy is positive, only the region, such as that shown in Fig. 21-1, where E — kx2/2 is positive corresponds to a real solution, namely

The mass cannot have potential energy greater than the total energy. The energy equation relates x and dxjdt. A typical curve, defined by equation 21.2 corresponding to one value of E, E = E0, is an ellipse in the phase plane shown in Fig. 21-2 with intercepts at x = ±.+/2E0/k and at dx/dt = ±^2E0/m.

71

Sec. 21

Phase Plane of a Linear Oscillator

Figure 21 -1

Potential energy of a spring-mass system.

Figure 21 -2

Elliptical trajectory in the phase plane.

How does this solution behave in time? Recall, in the upper half plane x increases, while in the lower half plane x decreases. Thus we have Fig. 21-3. The solution goes around and around (clockwise) in the phase plane. After one circuit in the phase plane, no matter where it starts, the solution returns to the same position with the same velocity. It then repeats the same trajectory in the same length of time. This process continues and thus the solution is periodic. Or is it ? How do we know the solution in the phase plane doesn't continually move in the direction of the arrow but never reaches a certain point? Suppose that, as in Fig. 21-4, the solution only approaches a point. As illustrated, dxjdt and x tend to constants as t —»• oo. Can x steadily tend to a

Figure 21-3 plane.

Oscillation in the phase

Figure 21 -4.

71

Mechanical Vibrations

constant, and dxjdt tend to a constant? It can, only if dxjdt —»0. Thus Fig. 21-5 appears possible:

Figure 21-5.

Figure 21-6.

However, if jc —> ^lE^k as t —> oo, then the above trajectory implies that dx/dt must depend on time as shown in Fig. 21-6. As illustrated dx/dt approaches 0 but does not reach that point in a finite amount of time. Clearly, d2x/dt2 —» 0. The differential equation, d2x/dt2 = —(k/m)x, then implies that x—> 0, i.e., x -A +/2EQ}k as t—» oo. Consequently, the solution never "stops." It oscillates between a maximum value of x, x = +*/2E0/k, and a minimum value, x = —^/2E0/k. (Note that at the maximum and minimum values, the velocity, dx/dt equals zero.) Since the solution oscillates, perhaps as shown in Fig. 21-7, the solution is periodic in time. Many of the qualitative features of the solution in the phase plane agree with the exact results. Furthermore an expression for the period, that is the time to go once completely around the closed curve in the phase plane, can be obtained with-

Figure 21-7.

73

Sec. 21

Phase Plane of a Linear Oscillator

out the explicit solution. The velocity v is determined from equation 21.2 as a function of x,

Since v = dx/dt, it follows that

which can be used to determine the period. If we integrate equation 21.5 over an entire period T, then the result is

where <j> represents the integral of dx/v as the displacement x traverses a complete cycle (the plus sign in equation 21.4 must be used in equation 21.6 if v is positive and vice versa). This calculation is rather awkward. Instead, the time it takes the moving spring-mass system to go from the equilibrium position (x = 0) to the maximum displacement (x •= /^/2E0/k) is, by symmetry, exactly one quarter of the period, as indicated by Fig. 21-8. Integrating equation 21.5 in this manner yields

since in this case the sign of v is always positive. The integral in equation 21.7 can be calculated (by trigonometric substitution or by using a table of integrals), yielding the result

Figure 21 -8

Maximum displacement of a spring-mass system.

74

Mechanical Vibrations

which agrees with the result obtained from the explicit solution. Interestingly enough the period does not depend on the energy (i.e., the period does not depend on the amplitude of the oscillation). This is a general property of linear systems. Can you give a simple mathematical explanation of why for linear problems the period cannot depend on the amplitude? [Answer: For a linear (homogeneous) differential equation, if x(t) is a solution, then Ax(i) is also a solution for any constant A. Thus if x(t) is periodic with period T, Ax(t) is also periodic with the same period. The period does not depend on the amplitude parameter A} However, it will be shown that for a nonlinear system the period can depend on the amplitude. The only information not determined from the phase plane curve is that the solution is exactly sinusoidal (rather than some other periodic function). Appropriate integration of equation 21.5 shows the solution to be sinusoidal. Instead let us briefly show that the ellipses in the phase plane also follow from the knowledge that the solution is sinusoidal. Since x(t) — A sin (cot -+- $0)> where co — */k/m, it follows that dx/dt = Aco cos (cot + 00). The phase plane is a curve relating jc and dx/dt, but not depending on /. Thus, / must be eliminated from these two equations. This is accomplished as follows:

or equivalently In exercise 21.5, it is verified that

Actually only one closed curve in the phase plane has been illustrated. A few more curves corresponding to other values of the energy E are sketched in Fig. 21-9 to indicate the phase plane for the linear oscillator:

Figure 21 -9

Linear oscillator: trajectories in the phase plane.

75

Sec. 21 Phase Plane of a Linear Oscillator

The initial value problem,

is satisfied as follows. First the initial values (x0, v0) are located in the phase plane. Then the curve (see Fig. 21-9) which goes through it is determined (note that E = (mj2)vl + (fc/2)xj>), representing how x changes periodically in time. In summary, for equation 21.1 the energy curves determined the trajectories in the phase plane. Those energy curves are closed curves implying that the solution oscillates periodically. The amplitude of oscillation is obtained from the initial conditions. Thus the entire qualitative behavior of the solution can be determined by analyzing the phase plane.

EXERCISES

21.1. Suppose the motion of a mass m was described by the nonlinear differential equation m(dzx/dt2) = —fix3, where fi > 0. (a) What is the dimensions of the constant fi ? (b) What are the equilibrium positions ? (c) Derive an expression for conservation of energy. (d) Using a phase plane analysis, show that the position x oscillates around its equilibrium position. (e) If at t — t0, x = x0 and dx/dt = v0, then what is the maximum displacement from equilibrium ? Also, what velocity is the mass moving at when it is at x — 0 ? 21.2. Suppose that the motion of a mass m were governed by m d2x/dt2 = kx, where k > 0. Show that the phase plane indicates the equilibrium position (x = 0) is unstable. 21.3. Assume that a mass m satisfies m (dzx/dt2) = — x2. (a) What are the equilibrium positions ? (b) Derive an expression for conservation of energy. (c) Using a phase plane analysis, show that for most initial conditions the mass eventually tends towards — oo. Is that reasonable? However, show that for certain initial conditions the mass tends towards its equilibrium position. (d) How long does it take that solution to approach the equilibrium position ? (e) Would you say the equilibrium solution is stable or unstable ? 21.4. Show that the test for the stability of an equilibrium solution by the linearized stability analysis of Sec. 18 is inconclusive for exercises 21.1 and 21.3. Can you suggest a generalization to the criteria developed in Sec. 18?

76

Mechanical Vibrations

21.5.

The motion of a frictionless spring-mass system with a linear restoring force is described by where co = */k/m. Show that the total energy E satisfies

wherex0 is the inital position and vo the initiaovelocity of the mss 21.6 the phase place equation for a oiear oscillatr or can be used to directly obtain the solution. (a) show that

21.7.

(b) Assume x = x0 at t = 0. (Why is E> (&/2)x§?) Integrate the above expression to obtain t as a function of x. Now solve for x as a function of t. Is your answer reasonable ? Evaluate the following integral to determine the period T:

21.8.

Show that the period does not depend on the amplitude of oscillation. Consider the linear spring-mass system, equation 21.1. Show that the average value of the kinetic energy equals the average value of the potential energy, and both equal one-half of the total energy.

**22. Phase Plane of a Nonlinear Pendulum
**

The energy integral sketched in the phase plane can be used to determine the qualitative behavior of nonlinear oscillators. As an example consider the differential equation of a nonlinear pendulum,

Multiplying each side of equation 22.1 by ddfdt and integrating, yields

Here the potential energy has been calculated relative to the natural position

22-2 yields an expression for L/2(d0/dt)2.2 must be analyzed. *E here does not have the units of energy. Unfortunately equation 22. we sketch in Fig. as in Fig. the kinetic energy.2 implied that the trajectories were ellipses).2 is used as the basis for sketching the trajectories. which is easily related to d0/dt dependence on 0. differentiate equation 22. 22 Phase Plane of a Nonlinear Pendulum of the pendulum 0 = 0: At 0 = 0. Drawing vertical lines whose length equals the difference between the total energy E and the potential energy. we immediately noticed equation 21. Figure 22-2 gives a graphical representation of L/2(dQ/dt)2 dependence on 0. Instead multiplying equation 22. 22-1 the potential energy #(1 — cos 6) as a function of 6: Figure 22-1 Nonlinear pendulum: potential energy. Figure 22-2 Nonlinear pendulum: kinetic energy. 21. Thus mLE is actually the constant energy. First. which must be positive. where Q0 = d0/dt(t<>) and 00 = 0(t0).1. the energy integral equation 22.2 does not represent an easily recognizable type of curve (for the linear oscillator in Sec.77 Sec. equation 22. To sketch the trajectories in the phase plane (d0/dt as a function of 0).2 by mL yields an expression for energy. The energy E is constant and determined from the initial conditions. Again as a check. the "energy"* E consists only of kinetic energy. Instead. .2 with respect to / yielding equation 22.

The calculation of the slope of the curve sketched in Fig. all values of 0 do not occur. Figure 22-3 Positions of zero energy. 22. 22-2). From Fig. .78 Mechanical Vibrations The solution curve corresponding to E = 0 is trivial (see Fig. If the initial energy is zero. To sketch the remaining curves. it is observed that cos 0 > 1 — E/g is equivalent to as long as — n < 0 < n. The solution can only correspond to these angles. Only angles such that E > g(l — cos 6) or equivalently cos 0 > 1 — E/g are valid.e. where as sketched 0 < E < 2g. Our sketch of the phase plane is improved by noting that the magnitude of dOjdt is larger in regions where the difference between the total energy E and the potential energy is the greatest. These isolated points are the only stable equilibrium positions of a pendulum. as marked on Fig. The pendulum will not move from that position. 22-4 is outlined in exercise 22. For 2g > E > 0. 22-4. then the pendulum must be at its stable equilibrium position. where we have used the fact that (d0/dt)z — 2E/L when 0 = 0.2. Also the curves are even in 9. curves in the phase plane are periodic in 0.2). for each fixed energy level such that 0 < E < 2g. The evenness in 0 Figure 22-4 Trajectory in the phase plane. L/2(d0ldt)2 = g(cos 0 — 1) + E yields Fig. it helps to notice that the curves must be even in dOjdt (i. with period In. 22-3. We have included an arrow to indicate changes in the solution as t increases.. 22-2. replacing dOjdt by —dO/dt does not change eq. dO/dt — 0 and hence cos 0 = 1. Thus. Furthermore. Thus the curves are sketched only for ddjdt > 0 and 0 < 0 < n.

22-7. 22-5. the motion represents a periodic solution (though not sinusoidal) with larger and larger amplitudes. 22 Phase Plane of a Nonlinear Pendulum and ddjdt yields solutions which as before must be periodic in time. Sketching the phase plane for other values of E such that 2g > E > 0. As E increases away from zero. kinetic energy. For each fixed E in this range. Figure 22-6 Oscillation of a pendulum. Figure 22-7 Nonlinear pendulum: trajectories in the phase plane (for sufficiently small energy). the largest angle is called 0max (see Fig. the solution is nearly the periodic solution of the linearized pendulum. and energy curves. The periodic solution oscillates around the stable equilibrium position. as shown in Fig. . For small energy.79 Sec. yields Fig. 22-6): Figure 22-5 Potential energy.

80 Mechanical Vibrations Figure 22-8. Figure 22-10 Energy curve. Figure 22-9 Phase plane if E = 2g. E = 2g. Figure 22-11 Phase plane for a nonlinear pendulum (for sufficiently small energies including critical energy). .

] . Consider the phase plane determined by equation 22. sketch the angular velocity d6/dt as a function of0. x = —y/k. where y is the vertical distance of the pendulum above its natural position) and the kinetic energy (\mv2. (c) If E = 2g. In the phase plane. 22-11. the curve corresponding to E = 2g is that shown in Fig.] (b) If dx/dt < 0.6. For a representative value of the total energy E. Suppose a spring-mass system is on a table retarded by a Coulomb frictional force (equation 10. 22.1. the energy is at the level necessary for all angles to be possible as illustrated in Fig. repeat the calculation of part (a). sketch the solution in the phase plane. EXERCISES 22. Determine how many times the mass passes x — 0 as a function of VQ.] (c) Using the results of (a) and (b). Consider a nonlinear pendulum.3): 22. (b) Verify d/d0 (dO/dt) = oo at d8/dt = 0 (if E 7* 0 and if E ^ 2g) as also assumed in the figures. The energy integral enables us to sketch the trajectories in the phase plane. Thus. From the kinetic energy. 22. Show that the mass stops in a finite time!!! (d) Consider a problem in which the mass is initially at x — 0 with velocity v0. the still incomplete phase plane is sketched in Fig.3.81 Sec. not centered at jt = 0. Show that the sum of the potential energy (mgy. which has been used in the figures of this section. diagram the kinetic energy (the difference between the total energy and the potential energy). we have Fig. 2. Sketch the potential energy as a function of 0. Using this last result. [Hint: See exercise 19. determine the energy equation. [Hint: By completing the square show that the phase plane curves are ellipses centered at v = 0. 22 Phase Plane of a Nonlinear Pendulum If E = 2g. where v is the speed of the mass) is a constant. Sketch the resulting phase plane curves. [Hint: The ellipses now are centered at v = 0.2. and briefly explain how that information is used in the last three sketches of Sec. jc — y/k. 22-10. 22-8. Note the key steps: 1. calculate (d/d9) (dOjdt). (a) Show that d/d& (d0/dt) = 0 at 6 — 0 (if E ^ 0). 22-9. 3. (a) If dx{dt > 0.2.

E = 2g corresponds to initially starting a pendulum at 0 = n/2 with just the right velocity such that the pendulum approaches the top with zero velocity. It appears to reach that position with zero Figure 23-1 Pendulum approaching unstable inverted equilibrium. There is some theoretical difficulty with this solution as the uniqueness theoremt for ordinary differential equations implies that if the pendulum ever reaches the top with exactly zero velocity.2. since E = 2g. We would like to know what happens as 0 —* n. Thus our phase plane picture is correct (but slightly misleading). then it would have to stay there (both in positive and negative time). How long does it take the pendulum to get there if it is initially close to 6 = n with exactly the critical energy? From equation 22. We will show this in two ways. 23-1. Expanding the above energy equation in a Taylor series around 9 = n.82 Mechanical Vibrations 23. . One solution is 6 = n for all time and another solution is one in which d ^ n (at least for t < t*). The first technique will involve an approximation. taking an infinite amount of time to reach the top.1 there is a unique solution (for all time) satisfying any given initial conditions. Can a Pendulum Stop ? The phase plane. For example. for the limiting energy curve E = 2g shows that the pendulum tends towards the inverted position (either 0 = — n or 9 = n). 22-10. since that point is an equilibrium position. instead it only approaches the top. yields tThe theorem states that for equation 22. JThe difficulty is that there appears to be more than one solution corresponding to the initial condition at t = t* that 9 = n and dBjdt — 0. Fig. angular velocity (dBjdt — 0). How do we remedy this nonuniqueness difficulty % ? It will be shown that the pendulum never reaches the top. as shown in Fig.

equation 23.83 Sec. If E = 2g. it never reaches n). then dO/dt > 0 and if 0 > n.e.. Since if 0 — 00 at t = 0.1 where the positive sign is again chosen to ensure that 9 increases (assuming 6 < n initially). 9 = n.2 shows that the trajectories (for E = 2g) can be approximated near 0 = n by two straight lines with slopes in the phase plane equal to ± */g/L (as sketched in Figs. rather than 0. In that manner the following is a reasonable approximation: The ± . then it takes an infinite amount of time for 6 to reach n. For the case we are investigating in which the pendulum swings towards the equilibrium position. This equation implies that if 0 is initially (at t = t0) near n. then d0/dt < 0). is considered as the dependent variable). Hence t* — °o. then from equation 23. Thus 0 — n — Ae~<t~to)*/s/L. EXERCISES 23. 23 Can a Pendulum Stop ? We neglect all terms of the Taylor series beyond the first nonzero one. (i.1. then the time t* at which 0 — n is given by This integral is divergent as 0 —» n. Furthermore. Assume that a nonlinear pendulum is initially at its stable equilibrium position. 22-10 and 22-11). (if 6 < n.sign indicates a pendulum can be swinging clockwise or counterclockwise either towards or away from the equilibrium position. (a) How large an initial angular velocity is necessary for the pendulum to go completely around ? . This first-order constant coefficient differential equation is easily solved (especially if 0 — n. A more rigorous derivation of this result is now briefly discussed.

d6/dt(0) = Q0. . then how large an initial angular velocity is necessary for the pendulum to go completely around ? Consider d2x/dt2 = —f(x). since E — L/2(d0/dt)2 + g(l — cos 6). (a) Show that (b) For what value of the constant E is it possible that the solution approaches this unstable equilibrium point. then Fig. \ 6 \ keeps increasing (and —> oo). The entire qualita- Figure 24-1 Rotating pendulum. then it will take an infinite amount of time for x to reach x = x0. the sketch of the phase plane for the pendulum is completed in Fig. 24-1) and (since there is no friction) continue revolving indefinitely. We thus expect the pendulum to complete one cycle around (see for example Fig. numerically integrate differential equation 22.3. Estimate the value of Q0 for which the pendulum first goes all away around.84 Mechanical Vibrations 23. If E > 2g. For E > 2g. 23. What Happens If a Pendulum Is Pushed Too Hard ? If E > 2g. (b) At what initial angular velocity will the pendulum never pass its equilibrium position again ? If a pendulum is initially at its unstable equilibrium position. the pendulum rotates around and around (clockwise or counterclockwise in real space depending on whether 6 is increasing or decreasing). 23. 24-3. 24.2.1 with initial conditions 0(0) = 0. Using this result. Thus in the phase plane we have Fig.[Hint: Recall Sec. Compare your answer to the exact answer. 24-2 shows that all angles occur. 16]. and if x is near x0. but approaches it in a manner such that the velocity approaches zero. Using a computer. 24-4. (c) Show that if E is the value obtained in part (b).4. As expected. then there is more initial energy than is needed for the pendulum to almost go around. with x = x0 a linearly unstable equilibrium position.

Note that the curve in the phase plane for which E = 2g separates the phase plane into two distinct regions of entirely different qualitative behavior. 24 What Happens If a Pendulum Is Pushed Too Hard? Figure 24-2. Figure 24-4 Nonlinear pendulum: trajectories in the phase plane. in this case .85 Sec. Figure 24-3 Energy curve (large energy). Such a curve is called a separatrix. live behavior of the nonlinear pendulum has been determined using the energy curves in the phase plane.

Repeat exercise 24.2.1. 24.86 Mechanical Vibrations EXERCISES 24.3. 24-7. Describe the different kinds of motion that can occur. Sketch the solutions in the phase plane. 24. Assume that the-following equation describes a spring-mass system: . 24-5: Figure 24-5. Figure 24-7. Repeat exercise 24. Assume that the forces acting on a mass are such that the potential energy is the function of x shown in Fig.1 for the potential shown in Fig.4. 24. Figure 24-6. 24-6.1 for the potential shown in Fig.

87 Sec. an expression for the period is obtained (see Sec.7.] (c) Sketch the solutions in the phase plane. was obtained in Sees. (d) Suppose that a mass starts at x = — 1. 24-8: Figure 24-8. Suppose that a spring-mass system satisfies m(d2x/dt2) = — kx + a*3. (a) Determine all (if any) equilibrium positions. (b) Sketch the force as a function of x [Hint: Your answer should be consistent with part (a). 24. (b) Formulate conservation of energy. Interpret the solution (see exercise 18. (a) Locate all equilibrium positions. where a > 0.5). (c) Sketch the solution in the phase plane. 25 Period of a Nonlinear Pendulum 24. the qualitative behavior of the nonlinear pendulum. Period of a Nonlinear Pendulum Using the energy integral. Referring to Fig.6.5. 22-24. Sketch the solution in the phase plane. Is the period unaltered as the amplitude becomes larger ? Using the energy integral. (d) Explain how part (c) illustrates which of the equilibrium positions are stable and which unstable. where a > 0. .4). Sketch the solution in the phase plane. Suppose m (d2x/dt2) = -ke2"* where a > 0 and k > 0. 21). For what initial velocities will the mass reach x — 0 ? 25. 24. Suppose that the potential energy is known. For infinitesimally small amplitudes the period of oscillation is 2n+/Llg. Interpret the solution (see exercise 18.

In order to determine the manner in which the period T depends on the energy E (or 0max). Although this is still difficult to evaluate analytically or numerically.) This expression is valid when the pendulum oscillates back and forth as in Fig. It even causes trouble numerically (rectangles. it is seen that for E > 0 the denominator of the integrand in (25. (Recall. it can be approximated for small values of E. Since 1 — u > 0. 0max is the largest angle of oscillation.) as the denominator of the integrand —» 0 at the endpoint 6 = 0max. corresponding to the period of an infinitesimal amplitude. This integral is not easily evaluated.88 Mechanical Vibrations where #(cos 0max — 1) + E — 0. a change of variables is employed. For E = 0. trapezoids. Let Under this transformation. 25-1 (that is.2) is smaller than that which occurs when E = 0: Thus showing that the period of a nonlinear pendulum is larger than that corresponding to an infinitesimal amplitude. if E < 2g): Figure 25-1. Evaluating 7"(0) involves methods of integration. etc. Let . Simpson's rule. the limits of the integral do not depend on E.

. 7X0) equals the period of a linearized pendulum. integration-by-parts. saving the tedious effort of actually calculating a Taylor series via its definition.3 is again made. In that manner The binomial expansion is an example of a Taylor series. this quantity is much smaller than 1. To evaluate this additional term. Let us attempt to calculate the first effects of the nonlinearity.* Thus or.89 Sec. or integral tables (for the lazy ones among us). 25 Period of a Nonlinear Pendulum in which case. It is valid for all n (including negative and noninteger «) as long as | a \ < 1.2. Many approximations requiring a Taylor series need only an application of a binomial expansion. as expected. the transformation given by equation 25. Using it yields This last integral can be evaluated using trigonometric substitutions. The use of the binomial expansion facilitates the above calculation. the first few terms of a Taylor expansion of the integrand (around E = 0) will yield a good approximation Only the desire to keep the amount of calculations to a minimum. The period is given by equation 25. prevents us from developing additional terms in this approximation. Consequently. E is small and thus For E very small. equivalently.

show that a nonlinear pendulum has a longer period than the linearized pendulum.2. Figure 25-2.2.2. determine an expression for the time it takes a pendulum to go completely around. (b) Show that dTjdE > 0. where cos k = 1 — E}g. equation 22. (b) Also determine the period as a function of 6max.7. If 0max = 5°. 25.1. 25. (a) Relate the new variable w directly to 0. For larger values of E (corresponding to a larger maximum angle). Give a physical interpretation of this answer. The transformation w = ul/z has been used in Sec. approximately what percentage has the period of oscillation increased from that corresponding to a linearized pendulum ? 25. 25 can be analyzed by the trigonometric substitution suggested by the triangle in Fig.6. 25-2.3. Consider formula 25. (a) If the initial energy is sufficiently large. The dependence of the period on the energy has been determined for small energies. (a) From equation 25.] (b) The integrations performed in Sec. EXERCISES 25. Evaluate the period as a function of the energy £"and of 0 max . 25. as defined in Fig.5. 25-2. (a) Using a computer.90 Mechanical Vibrations Thus. 25. [Hint: Assume initially 0 = 0Q and dBldt = 0]. Numerically integrate (using a computer) this equation. (b) Estimate this time if the energy is very large (E ~^> 2g). Show that. Briefly describe a physical interpretation of this result. .1. 25. 25.1. for E small. Compare the results of exercise 25. numerically evaluate the period of a nonlinear pendulum as a function of the energy E. an expression for the increased period is obtained.4. the period may be obtained by numerically evaluating either equation 25. Consider the differential equation of a nonlinear pendulum.2. [Hint: You will need the trigonometric identity cos 20 = 1 — 2 sin2 6.1 or equation 25.3 to exercise 25.

for sin k/2 small.91 Sec. directly transform equation 25. For nonlinear oscillators the period (of aperiodic solution) depends on the amplitude of oscillation. Nonlinear Oscillations with Damping In the last few sections. energy is not expected to be conserved.. 26 Nonlinear Oscillations with Damping (c) Using if/ as the new integration variable. we have analyzed the behavior of nonlinear oscillators neglecting frictional forces. recall that with no friction (but allowing a nonlinear restoring force) a significant amount of information was obtained by considering the energy integral as it related to the solution in the phase plane. with a frictional force. i. reconsider the linear oscillator with linear damping. Approximate the period for E small.e. Let us attempt to form an energy integral by multiplying both sides of this equation by dxjdt and then integrating: . However. 26. More than one equilibrium solution is possible. As an example. We have found that the properties of nonlinear oscillators are quite similar to those of linear oscillators with the major differences being: 1.2. Since even in linear problems we know that friction cannot be completely neglected. In order to understand how to analyze this type of equation. 2. The forces depend only on the position and velocity of the mass. we proceed to investigate systems in which the frictional and restoring-type forces interact in a rather arbitrary way.

The most general form is . It decreases in time (see equation 26. We are unable to sketch the phase plane from this equation. We will return to this example in a later exercise. the energy equation is not convenient for a more detailed understanding of the solution. This is not the most general form of a second-order differential equation. However. 26-1. if we sketched ellipses in the phase plane (corresponding to the conservation of energy). i.e.3) and is said to dissipate. Here we have incorporated the mass m into the function/. Thus the solution cannot be periodic. the energy at t = 0. is the initial energy. then the solutions would continually remain inside smaller and smaller ellipses. perhaps as illustrated in Fig. Explicitly we see that the energy depends on time.. A slightly different technique can be used to obtain information about the solution. Thus. Figure 26-1 Energy dissipation in the phase plane.92 Mechanical Vibrations where £"<. as in equation 26. it is not constant.3 dx/dt does not depend only on x. Reconsider the general form of Newton's law allowing restoring and frictional forces.

The simplest property of an autonomous equation is that translating the time origin does not change the equation (see exercise 26. then this first order differential equation could be directly integrated yielding solution curves as before. However. becomes Using v and x as variables. the general autonomous equation.10). the solution of first-order differential equations can always be sketched in the following way. a short straight line is drawn with slope equal to g(x. v) as illustrated in Fig. but it will be shown that an autonomous equation can be interpreted as a relationship between dxjdt and x. . If an energy integral had existed. Suppose that At each value of jc. there is no explicit dependence on t. An energy integral will not always exist. Here we are not necessarily as fortunate since first order differential equations cannot always be solved explicitly. Autonomous equations are quite important since in many physical phenomena translation in time is insignificant. Such an equation is called autonomous. Through each point in a v-x plane. 26-2. 26 Nonlinear Oscillations with Damping In the equation representing Newton's law. Let A simple use of the chain rule shows that In this way.93 Sec. Qualitative features of the solution of autonomous systems can be obtained by considering the equation in the phase plane. a first order differential equation is derived. and in this text we restrict attention to such equations. the differential equation prescribes the slope of the solution dvjdx (if v is known).

v — 0 is an isocline. Sometimes to confuse matters they are the same.5 is sketched by noting that along the x-axis (v = 0) any solution must have an infinite slope. curves along which the slope of the solution is a constant sometimes can be calculated.5. This graph is called the direction field of the differential equation. To indicate this small dashes are drawn with infinite slope on Fig. *lso meaning equal as in isotope or isobar. but usually they are quite distinct from each other. isoclines are curves along which f(x. Figure 26-3. One isocline is immediately obtainable. These curves are called isoclines* (Be careful to distinguish between isoclines and solution curves. 26-3. At each point the solution must be parallel to these "dashes. dine meaning slope as in incline. Let us consider a specific example. As a review of the method of isoclines. the solution to equation 26. For a spring-mass system with a linear restoring force but without friction. .94 Mechanical Vibrations Figure 26-2 Direction field. Solution curves which cross the x-axis must be parallel to these dashes. dv/dx = oo. v)/v is constant." By roughly connecting these line segments. To facilitate sketching of the direction field.) From equation 26. the solution in the phase plane (v as a function of x) can be sketched (given x and v initially). namely when v = 0.

most general isocline. In particular. Letting v = dx/dt and using the chain rule (d2x/dt2) = v(dv/dx) yields the first order differential equation (Explicitly integrating this equation is equivalent to the energy integral. indicated by the vertical slashes. Other isoclines.) We sketch the solution by first drawing the isoclines. note the difference between the solution curves (ellipses) and the isoclines (straight lines). We indicate in Fig. dv/dx = A. Solving for v. In this manner a solution in the phase plane can be sketched (as indicated by the darkened curve in Fig. as along x = 0. are v = (k/m)x along which dv/dx = — 1. along which the slope of the solution is always infinite. and v = —(k/m)x along which dv/dx = +1. we see these isoclines are v = — (kjmX)x. Furthermore x = 0 is another isocline. From equation 26. 26-4. curves along which the slope of the solution is constant. .95 Sec. Actually a rough sketch might not guarantee solutions are the ellipses which 'mr previous analysis tells us they must be. As above along v — 0. we look for the curve along which the slope of the solution is the constant X. The isocline is the straight line v = 0. For this example all the isoclines are straight lines (in general isoclines need nor be straight lines). To locate the Figure 26-4 Isoclines for md2 x/dt2 = -kx. we will illustrate the method of isoclines to sketch the solution in the phase plane. 26-5). These isoclines are sketched in Fig. 26 Nonlinear Oscillations with Damping Although we can solve this problem explicitly or solve it by sketching the solution in the phase plane using the energy integral. dv/dx = oo. dv/dx — 0 (a constant). Any solution must be tangent to the slashes.6. 26-5 some of these other isoclines. for example.

Figure 26-6. dv/dt = f(x.) From dx/dt = v. 0). What happens as time increases? As before when considering energy curves in the phase plane. We note (returning to the general time-dependent equations). (Note that dividing one of these equations by the other yields equation 26. . it follows that in the upper half plane (v > 0) x increases (arrows point to the right) and in the lower half plane (v < 0) x decreases (arrows point to the left). v). Recall that f(x. arrows are used to indicate the direction of changes in time. the direction in time is determined by considering the other time-dependent equation. Along the line v = 0. 26-6. dv/dt = f(x.96 Mechanical Vibrations Figure 26-5 Direction field sketched by the method of isoclines. 0) is the "force" assuming no friction. see Fig. If this force is restoring (which in general it does not have to be). Along v = 0.5.

Figure 26-8 Phase plane for a linear springmass system. and thus v increases in the left half plane. 26-8. 26 Nonlinear Oscillations with Damping then it is concluded that for x > 0. 0) < 0 (and vice versa). Similarly for x > 0 at v = 0. For the example m(d2x}dt2) = —kx. Thus the method of isoclines implies (a) For this problem. These results yield (if the force is always restoring) Fig. the same result as previously sketched in Sec.97 Sec. v decreases in the right half plane. Figure 26-7. f(x. EXERCISES 26. Curves along which the slope (dv/dx) is constant are called isoclines. arrows can be added to the phase plane yielding Fig. show that the isoclines are all straight lines. . 26-7. Consider dv/dx — (ex + dv)/(ax + bv). 21. Hence dvjdt > 0 for x < 0 at v = 0.1.

(a) Show that the isoclines are not straight lines.) (c) Is it possible for the slope of an isocline to be the same as the slope of the solution ? In this case show that the isocline itself is a solution curve. (b) Sketch the solution in the phase plane. Briefly explain why only one solution curve goes through each point in the phase plane except for an equilibrium point in which case there may be more than one. Suppose that dv/dx = v2 — x. c — — 1.2. then roughly sketch the solution in the phase plane. (c) Interpret the solution.4.98 Mechanical Vibrations 26. k = 1. 26. (b) Let v — dx/dt and show that dv/dx — ( — cv — kx)/mv. k = 1. k = 1. Reconsider exercise 26. Show that both values of A are negative. let v = dx/dt and show that 26. k = 1. (b) If m = 1. (a) Show that the energy is an increasing function of time. c < 0. [Hint: Use the results of part (e)]. . (f) If m = 1.2 if friction is negative. then roughly sketch the solution in the phase plane. 26.3. 26. (b) Sketch the solution. c — \.6. (Use known information about the time-dependent solution to improve your sketch.7. (b) What is the slope of each isocline? (Be careful to distinguish between the slope of the solution and the slope of an isocline. (d) If m = 1. 26.) (e) Show that if c2 > 4mk. c — 3. Consider a spring-mass system with cubic friction Show that E = m/2(dx/dt)2 + k/2 x2 is a decreasing function of time. (g) Explain the qualitative differences between parts (d) and (f). Consider the linear oscillator without friction: (a) Without forming an energy integral. then v = 'kx is not a solution in the phase plane. then v = kx is a solution in the phase plane for two different values of A. Consider a linear oscillator with linear friction: (a) Show that E = m/2 (dx/dt)2 + (k/2) x2 is a decreasing function of time. (d) Explain the qualitative differences between parts (b) and (c).5. c = -3. (c) Show that if c2 < 4mk. (c) Repeat part (b) if m = 1. then roughly sketch the solution in the phase plane.

8 if a < 0. oo. Consider the general second-order autonomous equation: Show that if x — g(t) is a solution. Assume k > 0. 26-9 implies that E—»0: Figure 26. 26. then x — g(t — t0) is another solution for any t0. m > 0. Assume with c> 0 and /(O) = 0. x —» 0. show that Fig. Show that the conclusion of the problem may no longer be valid. and a > 0. (b) Using the method of isoclines sketch the solution in the phase plane (for ease of computation.8. a > 0. it corresponds to x growing or decaying exponentially in time. Show that if a solution in the phase plane is a straight line. [Hint: At least sketch the isoclines corresponding to the slope of the solution being 0.9 Energy decay.14. let m = 1 and k = 1). [Hint: Form an energy integral and show dE/dt < 0. ±1.99 Sec. (a) Give a physical interpretation of this problem.9. 26. Suppose that m d2x/dt2 = kx with m > 0 and k > 0.11.13. 26. Reconsider exercise 26. Since E > 0 (why ?). 26. Why can't £—* E0 > 0 as t —» oo?] 26. Consider (a) What first-order differential equation determines the solution curves in the phase plane? (b) What curves are isoclines? 26. show that there are no periodic solutions (other than x(t) = 0). (c) Explain how part (b) illustrates part (a).10.] . (b) By considering the energy. 26 Nonlinear Oscillations with Damping Consider 26.12. Show that as / —> oo. (a) Briefly explain why x — 0 is an unstable equilibrium position.

These singular points thus represent equilibrium positions. stability can be investigated most easily by considering a linearized stability analysis. and there are no forces at any such singular point. As in that problem we are quite interested in determining which such equilibrium points are stable. then it is reasonable to expand .1. 26). Equilibrium Positions and Linearized Stability The general autonomous system. Singular points occur whenever In other words. not quite: at any point where both the numerator and denominator is zero. 18. The analysis here differs from the previous one only by certain mathematical details now necessitated by the possible velocity dependence of the force. the velocity is zero. yields a first-order differential equation for the phase plane where v = dx/dt (see Sec. Suppose that x — XE is an equilibrium time-independent solution of the equation of motion. For example. As has been shown in Sec. v = 0.100 Mechanical Vibrations 27. The slope of the solution in the phase plane is uniquely determined everywhere. If x is initially near XE with a small velocity. Such points are called singular points of the phase plane equation 27. dvjdx is not uniquely determined since dv(dx = 0/0 (it depends on how you approach that point). Well. values of x for which the forces cancel if there is no motion. such points were encountered in the discussion of a nonlinear pendulum without friction.

101 Sec. Again introducing the displacement from equilibrium. it follows that where The notation used has taken advantage of the analogy of equation 27.4 to a spring-mass system with friction. However. where *The formula for the Taylor series of a function of two variables is or equivalently . 0) = 0 (XE is an equilibrium solution). exactly the kind analyzed earlier. Solutions are exponentials ert. 27 Equilibrium Positions and Linearized Stability f(x. where higher-order terms in the Taylor series have been neglected since x is near XB and dxfdt is small. z. here it is not necessary that k and c be positive! This equation is a constant coefficient second-order homogeneous ordinary differential equation. dxjdt) in a Taylor series of a function of two variables*: Since f(xE.

This information can also be communicated using a stability diagram in c-k parameter space. Unstable if c < 0. Stable if c> 0. the displacement from equilibrium does not grow. 27-1. dx/dt) = —(x — 4) + x3(dx/dt). Unstable if c < 0. suppose that We see that x = 4 is the only equilibrium position (dxjdt = 0 and d2x/dt2 = 0). Figure 27-1 Stability diagram (the hatched region is unstable). As an example. The equilibrium solution is said to be linearly stable if. we can determine the stability of x = 4 by simply calcuh ting the partial derivatives of f(x.e. . Fig. Otherwise stable (i. for all initial conditions near x = XE and v = 0.. in which case the solutions are e~ct/2 and te~ct/2).102 Mechanical Vibrations (except if c2 = 4k. Also unstable if c > 0 but k < 0. The following table indicates the behavior of the equilibrium solution: Unstable if c < 0. Stable if c > 0 (sometimes said to be neutrally stable* if c = 0 since the solution purely oscillates if c — 0). dx/dt): *See page 58. The equilibrium position is stable only if the linearized displacement z satisfies a differential equation corresponding to a linear spring-mass system k > 0 with damping c > 0 (except if c = k — 0). Letting f(x. c > 0 and k > 0).

1. The solution may or may not continue to depart from the equilibrium position. Suppose that where c and k are parameters which can be negative. When this occurs we can not rely on the results of a linear stability analysis. In the case in which the linearized stability analysis predicts the equilibrium solution is unstable. Eventually the solution is perturbed so far from the equilibrium that neglecting the nonlinear terms is no longer a valid approximation. 27. and k < 0. the displacement grows (usually exponentially). Thus 27. then what can be said about f(x) and g(dx}dt) ? Analyze the linear stability of the equilibrium solution x = 0. If x = 0 is an equilibrium point. The nonlinear terms in the neighborhood of the equilibrium position have been neglected. For the moment. we see x = 4 is an unstable equilibrium position.3. This determines what happens near the equilibrium position. 27 Equilibrium Positions and Linearized Stability From the table or the diagram above. Where in the phase plane is it possible for trajectories to cross ? Analyze equation 27. EXERCISES 27. To analyze this situation. positive or zero.103 Sec. but will be analyzed in later sections on population dynamics. (a) Under what circumstances does z oscillate with an amplitude staying constant? growing? decaying? (b) For what values of c and k do there exist initial conditions such that z exponentially grows ? decays ? Assume that for a spring-mass system the restoring force does not depend on the velocity and the friction force does not depend on the displacement. k = 0. 27. . let us just say that in "most" cases the results of a linearized stability analysis explains the behavior of the solution in the immediate vicinity of the equilibrium position.4. the solution can be discussed in the phase plane.2.4 if c = 0 with k > 0. Are we justified in doing so ? A complete answer to that question is postponed.

dv/d& = oo. Since an energy integral does not exist for equation 28. Nonlinear Pendulum with Damping As an example of an autonomous system. that in the upper half plane 0 increases since v = dQJdt > 0 (and arrows thus . Before this problem is mathematically solved. the phase plane is sketched. We also recall. If the frictional force is proportional to the velocity of the mass with frictional coefficient c. Along v = 0.1 becomes a first-order differential equation. we cannot integrate this immediately to obtain energy curves.1. Instead.104 Mechanical Vibrations 28. can you describe what you expect to occur 1. Can you envision a situation in which damping occurs in this manner? We recall the phase plane for the nonlinear pendulum without friction sketched in Fig. then where k = cLfm is positive. 24-4. or equivalently Unlike the frictionless pendulum. the angular velocity. If an extremely large initial velocity is prescribed? Hopefully your intuition is good and the mathematics will verify your predictions. k > 0. In particular we are now interested in determining effects due to friction. d20/dt2 = dv/dt = (dO/dtXdv/dd) = v dv/dO and hence equation 28. If a small angle with small velocity is initially prescribed? 2. we must again introduce the phase plane variable. There the direction field is vertical. let us consider a nonlinear pendulum with a damping force. for example.

it is underdamped if £ 2 < 4Lg (sufficiently small friction) and overdamped i f k 2 > 4Lg. 28-2. Instead the time-dependent equation must be analyzed. For example. — n < 0 < 0) and vice versa. This can be verified in a straightforward manner by doing a linearized stability analysis in the neighborhood of the equilibrium positions. the next most important isocline (and also usually an easy one to determine) is the one along which dv/d0 — 0.2. near 6 = 0. dv/dt — —g sin 0 at v = 0. It cannot be determined from the phase plane differential equation. 10-13).1. 27. From equation 28. They occur where dv/d0 = 0/0 (see Sec. equation 28.2. Along the isocline v = 0. 27). Note again that 0 = 0 is expected to be stable. 28-1): Figure 28-1. should arrows be introduced on the vertical slashes pointing upwards or downwards? The sign ofdv/dt at v = 0 determines whether v is increasing or decreasing there. From equation 28.705 Sec. (Can you give a physical interpretation of this result?) Consequently we have Fig. 0 — 0 is a stable equilibrium position. as suggested in Sec. Other than the isocline along which dv/d0 = oo (namely v =• 0). Thus at v = 0. the solutions in the phase plane must have vertical tangents. 28 Nonlinear Pendulum with Damping point to the right as in Fig. at v = 0 is v increasing or decreasing? In other words. The equilibrium positions are marked with an V. However. The angle 0 is damped. the curve along which dvfd0 = Ois . and that 0 = n is expected to be an unstable equilibrium position. dv/dt is positive where sin 0 is negative (for example. sin 0 is approximated by 0 and hence This equation is mathematically analagous to the equation describing a linear spring-mass system with friction (see Sees.

in this example the sinuous isocline that has been drawn in Fig. An alternate method to calculate the sign of dv/dt is to analyze the sign for very large v (both positive and negative). 28-3. As this curve is crossed.106 Mechanical Vibrations Figure 28-2. On one side of this curve dvfdt is positive. and the corresponding arrows yields Fig. Thus trajectories go downward (as indicated by j) above the sinuous sketched curve and vice versa. which it is in this case). Sketching this curve. the direction field. Then. then we know that the sign of dv/dt usually changes at h(v. 6). Figure 28-3 Nonlinear pendulum with damping: direction field corresponding to dv/d& equal 0 and oo. 6) = 0. the sign of dv/dt changes (if the zero is a simple zero. the sign of dv/dt should be calculated from If dv/dt = h(v. then dv/dt < 0. 28-3. and on the other side dvfdt is negative. For example if v > —g/k sin 0. Before we attempt to make sketches of the solution. since .

we will discuss a wide range of possible phase planes in the vicinity of equilibrium points. then the symbol \ is introduced in the corresponding region of the phase plane. Let us use arrows in the following way to suggest the general direction of the trajectories: for example. Before doing so. 47). 9 = 0. Our results for the nonlinear pendulum with friction are shown in Fig. 28-4. however. . let us discuss the two specific equilibrium positions that occur for a nonlinear pendulum with friction. if 9 decreases in time <— and v increases in time t. we note that often the qualitative behavior can be more easily obtained by first analyzing the phase plane in the neighborhood of the equilibrium positions. the phase plane is as shown in Fig. Figure 28-4 General direction of the trajectories. Near the "natural" position of the pendulum. is Figure 28-5 Damped pendulum: qualitative behavior of the trajectories in the neighborhood of the natural equilibrium position. The curve along which dvjdO = 0. Eventually in the context of population dynamics (see Sec. In every region of the phase plane. 28-5. v = — g/k sin 9. 28 Nonlinear Pendulum with Damping v must decrease if v is sufficiently large and positive. (How large is sufficiently large?) This yields the same result. it has been determined whether both v and 9 are increasing or decreasing with time. we may further use the method of isoclines. To determine the accurate behavior of the trajectories. In the meantime.107 Sec.

the above diagram suggests the motion may be one of the four types sketched in Fig.5. The linearized stability analysis in the vicinity of 0 = 0 (see equation 28. Without the linearized stability analysis. 28.3) helps to determine the approximate behavior in the neighborhood of 0 — 0 of the phase plane. 28-6. In the vicinity of the equilibrium position. . Note that all four sketches satisfy the qualitative behavior suggested by the arrows in Fig. the trajectories (a) spiral Figure 28-6Some different types of trajectories in the neighborhood of an equilibrium position (not necessarily for a pendulum).108 Mechanical Vibrations approximated near 0 = 0 by the straight line v = —(g/k)0.

28-7. the trajectories must be as illustrated in (a) and spiral inwards. As we have suggested. in which case we would say the equilibrium position is stable. in which case we say the equilibrium position is neutrally stable.. we see that if there is sufficiently small friction (i. Let us roughly sketch in Fig. a different kind of equilibrium position (details of this last case. since the solution is an underdamped oscillation. the decay of the energy implies that either (a) or (d) is the correct case (see Fig. The trajectories seem to tend towards the equilibrium position if they are in certain regions in Figure 28-7 Phase plane of a damped pendulum : qualitative behavior near the inverted equilibrium position. are given in Sec. From equation 28. 26. 47).2 discusses this latter case. the type of equilibrium position is easily determined from the linearized analysis valid in the vicinity of the equilibrium position. or (d) the trajectories tend directly towards the equilibrium position without oscillating around it. 28-8 some trajectories in the neighborhood of 0 = n: Figure 28-8. .109 Sec. the trajectories will be different. the phase plane. if k2 < 4Lg). Alternatively. or (b) the trajectories could "circle" around the equilibrium position (that is. called a stable node. Exercise 28. the effect of friction (even the slightest amount of friction. If k2 < 4Lg.e. the trajectories would be closed curves). 27). or (c) the trajectories could spiral out from the equilibrium position.2d the same phenomena when comparing a damped linear oscillator to an undamped one. no matter how small) is to transform the trajectories from closed curves (k = 0) to spirals. in which case we would say the equilibrium position is unstable. 28 Nonlinear Pendulum with Damping in towards the equilibrium position. while in other regions the trajectories move away. Clearly this indicates that 0 = n is an unstable equilibrium position of the pendulum (as we already suspect on physical grounds and can verify using the linearized analysis of Sec. If k2 > 4Lg.3.1). Can a similar analysis be done near 9 = nl See Fig. k > 0. This is not surprising because we observed in exercise 26.

Energy considerations. since 6 = n is unstable.110 Mechanical Vibrations Consider trajectories near the area marked © in the figure. What might this correspond to physically? Do you expebt that it ever gets there ? 2. must turn towards the right as illustrated by curve b. The motion would not be restricted to the area near 0 = 0. Thus there must be a trajectory (in between) which "enters" the unstable equilibrium position. Can you explain what is happening? What do you expect will eventually happen to that solution ? Determining the phase plane in the neighborhood of the equilibrium position is not sufficient to completely understand the behavior of the nonlinear pendulum with damping. For this case additional analysis would be necessary. Two of the trajectories seem to be tending towards the unstable equilibrium position. the trajectories that enter a saddle point can be shown to be approximated by two straight lines as sketched (see exercise 26. We can easily imagine an initial condition near 6 — 0 with a large angular velocity such that the pendulum does not remain near 6 = 0. In Fig. To investigate solutions for which the angle is at some time far away from an equilibrium. Figure 28-9 Trajectories for a damped pendulum in the vicinity of the unstable equilibrium position. 28-10 (assuming that k2 < 4Lg). show that the trajectory emanating from 0 = —00 (0 < 00 < n) with . Some trajectories on the left must cross the isocline along which dv/d0 = 0 and then curve. for example. 28-9. 28-9 consider the trajectory marked (2). In the neighborhood of the equilibrium. 47B we will call such an equilibrium position a saddle point. In Sec. more to the right. downward as illustrated by curve a. in the phase plane we can roughly connect the solution curves that are valid in the neighborhood of 6 = 0 and 0 = n as illustrated in Fig. In a similar manner we can easily see that there are four trajectories which enter this unstable equilibrium position (two enter backwards in time) as illustrated in Fig.1).2 and the further developments of exercise 28. Furthermore. Some thought questions (not intended to be difficult) follow: 1. Others. we will rarely be interested in trajectories that remain near 6 = n.

the pendulum oscillates with smaller and smaller amplitude. the pendulum will go around a finite number of times (this contrasts with the frictionless nonlinear pendulum in which the pendulum continually goes around and around). Figure 28-10 Phase plane if k2 < 4Lg: sketch illustrating trajectories in the neighborhood of both equilibrium positions. but never gets there). Along the curve More generally along the curves Using these isoclines we obtain an improved sketch of the phase plane in Fig. If a sufficiently large initial angular velocity is given. The sketch of the trajectories of the slightly damped nonlinear pendulum shows that we can understand the solution of a complicated mathematical problem without obtaining an explicit solution. 28-11 (if A:2 < 4Lg). The maximum displacement of a pendulum diminishes after each oscillation due to the small friction (as illustrated in Fig./// Sec. Under certain determinable circumstances. 28 Nonlinear Pendulum with Damping v — d0/dt = 0 must fall short of 6 = 90. in particular. the method of isoclines should be systematically employed. when again v = 0.2 modifies the above discussion when k2 > 4Lg. The pendulum will be continually slowing down and eventually it will not be able to go completely around. to sketch the phase plane for large velocities. To improve this rough sketch. Then the pendulum will oscillate around its natural position with decreasing amplitude (except in the very special cases in which the pendulum approaches the inverted position with zero velocity. and. Exercise 28. . 28-10).

1. [Hint: See exercises 26.2 and 28.1. sketch the solution in the phase plane if friction is sufficiently large. Pay special attention to the phase plane in the neighborhood of 9 = 0 and 0 = n. g = 1.3. k = 3. Show that one straight line has positive slope and the other negative. (a) Approximate sin 0 in the neighborhood of the unstable equilibrium position.1.2. . 28. equation 28. (c) Show that two solution curves are straight lines going through the origin (v = 0. If a spring-mass system has a friction force proportional to the cube of the velocity.2]. you may assume that L = 1.] 28. (b) Show that the resulting approximation of the phase plane equation can be put into the form where 0* = 0 .1. Show that there are straight line solutions in the neighborhood of both 0=0 and 0 = n. For the nonlinear pendulum with friction. then (a) Derive a first-order differential equation describing the phase plane (dx/dt as a function of x). For sketching purposes. [Hint: See exercise 26. k2 > 4Lg. EXERCISES 28.n. 9* = 0).112 Mechanical Vibrations Figure 28-11 Trajectories of a pendulum with damping (if k2 < 4Lg). Consider equation 28.

Describe any interesting features of the solution.7. roughly sketch the phase plane.4. what happens ? (c) Is the equilibrium position x = 0. show that (c) Instead of sketching the isoclines. 28 Nonlinear Pendulum with Damping 28.6): where ft > 0. Using this solution. (d) Show that the solutions of parts (b) and (c) verify part (a). what do you expect happens ? (e) Sketch the trajectories in the phase plane if a> = 1 and € = ^. Consider the spring-mass system of exercise 28. k = 0). Under what condition does the pendulum continually oscillate back and forth with decreasing amplitude? Consider a nonlinear pendulum with Newtonian damping (see exercise 10. (b) By introducing the phase plane variable v = dO/dt. Under what conditions does the + or — sign apply? This is a linear differential equation for v2. Consider a linear pendulum with linearized friction 28.e. 28 ? The Van der Pol oscillator is described by the following nonlinear differential equation: (d) (e) (f) (g) where € > 0. . (a) Briefly describe the physical effect of each term. show that 28. linearly stable or unstable ? (d) If displacements are large. (b) Sketch the solution in the phase plane. Solve this equation. (a) How do you expect the solution to behave ? (b) Let v = dx/dt and sketch the solution in the phase plane.6. (a) Show that an energy integral does not exist. (b) If € = 0. What qualitative differences do you expect to occur between this problem and the one discussed in Sec. (c) Let v — dx/dt and solve the problem exactly. 28..113 Sec.3 without a restoring force (i.5.

our goal is only to introduce the concepts of applied mathematics. only some of the simplest models of mechanical vibrations have been introduced.1. Assume A 2 < 4Lg. Rescale equation 28. 29.J. 28. However. New York: Interscience..iV Determine how many times the pendulum goes completely around as a function of Q 0 28. Nonlinear Vibrations. Reconsider the Van der Pol oscillator of exercise 28. CHAIKIN. A.L = l.: Princeton University Press.. and thus we may end our preliminary investigation of mechanical vibrations at this point.g = 1).114 Mechanical Vibrations 28. Using a computer. (a) Numerically integrate the differential equation with CD = 1 and (b) Can the scaling of time justify letting CD = 1 always ? (c) Compare your numerical results to the sketch in the phase plane (exercise 28. A.1 to determine the important dimensionless parameters. (This includes a good discussion of the theory of the clock!) STOKER. 28. J.7.10. 1950. Further Readings in Mechanical Vibrations In the preceding sections.7e). 1949. N. We progressed from linear undamped oscillators to nonlinear ones with frictional forces.11.9. Princeton. The behavior of the nonlinear systems we have analyzed seem qualitatively similar to linear ones. I refer the interested reader to the following excellent books: ANDRONOW. solve the initial value problem: f . However.. show that 0 — n is an unstable equilibrium position of a nonlinear pendulum with friction. Theory of Oscillations. By including other types of nonlinear and frictional forces (for example. J. (Let A: = 1.8. For further studies. A. A. we could find behavior not suggested by linear problems especially if external periodic forces are included. Using a linearized stability analysis. E.. Consider equation 28. C. as occur in certain electrical devices). we have not made a complete mathematical analysis of all possible problems. Other interesting problems involving the coupling of two or more springmass systems and/or pendulums as well as problems involving rigid bodies in . and WITT.

Mass. H. D.. LANDAU. L." In particular. Mechanics. Reading. Classical Mechanics. two well-known ones are: GOLDSTEIN. 29 Further Readings in Mechanical Vibrations two and three dimensions can be found in a wide variety of texts. whose titles often contain "mechanics" or "dynamics. M. ..: Addison-Wesley.: Addison-Wesley. 1950. 1960. E. and LIFSHITZ. Readings Mass.115 Sec.

This page intentionally left blank .

Population Dynamics— Ma th ema tical Ecology .

This page intentionally left blank .

In the following sections we will develop mathematical models describing one aspect of the biological world. population dynamics. and nuclear physics. There are probably many reasons for this. mathematics has not been as successful a tool in the biological sciences as it has been in the physical sciences. We cannot expect simple models of the biological world to accurately predict population growth. We will analyze the fluctuations of these populations. We might want to try to understand the complex ecosystem of a lake. many assumptions will be made. Furthermore. we might model the entire world's plant and animal populations (consisting of many smaller ecosystems). hence the other part of the title of this chapter. The mathematical models that are developed are frequently crude ones. we should not hesitate to discuss simple models. it is not always apparent what factors account for observed population variations. On the most grandiose scale.30. an ability that cannot be attributed to a spring-mass system nor a pendulum. the simplest of which man is just learning to comprehend. 119 . and their environment. Here we will attempt to mathematically model processes involving a few species. for it is reasonable to expect that such models may be as significant in the biological sciences as spring-mass systems are in contemporary quantum. atomic. the mutual relationships among plants. many animals possess an ability to choose courses of action. animals. In addition. made up of many components. a field of study known as ecology. There seems to be no fundamental biological law analogous to Newton's law. We will formulate models based on observed population data of various species. As in the modeling of any problem. However. Observations of population changes are limited. involving the interaction of many species of plant and animal life. Thus the scientific community is a long way from understanding cause-eifect relationships in the biological world to the same degree that such laws exist for spring-mass systems. Surely it is reasonable to proceed by asking simpler questions. the number of individuals of different species is investigated. such as human population in specified areas or the population of fish and algae in a lake. Plants and animals are complex. As a means of quantifying this science. Another example might be to study the growth of a forest. Introduction to Mathematical Models in Biology In the past.

The previous population data were observed continuously in time. which must be modified to model more realistic environmental influences (Sees.e. or the number of bacteria in an experiment. This curve is discontinuous since changes in the monkey population occur in integral units (+1 for single births. the limitations of the observed data suggest that it might be only necessary to model the population at certain discrete times.. For example. What kinds of data might be observed ? Perhaps the number of monkeys in a laboratory as a function of time would be as shown in Fig. In many situations involving a large number of a species. and so on). age differences). it is reasonable to approximate N(t) as a continuous function of time. where each ti represents the time at which a measurement was made. 45-47). 32-36). 31. 31). the bear population in a forest might be estimated only once a year. Specific predator-prey and competing two-species models accounts for the concluding parts of our discussion of population dynamics (Sees. However. +2 for twins. it is frequently possible to use continuous functions of time to represent populations. 37-39). In this manner. . 43-44). 48-54). male-female differences. Our first models involve the simplest birth and death processes of one-species ecosystems (Sees. Experiments suggest that these models are also at times inadequate. 31-1. explicit solutions. Throughout our study we focus our attention on the modeling process. We ignore any differences in the individuals comprising the group (i. 40-42). qualitative behavior.120 Population Dynamics—Mathematical Ecology We start our study of mathematical ecology by briefly discussing some basic assumptions of the population models utilized in this text (Sec. whether it is the number of people in the world. some populations are normally measured periodically. the number of monkeys. Population Models We begin the. study of ecosystems. Fundamental concepts (as with mechanical vibrations) involve equilibriums and their stability. Although this could be modeled as a continuous function of time (by again fitting a smooth curve through the data points). motivating our investigation of population growth with time delays (Sees. — 1 for deaths. Then we begin to model more complex ecosystems involving the interaction of two species (Sees. Thus data might be N(ti). pausing to discuss some necessary mathematics (Sees. and the resulting ecological interpretations. can only be an integer. perhaps by fitting a smooth curve through the data. Furthermore. by considering the population of one species in a specified region. N(t). the number of pine trees in a forest.

It would be incorrect to assume that the population of sharks is affected only by other species. Data is gathered through a census taken every ten years. trying to make no variation in the initial population nor in the laboratory environment. What caused the differences ? It seems impossible to exactly repeat a given experimental result. The data to which a mathematical model is compared may be inaccurate. We might observe the graph in Fig. we do not have complete control of the experiment. sharks would become extinct). Clearly in some cases it depends on many quantities. (Is this also true for a . Suppose we perform such an experiment starting with N0 animals and model population as a continuous function of time N(t). For example: the population of sharks in the Adriatic Sea will depend on the number of fish available for the sharks to consume (if there are none. We might observe the second graph in Fig. In addition. We should not be surprised if. let us rerun the hypothetical experiment. Now we will study a simpler species. Thus. Other factors may also be significant. Before attempting to analyze this situation. a field of study in itself. We will model the population of sharks later. we must decide what factors affect that population. the water temperature and salinity (salt content) are important in determining the shark population. for example. 31-2. Here. Such a species might be observed in a laboratory experiment of well-fed animals. We would be quite surprised if we got identical results. 31-2. we will not pursue the question of how to analyze data with inaccuracies. when considering the population of the United States as a function of time. However. 31 Population Models Figure 31 -1 Discontinuous population growth. the presence of a harmful bacteria will affect a number of sharks.121 Sec. An additional difficulty occurs. for example. one not affected by any others. In formulating a model of the population growth of a species. the accuracy of these census figures have been questioned.

122 Population Dynamics—Mathematical Ecology Figure 31-2 Experimental variability. However. The rate of change of the population as measured over the time interval A/ would be This indicates the absolute rate of increase of the population. Typical data on the variations of the population of a species in a specified region might be as represented in Fig. if a later experiment or observation does not correspond precisely to a prediction of a mathematical model. one of the simplest models of population growth of a species is developed. This is called the growth rate per unit time (for example. A quantity which will prove to be quite important is the rate of change of the population per individual. Instead. This randomness in the model would predict different results in each experiment. A Discrete One-Species Model In this section. 32-1. we will model some type of average in many experiments. To account for the observed variability from experiment to experiment. . we might introduce some random quantities into the mathematical model. 32. we will almost exclusively pursue deterministic models. then it may be the result of some randomness rather than some other inherent failure of the model. R(t). rather than attempting to model each specific experiment. However. Thus. because of the author's own orientation and because a complete discussion of probabilistic models requires a previous course in probability (a prerequisite the author did not want for this text). in many ways the best agreements with experiments occur with probabilistic models. where measurements might have been taken over an interval of time At. there will not be much discussion of such probabilistic models. in this text. spring-mass system?) To account for this (perhaps caused by an indeterminateness of some environmental factors).

. if the growth rate and the initial population were known. is The growth rate /?. the population at a time Af later. Thus one hundred times the growth rate R(t) is the percentage change in the population per unit time. then R(t) = f and the growth rate is 40% per year (as measured for one-half year). N(t + Af). There is no migration into or out of the region. Thus The reproductive (birth) rate b per unit time measured over the time interval Af and the death rate d are defined as Consequently. 32 A Discrete One-Species Model Figure 32-1 Typical data on population N(t). However.725 Sec. per year) as measured over the time interval A/: The percentage change in the population is 100 AN/N(t) = 100 /?(/) Af. if in one-half year the population increases by 20%. then the population at later times could be calculated: We assume that the population of the species only changes due to births and deaths. For example.1 cannot be used to determine the population at future times since it is just the definition of R(t). No outside experimenter slips some extra species into the system. Equation 32.

If the growth rate is constant. In recent years. the future population can be easily computed.019. As a first step in the mathematical modeling of population growth. A difference equation has certain similarities to a differential equation. However. Thus the growth rate R is a constant. the unique solution can always be directly calculated. averaged over this entire population. We are not distinguishing between older or younger individuals. it is assumed not to change in time. This figure gives no other information concerning the birth and death rates. the birth and death rates are averages. A twofold increase in the population yields twice as many births and deaths. In discussing human population growth. This means that the growth rate (the birth rate minus the death rate) is 1. Two populations are likely to grow quite differently if one has significantly more senior citizens than the other. assuming the effects of a possibly changing age distribution can be neglected. R = RQ.9 percent per year. we assume that the number of births and the number of deaths are simply proportional to the total population. We now proceed to discuss the total population of a species. They realize that accurate predictions of future growth depend on a thorough knowledge of the age distribution within the population. then for any t This can be expressed as a difference equation for the population The population at a time At later is a fixed percentage of the previous population. that is given an initial population at / = f 0 . the world human population growth rate approximately equals . Without arguing the merits of such an assumption. Thus the mathematical model we are developing can be improved to allow for an age distribution in the population. We will show this difference equation can be solved as an initial value problem. None of the "tricks" of . Since we focus our attention on the total population in a region. for the initial value problem of this type of difference equation. actuaries and demographers would be upset with our approach. let us pursue its consequences. 35). This will be briefly discussed in a later section (Sec.124 Population Dynamics—Mathematical Ecology is the birth rate minus the death rate.

Not only did the death rate dramatically increase. the population grows. it is clear that a general formula exists. amount. the British economist Malthus used this type of population growth model to make the pessimistic prediction that human population would frequently outgrow its food supply. Growth occurs over each discrete time interval of length A/. At m units of At later. A sketch of the solution is easily accomplished by noting where a is a constant. Around 1800. We illustrate some environmental factors which have caused human growth rates to vary: 1. if R0 > 0). 32 A Discrete One-Species Model differential equations are necessary.125 Sec. but immigration to the United States (and elsewhere) was so large that during the years that followed the population of Ireland . In each interval of time At the population increases by the same rate. or equivalently. The assumption that the growth rate is constant frequently does not approximate observed populations. Although this method gives a satisfactory answer for all times.e. For a constant birth rate. 32-2. but not by the same Figure 32-2 Constant growth rate. if R0 > 0 we have Fig. a = ln(l + R0At). Malthus did not foresee the vast technological achievements in food production. If the birth rate is greater than the death rate (i. t = f 0 + /wAf.. The failure of the potato harvest (due to blight) in Ireland in 1845 resulted in widespread famine. Thus. rather an increasing amount.

5 6. The average number of desired children seems to depend on economic and other factors. 32. the reproductive capacity) and fecundity (the actual rate of reproduction). Examples (2)-(4) vividly illustrate the difference between fertility (the ability to reproduce. but negative.2.5 8. as the result of curfew laws in Chile in 1973. Assume that a savings bank gives / percent interest (on a yearly basis) and compounds the interest n times a year.7 4.3.5 6. for example.5 2. (a) Formulate the equation describing the change in the population. A famous long power blackout in 1965 in the northeast United States resulted in an increased growth rate nine months later. that there is an increase in the population of a certain species due to the migration of 1000 individuals in each Af interval of time. (b) Explicitly solve the resulting equation. in addition to births and deaths (with constant rates b and d respectively). Assume that the growth rate of a certain species is constant. Population estimates for Ireland speak for themselves: Estimated Population of Ireland (Including Northern Ireland) Year Population in millions 1800 1845 1851 1891 1951 1971 4. 3. The pill and other birth control measures have contributed to decreases in the 1960s and 1970s in the growth rate in the United States. What is the predicted population as a function of time if the species numbers 2000 in 1950? Suppose. . During the depression in the 1930s.3 4. Sketch the population if at t — 0. your answer reduces to the correct one.1. Assume that the initial population is AV (c) Verify that if b = d. birth rates in the United States were lower than they were both before and after. 32. If N0 dollars is deposited initially. 32. What happens as t —> oo ? Is your answer expected ? Suppose that the birth rate of a species is 221 per 1000 (per year). N = N0. EXERCISES 32.126 Population Dynamics—Mathematical Ecology significantly decreased.4. and the death rate is 215 per 1000 (per year). 4. This effect also occurred.

Assume that a savings bank gives 6 percent interest (on a yearly basis) and compounds the interest 4 times a year. 32. [Hint: For y small. (c) For a typical interest rate of 5 percent. For example. 32. how much more money would you save in the bank that compounds 4 times a year? Most banks advertise their interest rate (on a yearly basis) and also advertise their method of compounding. If you deposit $100. ev = 1 + y + . 32. We describe in this problem how to approximate the yield if I/n is small (this occurs often as either « is large or the interest rate / is small). (a) Determine the Taylor series of In (1 + x) around x — 0 by term-byterm integration of the Taylor series for 1/(1 + x).5. . some banks compound interest daily. For one of your local banks. 32 A Discrete One-Species Model then show that 32.127 Sec.7 is in competition with a bank which offers the same interest rate. Show that your cal- .7. We now will approximate In (1 + I/n) for IIn small..8. approximately how much more money is made in a year in a bank compounding every day at 5 percent than one compounding every month at 5 percent ? The hint of part (c) may again by helpful. .6. The yield (derived in exercise 32. let and hence In C = n In (1 + 1/n). but only compounds 3 times a year. or equivalently that C = el-I i /2n + . where t0 is the time of the initial deposit. how much money is in your account after one year? Suppose that the bank described in exercise 32.9.5) can be difficult to evaluate if n is large.] (d) If $100 is the initial deposit.. [Hint: The sum of an infinite geometric series implies that (if \ x \ < 1) 32. . (a) What is the yield? (b) Evaluate the yield if n = 1 or n = 2. compute their annual yield..] Referring to exercise 32.. To analyze (1 4.///*)" for I/n small.72/2« + . taking their advertised information. show that after one year the total interest paid is The yield is defined as the total interest at the end of the year divided by the money on deposit at the beginning of the year. (b) Thus show that In C = I . In the first year. and for that reason we may wish to evaluate the yield if n = 365. compare the yield of daily compounding (n = 365) to the yield of continual compounding (the limit as n —> oo). [Hint: First show that A/ = l{n and N(t + Af) = N(t)(l + I/n).4.

If this is the case. then how much money must be initially deposited such that the $50 can be withdrawn every month FOREVER! 32. Suppose that each time a savings bank compounds the interest. check your work carefully. Suppose that you borrow Pa dollars (called the principal) from a bank at / percent yearly interest and repay the amount in equal monthly installments of M dollars.10. then how much money have you paid in total for the car? 32. and you may actually receive a higher yield than you should mathematically.13.6a. (b) Show how part (a) suggests the desired result. Some banks call 360 days a year.5 percent yearly interest. (d) How much should your monthly payments be if the money is to be completely repaid to the bank in N years ? (e) What is the total amount of money paid to the bank ? (f) If you borrow $3000 for a car and pay it back monthly in 4 years at 12.12. we wish to prove (a) If you have not already done so. show that . Show that (if D(t) is negative. At 5 percent interest compounded 4 times a year. this is a withdrawal). (a) If P(t) is the money owed at time t. How much of your first payment goes towards reducing the amount owed ? (c) Solve the equation for P(t + i A/). If an initial deposit is made and $50 is withdrawn every month (with / percent interest compounded every month). (c) Show that (d) By integrating the result of part (c). state and federal banking laws can be tricky.11. in how many years does your money double? 32. you make an additional deposit D(t). However. do exercise 32. ask your bank for the explanation. in which case you benefit from 5 or 6 extra days of interest!] 32.128 Population Dynamics—Mathematical Ecology culation agrees with their advertised yield! [Warning: If they do not agree. In this problem. show that P(0) = P0 and What is A/and «? (b) For example. part of the first payment consists of P0 I/n interest.

Constant Coefficient First-Order Difference Equations When discussing a population at various discrete times. If the initial population is N0. t + Af and t. 32.2 is called a linear difference equation of the first order with constant coefficients.729 Sec. then Use the Taylor series with remainder of In (1 + x) around x = 0 to prove the result of exercise 32. the population at the m + 1st time is determined from the previous population where as derived in Sec. i. prove that if x> 0. .e.13.14. 33. in how many years will the population double? 32. A species has a growth rate (measured over one year) of a percent. 32.2 involves one difference in time. 33 Constant Coefficient First-Order Difference Equations (e) Using the result of part (d). It is called first order since equation 33. An example of a difference equation without constant coefficients (but still linear) is is an example of a nonlinear difference equation. it is convenient to introduce the following notation: Thus the population at the mth time is For a constant growth rate R0.. Equation 33.15.

Thus Nm = am is a solution. Suppose that the growth of a population is described by where R0 < 0. equation 33. EXERCISES 33.2. (d) Sketch the solution if R0At < — 2. The result is called a divergent oscillation. if initially the population is N0.2. .3 is valid. is also a solution. yields Dividing both sides by rm. the arbitrary constant is determined Thus equation 33. (c) Sketch the solution if —2 < R0At < — 1: The result is called a convergent oscillation.1. Substituting this expression into the difference equation 33. (b) Sketch the solution if -1 < R0At < 0. was constructed earlier: A general technique to solve constant coefficient difference equations (as will later be shown for a higher order difference equation) is to try a solution to the difference equation in the form: an unknown number raised to the mth power.130 Population Dynamics—Mathematical Ecology The solution of the linear first order difference equation with constant coefficients. This method is quite similar to the technique of substituting an unknown exponential into a constant coefficient differential equation. The linearity property implies that any multiple of that solution. (a) Determine the population at later times. derived by a general technique for constant coefficient difference equations. determines r. Since at m = 0 (corresponding to t = to) the population is known.

2. Alternatively. this growth rate can depend on time. substitute an exponential (Nm = esm). By this definition. As a first model. while part (b) is? What ecological assumption of the model caused parts (c) and (d) to yield unreasonable results ? Consider Nm+i = ttNm. For this to be meaningful. However. the rate of change of the population. Instead of substituting Nm — rm.3.131 Sec. Consider a species of animal which only breeds during the spring. Determine the female population as a function of time. If this growth rate is a constant R0. Show that the result is the same. This approximation is most reasonable for large populations. More likely of interest is the instantaneous growth rate (which we will now refer to as the growth rate). Suppose that all adults die before the next breeding season. dN/dt. 34. 34 Exponential Growth 33. It is calculated over a time interval of length Af. the population must be approximated as a continuous function of time. 33. the growth rate also depends on the measuring time interval. we again assume the growth rate is a constant. which is assumed to be differentiable. then the population growth is described by the solution to the first order linear differential equation with constant coefficients . The growth rate is the rate of change in the population per individual. (e) Why are parts (c) and (d) not reasonable ecological growth models. Exponential Growth The definition of the growth rate is In general. equals the growth rate R times the population N. assume that every female produces (on the average) R female offspring which survive to breed in the next year.

A^/i) = 2N(t0). (It is often convenient to let the initial time t0 = 0. 34-1 for 7?0 > 0. .132 Population Dynamics—Mathematical Ecology which satisfies the initial condition The solution exhibits exponential behavior as sketched in Fig. The length of time. Hence. 34-2. Of interest is the time necessary for a population to double if the growth rate is a positive constant. In particular. the time it takes to double does not depend on the initial population. a population decays exponentially if its growth rate is a negative constant as shown in Fig. tt — t0. such that the population doubles. Figure 34-2 Exponential decay. A population grows exponentially if the growth rate is a positive constant. Similarly. is obtained from the expression A^ cancels.) Figure 34-1 Exponential growth.

the population will have grown from N0 to N0eR<>. then from equation 34.) . If a population grows continually at the rate of 2 percent a year (R0 = .69315. the population doubles in approximately 35 years. then after one year an original population of 1. which we have called the doubling time. An accurate rule of thumb (very useful as described in the exercises for savings bank interest rates.3.3 rather than 1. then the number of years to double is approximately 70/R0 years. then in how long will the population double? The required time is Thus. not for an exam. If a population grows continually at the instantaneous rate of 2 percent a year (R0 — . the exponential growth equation 34. In one year.020. becomes *This is one of the more practical formulas we offer in this text.000 grows to 1.* How many years would it take the population to quadruple ? R0 is the instantaneous growth rate per year.000.02). (Memorize it.4 In terms of this parameter. 34 Exponential Growth where.000. and so on) is to note that if R0 is the instantaneous rate of growth per year measured as a percentage. your everyday experiences. from a table of natural logarithms. If td is the mean generation time.02).133 Sec. since Biologists frequently speak of the mean generation time. that is the time necessary for a population to reproduce itself. This result can be applied to the following problem. The measured growth rate over that one year is Since the Taylor series of eR° is there is a small difference between the instantaneous growth rate and the resulting growth rate measured over one year only ifR0 is small.020. inflation rates. but fc. In 2 ^ .201.

the population is also known JV(/.RQAt). Alternatively. using properties of logarithms and exponentials. Let us use this information to determine R0: Although we can directly solve for R0. Thus. We verify this using L'HopitaFs rule. In a problem. Suppose at a later time. An equivalent expression for the discrete case is For both models.14. we might wish to assume a constant growth rate R0. this can be shown using the Taylor series of In (1 -f.) = TV. while for population growth (with growth rate R0) occurring over each discrete Af time interval. We now compare the population growths predicted by the discrete and continuous models. In this latter form growth is measured in intervals of the doubling time. The discrete growth process causes a slower population growth as expected since as is shown in exercises 32.. the population grows exponentially. For a continuous growth model (with growth rate . but not know what value to take for it. The exponential coefficient for a population increasing every A/ time is ln(l + R0At)/At as compared to R0 for continual growth. Another condition is necessary to determine the growth rate. becomes or finally an equivalent formula which is often easy to evaluate. .134 Population Dynamics—Mathematical Ecology which.13 or 32./?„). These two models should give the same result in the limit as A? —» 0. the population would be where N0 is the known initial population at t = 0. t = tt.

.5. How many bacteria would you predict there are four hours after the experiment started ? 34... The growth rate of a certain strain of bacteria is unknown. f(t). A certain bacteria is observed to double in number in 8 hours. determine Nm. 34. 34. Show that the population grows or decays with an exponential growth coefficient.2. show that the solution of the discrete growth model is analogous to integration. N2. In this problem we study the effect of any time-dependent migration. Consider both the resulting discrete and continuous growth models: (a) If the growth rate is zero (R0 = 0). then the problem may be over determined. and an hour later 2000. If additional data is known. By explicit calculation of NI. it was estimated that there were about 1500 bacteria. A population of bacteria is initially N0 and grows at a constant rate R0. N3. if N0 is known. (b) If the growth rate is nonzero. Determine the population of bacteria for all time.1. 34 Exponential Growth a simpler approach is to note eR> = (Ni/Noy/tl. . equal to the average of the time-dependent growth rate. where N(t) oc eRs(t)t). as this is frequently the case. but assumed to be constant. Suppose r hours later the bacteria is put into a different culture such that it now grows at the constant rate RI . we might want to know the exponential curve that best fits the data.3. However. but depends in a known way on the temperature of its environment. (/?£(f). This can be done using the method of least squares as is discussed in the exercises.135 Sec. EXERCISES 34.. If the temperature is known as a function of time. Suppose the growth rate of a certain species is not constant. and hence This gives the expression for the population at all times if it is N0 initially and NI at time / ( (assuming a constant growth rate). derive an expression for the future population (which is initially N0).4. What is its growth rate? 34. When an experiment started. show that where a = 1 + /?0Af.

solve equation (iv).N.11.136 Population Dynamics—Mathematical Ecology (c) For the differential equation (ii). At that rate. If discrete population data is known (not necessarily measured at equal time intervals). One year later the cost had increased to 370 each. Nd(tm~). If it continued at that rate.6. 34. estimate the difference between the measured growth rate over one year and the instantaneous growth rate per year.4 percent. Show that your answer agrees with part (b). and show how your answer is analogous to the results of part (c). 34. If this rate of increase continued.000 to $11. Another species grows at 3 percent per year when measured every year. (Gross National Product) of a certain country increased by 6.7. 34. then the mean-square deviation between the data and a theoretical curve. Suppose that a species is described by equation 32. and thus determine Nm. (a) Assume that the initial population is known with complete certainty. analogous to the integrating factor method used in part (c). what is the instantaneous rate of increase of the cost of living in that year? The parameters of a theoretical population growth curve are often estimated making the best fit of this curve to some data. (d) In this part of the problem. For small growth rates.P.13.12. Suppose the second species has four growth periods a year. By dividing equation (iii) by am+1. How long does it take the population to double? Suppose that one species has an instantaneous growth rate of a percent per year while another species grows in discrete units of time at the annual rate of /? percent per year. obtain an equation for . show that 34. so that we insist that the theoretical population curve initially agree exactly. If the cost of living rose from $10. 34. where Qm = NJa1". 34.9. approximately when would soda be 500 each.8. Consider a species which is modeled as growing at an instantaneous rate of 3 percent per year. Compare the time it takes both species to double. calculate the solution using the integrating factor e~R>t.14. to double? One year food prices increased at a yearly rate of 15 percent.4 with R0 > 0. is the sum of the squared differences. By minimizing the above mean-square deviation. The "best" fit is often defined as those values which minimize the above mean-square deviation. Assume the theoretical curve exhibits exponential growth. we wish to develop a technique to solve difference equation (i). 34.N. in approximately how many years would food prices double. 34. The G. What relationship exists between a and ft if both species (starting with the same number) have the same number 5 years later? The cost of a large bottle of soda was 320 each. approximately how many years would it take the G.P. N(t). Using the result of part (a). 34.10.000 in one year (a 10 percent net increase).

16. Using the ideas presented in exercise 34. 34 Exponential Growth the best estimate of the growth rate.137 Sec.14b and c. determine the best estimate of the growth rate using this criteria. The growth rate of a certain bacteria is unknown. if higher order terms are neglected : (a) Show that Replacing the derivative by this difference is known as Euler's method to numerically solve ordinary differential equations.14b and c. Predict the population at / = 4.15. then Thus. 34. we will show how to calculate a difference equation which corresponds to it. In this way the mean-square deviation is Show that this method now is the least squares fit of a straight line to data.e:. for example. (b) One way to bypass the difficulty in part (a) is to fit the natural logarithm of the data to the natural logarithm of the theoretical curve. If Wo is known (N0 = Nd(t0}). Suppose an experimentally determined growth curve Nd(t) was known. minimize the mean-square deviation with respect to both N0 and R0). In this section it was shown that the difference equation of discrete growth becomes a differential equation in the limit as A/ —* 0. Suppose that the following data were obtained: time number 0 100 1 120 2 160 Using the ideas of exercise 34. estimate the best fitted exponential growth curve. . (c) Redo part (b) assuming that a best estimate of the initial population is also desired (i. determine a best estimate of the initial population and the exponential growth rate of the population.17. This idea is useful for obtaining numerical solutions on the computer of differential equations. If a Taylor series is used. 34. (b) Show that the resulting difference equation (called the discretization of the differential equation) is itself a discrete model of population growth. Show that this is a transcendental equation. 34. If a differential equation is known.

Instead of the total population. it is necessary to discuss the birth and death rates for different age groups. Furthermore let bm be the birth rate of the population m years old and dm its death rate. are there thus any restrictions on the discretization time A? for this problem? 35. the first neglected term in the Taylor series (A/)2 d2N/dt2/2 must be much less than At(dN/dt). We formulate this problem using the discrete population changes as measured every year. for this to be a reasonable procedure. Discrete One-Species Models with an Age Distribution For accurate predictions of the future population of species with variable age distributions.e. where m terminates at N. the oldest age at which there is an appreciable population..138 Population Dynamics—Mathematical Ecology (c) At the very least. we must know the population in each age group. Using the differential equation. Let N0(t) = number of individuals less than one year old Ni(0 = number of individuals one year old N2(t) = number of individuals two years old Nm(t) = number of individuals m years old. A? = 1 year (for many species this measuring interval is too long). i. The difference equations describing the populations (assuming there is no migration) are .

139 Sec. Consider the «-dimensional population vector (n = N + 1) Then equations 35.0 states that the number of individuals less than one year old at the time t + At (i. one year after the time /.2-35. it is the average number of births between / and t -f At who survive until t -f. N25(t).At. 35 Discrete One-Species Models with an Age Distribution Equation 35. b2s.0-35. times the birth rate for 25 year olds.N are derived using the same principles as equation 35. since one minus the death rate (1 — d0) is the survivorship rate.. For large N such problems are ideally suited for computer calculations. For example. The birth rate as used in this section is a slight modification from that discussed in other sections. Equation 35.2 we explain how to solve this linear system of difference equations using the matrix notation if the birth and death rates do not change with time. Knowing all the birth and death rates and the initial population in every age group is sufficient to predict the future population. . Instead of the average number of births per individual between t and t + At. Equations 35. since At = 1) equals the sum (over all age groups) of the individuals born during the preceeding year (who survived until the time / + A/). Matrix methods were introduced by Leslie in 1945 to simplify these types of calculations.1 states that the number of one year old individuals at time / + A? (now) equals the number whose age is less than one year at time t (one year ago) times the survivorship rate for those less than one year old.N become where the n x n matrix A is given by In exercise 35.e.1. the number of individuals born to 25 year olds during the year is the number of 25 year olds.

We assume that in each population group the ages are evenly distributed within the group (this is not entirely accurate). Consequently. b3 and </.140 Population Dynamics—Mathematical Ecology A simplified model with age distributions is sometimes desired. We assume we know the average birth and death rates for each age gro^p (b^bz.1. EXERCISES 35. Instead of separating the population into each age division. Some of these advance age groups. As a very rough model of human population growth suppose we divide the population into three age groups (group 1 aged 0-14. in this example as an additional approximation b^ and b3 can be set equal to zero. The number born is blNl(t) + b2N2(t) + b3N3(i). These types of models of age-dependent population growth thus enable demographers to predict the future need of resources ranging from elementary schools to old-age facilities. Thus since the first group has 15 different years. (This problem requires a knowledge of elementary linear algebra. then . d3). 35. We could calculate how much faster a population predominantly of child-bearing age would double. Since both bv and b3 are much less than b2. Can we predict the future population if we know the present age distribution? The number in group 1 consists of the number born (and surviving a year) plus the number who have survived from group 1 but not "advanced" into group 2. the total number from group 1 who survived is (1 — d^N^i). approximately |f remain after one year (At = 1). Besides predicting more accurately the growth of the total population. Using these types of models we can compare the growths of two similar populations which differ only in their age distribution.. group 2 of child-bearing age 15-39. If A is a constant matrix. and group 3 aged 40 and older). the population is divided up into larger groups.) If A is a constant matrix. show that N(t + wAf) = \mN(t).2. However. d2. these techniques can be used to estimate the future population in each age group. We estimate the number that did not advance in a very rough way.

then a stable age distribution exists. In fact if any of the eigenvalues are complex. i — 1. The population in each age group grows or decays depending on the eigenvalues rt. an age distribution independent of time. exponentially decays if r\ < 1.. obtained by insisting A — ri has a zero determinant. which we call an eigenvector. (c) We say the populations represented for example by and have the same age distribution. The solution can be obtained by a method similar to that developed in Sec. where Ci is an eigenvector corresponding to the eigenvalue r\. then there will be n eigenvalues (call them rt. 41). Show that where I is the identity matrix. . (b) The general solution of the system of difference equations is an arbitrary linear combination of the solutions obtained in part (a).141 Sec. then the populations can also oscillate (the mathematics necessary to analyze this case is discussed in Sec.. Show that //there exists a real positive eigenvalue (call it r\) which is greater than the absolute value of the real part of all the other eigenvalues. Note that such a stable age distribution is independent of the initial age distribution. Show that a stable age distribution is Ci. where C is a constant vector and where r will be determined. Try a solution to this system of difference equations in the form Nm = rmC. and approaches a constant if r\ = I . If A is an n x n matrix. (d) A stable age distribution exists if the populations approach.e. Corresponding to each eigenvalue r{. C ^ 0). Show that in general the age distribution derived in part (b) changes with time. 33. show that the total population grows like e".. (a) Use the notation Nm = 7V(mA/)> and hence Nm+. Thus show that the population exponentially grows if ri > 1. ri).. we know that for nontrivial solutions (i. as time increases. . r can be only certain values called eigenvalues of the matrix A. We assume (for mathematical simplicity) that the eigenvalues of A are all distinct. From linear algebra. there is a vector Ct satisfying (A — rtlL)Ct — 0. it only depends on the birth and death rates of each age grouping! (e) If a stable age distribution exists. = \Nm. Thus show that since / = wAf and At = 1. where a = In r\. 35 Discrete One-Species Models with an Age Distribution is a linear system of constant coefficient first order difference equations.

010 </3 = .2a-b. N{ = 100. How long does it take each total population to double once? Reconsider exercise 35.. N2 = 100. Assume new medical advances cut each death .04 b3 = 0 di = . show that the eigenvalues r of the matrix A satisfy: 35.015 Using a computer.142 Population Dynamics—Mathematical Ecology 35.b2=Q. qualitatively and quantitatively compare the growth of the following three initial populations (measured in thousands).4.005 d^ = . 35. For a species that lives at most four years (i.5. Consider a species in which both no individuals live to three years old and only one-year olds reproduce.5. What is the A matrix? (c) Determine (a computer is not necessary) the time development of a population initially consisting of N0 = 100.d2 = l satisfy both conditions. Explicitly calculate the following three cases of birth and death rates: 35. (b) Let bi = b.6. Case ( 1 ) ( 2 ) ( 3 ) Ni N2 N3 Total 1000 1000 1000 200 400 200 400 300 500 400 300 300 35.3. (d) Compare your results in part (c) to that predicted by the theoretical considerations of exercise 35. (a) Show that b0 = Q. Consider the growth of a human population split into the three age groups discussed in Sec. only four age groups).e. Assume: bi = 0 b^ = .

we assume the probability of one birth is proportional to Af. Are there any significant changes in the age distribution ? 35. for every yearly age group. in this section we formulate a simple model of a stochastic (probabilistic) process. 36. and predict the population for the next 25 years ignoring migration.11. Are there any significant changes in the age distribution ? 35. but at the same time people's attitudes change reducing the birth rate by 25 percent.7 without ignoring migration. In many circumstances this model may need to be replaced by other stochastic growth models (usually more difficult to analyze). Collect human population data in the United States. Instead.7 to exercise 35. In a sufficiently small time interval Af. Suppose an "x" represents the hatching of a baby chick from one hen as in Fig. This is a reasonable approximation for bacteria and other species which grow by cell division. 35. 36 Stochastic Birth Processes rate in half. 35. In order to facilitate the analysis. Stochastic Birth Processes The previous approximations considered birth and death processes to be deterministic. Redo exercise 35. let it be Figure 36-1 Random births. pick more convenient age groupings to do a computer calculation. Redo exercise 35.8.8. including the present population and the birth and death rates.8 without ignoring migration. Suppose we cannot say with certainty that there will be a population increase of exactly /?A/ percent in time At (later in the analysis Af will be infinitesimally small). 36-1. (b) How long does it take each to double ? (c) Compare the results of this problem to exercise 35.7. predict the population for the next 25 years ignoring migration. (a) Analyze the three cases of that exercise. . Instead we assume the birth process is random.9. for every yearly age group.10. 35. including the present population and the birth and death rates.5. we will assume that there are no deaths or that the deaths have a negligible effect as compared to births. Collect human population data in the United States. Do not predict the change in number of each age group. A birth associated with an individual might randomly occur at any time with equal probability.143 Sec. Using a computer. In contrast to that. Compare exercise 35.

. but A is also the birth rate (if growth occurs in a deterministic manner). The probability of one compound event is the product of the individual probabilities (if the events are independent). or the population was N at time t and no births occur during A?. Thus where erN_l is the probability that exactly one birth occurs among the N — 1 individuals. In At time. The probability of it not giving birth in time Af is thus 1 — AAf. Thus not only is X the probability of a birth per unit time. In this case either a birth occurs or it does not. we estimate AA? be dividing the total number of births in Af time by the total population (if the time A? is small enough such that the number of births is a small percentage of the total population). We cannot calculate the exact population at a given time. This shows that the birth rate (AN/(N0At). then the probability of no births among N *This can occur in two different ways.e.144 Population Dynamics—Mathematical Ecology The probability of two or more births in a time interval A/ is considered to be negligible if Af is small enough. We have ignored the possibility of two or more births from different individuals (in exercise 36. Consequently. For example. i. since there are no deaths in this model and hence the birth rate equals the growth rate) is A. it is shown that this assumption is a good approximation if A? is small enough).. In the exercises. since a fundamental law of probability states that the sum of all the probabilities must be one. (We are assuming there are no multiple births). it will be shown that A is the expected growth rate for this stochastic birth process. if the probability of one birth from one hen is AAf. A = -^ per hour. and where VN is the probability that no births occur among the N individuals. then we "expect" if there are a large number of hens TV. If the probability of one individual not giving birth is 1 — AAf. Let PN(t) be the probability that at time t the population is N. and then formulate an equation describing it..* Let us calculate these probabilities. We can only discuss probabilities. if A is unknown. then the birth rate is estimated as ^ per hour and also the probability of a birth is estimated as -j\j per hour. it is the sum of the probabilities of the two compound events. then there would be JV0AAf births. . At time t + A?. there are no other possibilities. How is it possible to have N of the species at time t + Af? A population N can come about in two ways: either the population was N — 1 at the previous time and a birth occurs in time interval A/. what is the probability of the population being Nl We will discuss the answer first. if in one hour 20 hatchings occur from a population of 600.2.

from equation 36. The probability of this is Thus the probability of exactly one birth among m individuals is For N — 1 individuals . The probability of exactly one birth among m individuals is calculated in the following way. 36 Stochastic Birth Processes independent individuals is (1 — hAf)1*. For this last expression to be valid At must be extremely small. Thus. If At is small enough so that the probability of two or more births even from N individuals is negligible. Then (from the binomial expansion) In a similar manner. that is The probability of at least one birth occuring among N individuals is then (one minus the probability of no births). Using a Taylor expansion of the left hand side of this equation. Also the probability of no births is approximately 1 minus the probability of exactly one birth (since in the limit of a small time interval two or more births are virtually impossible). then the probability of exactly one birth is approximately the same as the probability of at least one birth. One way is for an individual (numbered) 1 to have a birth and the other m — 1 individuals not to. That is.1 This equation becomes more accurate as At becomes smaller. yields *We can be more precise in this calculation. Let us consider these probabilities if At is that small. the probability of exactly one birth among m individuals is approximately m times the probability of an individual giving birth. Thus. for At small.145 Sec. Exactly one birth can occur in m different equally likely ways.

For example. since there are births but no deaths! PNt(i) is always a . we are interested only in populations greater than or equal to the initial population N0. a system of ordinary differential equations result: This equation describes how the probability of a population of N changes in time due to a random birth process. With these initial conditions we will show that the system of differential equations can be successively solved. the probability of NQ individuals is given by But PNt-i(t) — 0 since it is impossible for the population to be less than N0. the procedure to solve the system of differential equations is clearer.146 Population Dynamics—Mathematical Ecology After canceling PN(t). 36-2. dividing by At and taking the limit as Af —» 0. Upon solving the resulting constant coefficient differential equation. The probability of the population being N0 decreases in time. initial conditions are necessary. To solve this system. The problem we will solve is one in which the initial population (t = 0) is known with certainty. the likelihood of the population remaining the same diminishes. These are the initial probabilities. In the form. As time goes on. being some value N0. First determine /V-i(0» and then use it to determine PN(t). Since there are no deaths. Thus which is sketched in Fig. In that case all the initial probabilities are also known. where the initial condition implies PN<>(0) = 1.

For example. The probability of A^0 + 1 individuals initially increases (from zero). where N = N0 + 1 and where PNt(t) is eliminated using equation 36. probability function. You should be able to explain. why this is so. The probability of the population being N0 + 1 is determined from equation 36. namely it never is greater than 1 and is always nonnegative.02). /».4: This is a nonhomogeneous constant coefficient first order differential equation. the probability that the population will remain 10 one year later is PIO(!) = e~'z ^ 82 percent. It is unlikely that a population of 10 will change in one year since the likelihood of a birth is only 2 per 100. A = . At what time is it most likely that Figure 36-3 Probability of population increasing by one. then the probability of it remaining 1000 is as we expect extremely small..ooo(l) = e~2° ^ 2 X 1°~9 = 2 X 10~7 percent. if the probability of a birth is 2 percent in a year (i. if the initial population had been 1000. 36-3. based on the birth process. yielding The homogeneous solution is multiplied by the arbitrary constant ANt+lt determined from the initial condition that PNt+i(0) = 0. then for an initial population of 10.Figure 36-2 Probability of population remaining the same. 147 . However. Thus.3. sketched in Fig.e. but eventually diminishes to zero. which can be solved by the method of undetermined coefficients.

As a review. In order to determine the probabilities for all cases we could continue this type of calculation. we outline the concepts behind a mathematical proof by induction.2 with N = N0 + 2. which is good practice in obtaining particular solutions by the method of undetermined coefficients). If a population grew in a deterministic manner with an instantaneous growth rate of A.5.(/). However.148 Population Dynamics—Mathematical Ecology there are N0 + 1 individuals? [Answer: The time at which N0 -f 1 individuals is most likely occurs at the maximum of/ > Afo+ . Let us prove it using induction. The initial condition is again PNt+2(Q) = 0 and thus a few additional algebraic manipulations show that We have determined the probabilities of the first three possible populations. The differential equation 36.+2(t). after using equation 36. Thus the most likely time the population will be A^0 + 1 is the same time that a population growing via a deterministic model would exactly equal N0 + I. becomes Solving this we obtain (with a little bit of algebra.] Let us continue the calculation to determine PN. "a little bit" of insight* suggests that the solution for ally might just be (for7 > 1) At this point this is only a hypothesis. then at this time an original population of N0 would have become N0 + 1 since N0e** — N0(N0 + 1)/JV0 = NO + 1. Thus After some algebra Note that eif = (N0 -f 1)/W0. *A considerable amount of insight is probably required! .

from the differential equation 36. since . why have we proved that the statement is valid for all finite values ofy? ' Let us prove our proposed formula. (Sometimes it is advantageous to prove the statement for the first few values of/) 2. Using the formula for PNt+Jt(t).If we know all of these probabilities.3 with N = N0 -\-j0 -f 1.149 Sec. where the integral is easily calculated. and (3) have been followed. equation 36. 3. e*Wt+Jt+l}t. yields Then by integrating. (2). We have already verified this is valid for j =1. let us determine the next one. Using the assumption prove it holds for the next value. PN. which is valid by the induction assumption.7. by induction. we have To solve this differential equation by undetermined coefficients is difficult (Why?). If steps (1). 36 Stochastic Birth Processes We would like to prove the validity of a statement for all integers j. A proof by induction involves three steps: 1. Explicitly prove the statement is true for the first value of j. Then assume it holds for ally (less than or equal to any value y0).+j. Let us assume it is valid for all j <7o.+i(t).2. usually j = 0 ory = 1. Instead. jQ + 1 (using ordinary proof techniques). Multiplying both sides of the above equation by this factor. note that a simple integrating factor exists.

further analysis of this stochastic birth model is undertaken. This ends the discussion of stochastic models in this text.7 is the probability distribution for a population initially TV.. Show that equation 36. where the constant has been chosen such that PNo+j.+ i(ty = 0.1. EXERCISES 36.7 has been derived for N0 +y'0 + 1 assuming it is valid for N0 + y'0. Thus which is the proposed formula for N0 +j0 + 1. fluctuating in time over different values. In the exercises.. We could also assume the growth rate R. Formulate a stochastic model for population growth including births and deaths (i. (a) Based on probabilistic ideas.2. Let us just mention another way without pursuing its consequences. itself. Thus by induction equation 36. Those interested in additional knowledge in this area can consult the books referred to in Sec. 36. Equation 36. perhaps due to uncontrollable changes in the environment.750 Population Dynamics—Mathematical Ecology and (which can be verified by differentiating or by making a change of variables).2 still remains valid. what system of differential equations determines /VO?) What mathematical difficulty is presented by this system? 36. Thus. We have assumed that each individual randomly gives birth with certain known probabilities. Additional probability concepts imply that the deterministic model is a better and better approximation to the stochastic model as the size of the initial population increases! This model of stochastic births shows one manner in which uncertainty can be introduced in a mathematical model. Modify equation 36. Suggestions are made relating this model to the deterministic model of exponential growth.3. why should . is a random variable. 55.e.1 to take into account the possibility of exactly two births occurring among N — 2 individuals in time Af. with births occurring randomly with probability Xkt (for a short time At).

7. Assume A^ = 10. Consider equation 36. Estimate A. 25)]. The expected population at time t. and explain its significance. verify that part (a) is valid. If A = .4 by an alternative method. compare this to the derivative of the binomial expansion (Sec. (a) If you have not already done so. 25)]. experiments have indicated that the .000. The growth rate cannot remain constant. if it is observed that a total of 4500 births occur in 20 days.151 Sec. [Hint: Explicitly calculate the first few terms of the summation. In this problem we will derive the results of exercise 36. 36. E(t). Density-Dependent Growth The fundamental prediction of the constant growth rate population model. A more complex population growth model is needed.4. compare this to the binomial expansion (Sec. (b) Using the known expression for PNi+j(t\ equation 36.4b. do exercise 36. [Hint: See exercise 36. [Hint: Explicitly calculate the first few terms of the summation. what is the probability that N = 12 one year later? What is the probability that N = 12 two years later? 37. is that the population exponentially grows without limits (if JK0 > 0).4.6. What might prevent a population from growing without a bound? Essentially we suspect that once a population grows sufficiently large it will begin to interact in a different way with its environment or with other species.7. 37 Density-Dependent Growth 36. (c) Explain the significance of part (b). (b) Using the differential equation for PNo+J(t\ show that (c) What is £(0)? (d) Derive the result of exercise 36.5. Although this model may accurately reflect experiments in the initial stages.7. we realize that no population will grow exponentially indefinitely. Laboratory experiments have shown that the lack of food (nutrients) to sustain an indefinitely large population can limit the population growth.4a.] 36. (a) Can you explain why this is valid using probabilistic ideas ? (b) Show that E(t) — NQeif. is obtained from the formula 36. Even if the food supply is sufficiently increased.09 and initially N = 10.

crowding may have the same effect as limiting the food supply. but might depend on the population: What mathematical properties might the function R(N) have? We must remember that we have already assumed that the population is large enough so that we may model TV(/) as a continuous function of time. In some manner. as TV diminishes R(N) should approach the growth rate without environmental influences. Let us attempt to mathematically model this process. then it may be necessary to introduce a more complex mathematical model than the ones developed in this text. If we assume that the growth rate is continuous. In general.152 Population Dynamics—Mathematical Ecology growth rate still diminishes as the population density* increases. but at a smaller rate due to the limitation on growth caused by the increased population density. or both. 37-1: Figure 37-1 Possible density-dependent growth rates. Thus. As the population increases. the death rate to increase. . the birth rate equals the death rate and the resulting growth rate is zero. Thus ./?(TV) decreases as TV increases. Thus we are not particularly interested in R(N) for TV extremely small. still being investigated by researchers. At some population. the increase in density causes the birth rate to decrease. For a much larger population. we still expect it to grow. growth occurs with only slight limitations from the species' total environment. then we know there is a population at which the growth rate is zero as sketched in Fig. For moderate size populations. experiments show the growth rate to be negative (more deaths than births). the growth rate (\/N)dN/dt may not be constant. Space can be considered necessary to sustain life for certain species. *Population density is proportional to the total population since the region is assumed fixed and since the population is also assumed uniformly distributed throughout the region. If instead the population density significantly varies in the region. fl(/V).

37-2. We cannot expect this model to always make accurate predictions if the population ever gets sufficiently small. However. a is the growth rate without environmental influences. This again yields the logistic equation . note that we have not attempted as yet to give a specific model of the growth rate for extremely small populations. Note that a and b are positive constants. Perhaps the growth rate is proportional to the difference between the available food and the food necessary for a subsistence level of food consumption. 37 Density-Dependent Growth In particular. Before solving this equation let us indicate a more specific model from which it may arise. Under these assumptions. Fa. The subsistence level of food consumption can be assumed to be proportional to the population. Fc. Suppose that the available food per year Fa is fixed. If growth was limited by the supply of food. The simplest function with this property is the straight line. for simplicity we now model the growth rate for very small populations in the same manner (solid curve). Figure 37-2 Logistic growth rate.153 Sec. then another variable can be introduced equal to some measure of the yearly available food. This model was first investigated by Verhulst in the late 1830s and later "rediscovered" by Pearl and Reed in the 1920s. sketched in Fig. and b represents the effect of increased population density. yielding the nonlinear first order differential equation known as the logistic equation.

154 Population Dynamics—Mathematical Ecology The population at which the. fc. However. the number of births would exactly offset the number of deaths. What equation describes this situation ? What properties do we suspect are valid for this functional relationship ? Is there an equilibrium population ? F.1.G. the so-called carrying capacity of the environment. (zero population growth).P. Smith suggested a different simple model of the population growth of a species limited by the food supply based on experiments on a type of water bug. then would the population this time increase towards the equilibrium population? EXERCISES 37. As in the logistic model. if there were more than the equilibrium number. Using the logistic model. That is. Consequently a simple model would be with y > 0. the major interest is in the case in which N = a/b. Let us modify our derivation of the logistic equation. fa.2. E. What differential equation describes this model ? What are the equilibrium populations ? . A question we will answer in the next section is whether this equilibrium population is stable or unstable. However. Again assume fc = flN. This is the largest population which the environment can sustain without loss. This theory predicts that the population N = a/b would correspond to Z. if there were initially less than this "crowded" population. then would the population eventually decrease and approach this equilibrium figure? Also. the equilibrium populations are Zero population is certainly an equilibrium population. That is. and a subsistence level of yearly food consumption. the growth rate is proportional to the difference between the available food fa and the subsistence level of food consumption fc: 37.2. equation 37. Smith instead assumed that more food is necessary for survival during the growing phase of a population. previously fc was assumed proportional to the number of individuals of the species.growth rate is zero is an equilibrium population in the sense that if the population was initially at that value it would stay there. Suppose the growth rate is a function of the difference between the available food.

The logistic equation is a first-order differential equation that does not explicitly depend on time—i. The growth rate is defined by equation 32.1. for each 1000 of the species the birth rate drops by 12 per 1000 per year..4. we will explicitly solve this equation. 38 Phase Plane Solution of the Logistic Equation 37. The solution of firstorder autonomous equations can be understood using a phase plane analysis having certain similarities to the approach which we developed concerning . What nonlinear difference equation models population growth of a species if the basic idea behind the logistic equation is applied to the discrete model ? 38. Which of the following are reasonable models of the spread of a disease among a finite number of people: (1) dN/dt = &N (2) dN/dt = &(NT .bN2. let us determine from the differential equation the qualitative features of the solution.155 Sec. 37.bN2 (3) dN/dt = aN + bN2 (4) dN/dt = aN .5. Phase Plane Solution of the Logistic Equation The logistic equation. What is the expected nonzero equilibrium population? 37. Consider the following models of population growth: (1) dN/dt = -aN + bN2 (2) dNjdt = -aN .NT). 37. and the death rate increases by 50 per 1000 per year. with a > 0 and b > 0. Experimentally. Suppose we are considering the growth as measured over a time A/. where N is the number of infected individuals and NT is the total population. However.e. before doing so. describe possible birth and death mechanisms. describes the environmentally limited growth of a population.3.6. For each case.N) (3) dN/dt = ct(N . In the next section. it is autonomous. Determine the parameters of a logistic equation which models this species. A certain species has an instantaneous growth rate of 27 percent per year when not affected by crowding.

If initially greater than the equilibrium population. 38-1: Figure 38-1 Phase plane of the logistic equation. Not surprisingly. (Hopefully only the right half plane is necessary since TV represents the number of the species and must be non-negative. yields Fig. designating how the solution changes in time. the population decreases. We wish to analyze the solution in the neighborhood of the equilibrium population. If we approximate the phase plane curve in the vicinity of .156 Population Dynamics—Mathematical Ecology vibrating mechanical systems. The population level a/b is sometimes called the saturation level. and for populations more than the equilibrium. but we will show it never reaches the equilibrium population. N = a/b. the population increases. as shown in Fig. the population continually grows. we will be able to determine the qualitative behavior of the solution quite quickly. the population continually diminishes towards the equilibrium population. For populations less than the equilibrium population. Graphing dN/dt as a function of N. N increases if dN/dt > 0 and vice versa. this diagram indicates that the model has the desired qualitative behavior. since for larger populations there are more deaths than births.) Only points on the sketched curve correspond to a possible solution. The equilibrium population is clearly stable. If initially less than the equilibrium population. Again arrows are introduced. This can be further mathematically demonstrated in two equivalent ways: 1. As in that case. 38-2: Figure 38-2 Logistic growth model: approach to equilibrium.

(since a < 0). let eNt is the displacement from equilibrium and must be small. the equilibrium population by a straight line as in Fig. (where 6 is a small parameter). It can be more easily solved than the logistic equation. where NQ is the initial population (close to equilibrium and either less than or greater than the equilibrium population). | eNt \ <C a/b. 38-3 (the first two terms of a Taylor series).757 Sec. then we derive the following first-order linear differential equation with constant coefficients: where a is negative (the negative slope of the curve). For any initial population (near equilibrium) the displacement tends to zero. we neglect the nonlinear term (corresponding to the linearization done geometrically in (1)). but never reaches it in a finite time. Using the perturbation method. N —» a/b. The equilibrium population is N = a/b. Thus . Equivalent to this method. The equilibrium population is thus stablel 2. we use a linear stability analysis as developed in the discussions of nonlinear vibrations. 38 Phase Plane Solution of the Logistic Equation Figure 38-3 Population growth near equilibrium. Explicitly as t —> oo. Substituting this into the logistic equation yields or equivalently Since €Nt is small. yielding the behavior of the population in the neighborhood of the equilibrium.

6.2. the population nevertheless does not ever attain the equilibrium population. again shows that the equilibrium population is stable.1. 38-4: Figure 38-4. 38. but at both 12 and 24 hours after the beginning of the experiment there were only approximately 8 times as many as there were initially.2) 38. Sketch A' as a function of t using the direction . 38. the initial number doubled in approximately 2 hours. However. Explain why the sketches in Fig.e. Using Taylor series methods. show that the nonzero equilibrium population of the logistic equation is stable. these values are the same since a is the slope of the phase plane curve at N = a/b. 38. Consider the logistic equation. The exponential decay constant is a.4. 38-2 do not have inflection points. 38.3.158 Population Dynamics—Mathematical Ecology The solution of this differential equation. Which (if any) are stable? Show how the eventual population of the species depends on its initial population. Model this phenomena using a logistic equation (i.5. the decay constant was —a. Suppose the growth rate of a species is as sketched in Fig. Using the geometric argument. estimate the parameters). Consider Smith's model of population growth (see exercise 37. Show that if the initial population is near (but not at) this equilibrium population. In an experiment on fruit flies. Are the equilibrium populations stable? Show that the solution of the logistic equation has an inflection point at a population equal to ^ the saturation level. Show that there are 3 equilibrium populations. EXERCISES 38.

to be solved. Af(0) = 7V0. How does this differ from the first order differential equations analyzed in the sections on mechanical vibrations? 39. was qualitatively analyzed in the previous section. yields . not a desirable form. Eliminating c in that way yields Since both N and N0 must be positive.759 Sec. An explicit solution to the logistic equation can be obtained since the equation is separable: The method of partial fractions will be successful in integrating this equation Since integration yields where the absolute values in the resulting logarithms can be very important! The arbitrary constant c enables the initial value problem. This equation gives t as a function of N. more precise quantitative behavior may at times be desired. Multiplying by a and exponentiating. Explicit Solution of the Logistic Equation Although the logistic equation. 39 Explicit Solution of the Logistic Equation field and the method of isoclines.

One example is sketched in Fig. As we know. that is.2) show how this solution verifies the qualitative results obtained from the phase plane. if the equilibrium population is reached in a finite time. have indicated good quantitative agreement to logistic curves. such that a — bN = 0. then equation 39.1. b. Consider dNjdt = UN2 .bN) remains positive for all time. .160 Population Dynamics—Mathematical Ecology a — bN and a — bN0 have the same sign* and hence. Specific logistic curves depend on the three parameters a.2 shows t = +°°. "Initially a — bNo and a — bN have the same sign. this cannot occur. Specifically. or equivalently This equation can be solved for TV or As an exercise (see exercise 39. 39-1: Figure 39-1 Typical time-dependent logistic growth curve.ftN (with a > 0. on the growth of yeast in a culture and on the growth of paramecium. Laboratory experiments.bN = 0. for examples. (a) How does the growth rate depend on the population ? (b) Sketch the solution in the phase plane. EXERCISES 39. Thus the sign of (a — bN0)/(a . if a . ft > 0). The sign of (a — bNo)l(a — bN) can change only if there is a finite value of t. and N0.

and /0 ? [Hint: Put the above expression over a common denominator. (a) Show that 39. (d) By considering the exact solution.e. (a) How does the growth rate depend on A7? (b) What would you expect happens to the population ? (c) By using the phase plane.] (b) Recall that the hyperbolic functions are defined as follows: Thus show that (c) Sketch tanh x as a function of x. equation 39. What is a. tanh (— x) = —tanh x). show that the population reaches infinity in a finite time. 39. then N —» 0. what might be called a population explosion.3. then TV—> oo.. (ii) Show that tanh x is an odd function (i. Consider the following growth model: 39.2..) The logistic curve.3. (iii) Show that lim tanh x = 1. has been shown to be equation 39. .ft = 0. Show that thi? exact solution has the following properties:j (1) It is defined for all / ^ 0. (At what time does N—* oo?) (ii) If N0 < fry. N —> a/b for all initial conditions except A^ = 0.3. Hints: (i) Show that tanh 0 = 0. 39 Explicit Solution of the Logistic Equation (c) Obtain the exact solution.4. is sometimes referred to as an S-curve for the reasons to be described. show that the population tends towards infinity.1. and thus what is lim tanh x? (d) Now sketch the logistic curve.161 Sec. equation 39. (Note: It takes an infinite amount of time to reach the equilibrium population. (e) Show that a + ft = a/b and <X . Show that it might be called "Sshaped". Multiply numerator and denominator by e~(a/2)('~'o). (d) Show how both parts (b) and (c) illustrate the following behavior: (i) If ATO > /?/<x. (2) As t —» oo. (iii) What happens if A^ = /?/<x? The general solution of the logistic equation. ft.

It is impossible for a population to vary in an oscillatory . Growth Models with Time Delays Although some experiments on environmentally limited population growth have shown good qualitative and quantitative agreement with the logistic growth model. The preceding types of experiments motivate us to develop other models of population growth of one species. some of these experiments show the types of behavior sketched in Fig. 38 that for any population growth model of the above form. Consider a model in which the rate of change of the population depends only on the present size of the population in an otherwise arbitrary way. we can verify using the phase plane analysis of Sec.162 Population Dynamics—Mathematical Ecology 4O. especially since in many cases it is so complicated as to disguise the frequently simple qualitative behavior of the solution. However. 40-la the population oscillates around its carrying capacity in a manner similar to the decaying oscillation of a spring-mass system with friction. a stable equilibrium population will be approached monotonically. Of secondary interest is the explicit solution. 40-1: Figure 40-1 Oscillations of experimentally observed populations. other experiments have shown that the carrying capacity of the environment is not always reached monotonically. In Fig. Of most importance for any specific model is knowing the possible equilibrium populations as well as which (if any) are stable. Other experiments show wild oscillation as in Fig. Instead. 40-lb.

the delay time may be . equation 40. and hence would continue for a few years to produce large numbers of newer trees. Instead there is a delay. The total tree population might easily be larger than the environment could sustain. This is a reasonable ecosystem to investigate since we expect the population to exponentially grow when the forest is sparse. However. Consider the growth of a small forest composed of one species of trees. where td is the delay time.the time it takes an egg to develop into a . Furthermore. Although the situation described above might not happen under all circumstances. The delayed response of the tree population to its environment could cause the population to go beyond its saturation level. but a past population N(t — td). To mathematically model this phenomena. Perhaps the understanding of one such ecosystem will suggest to us ways to modify the general one-species deterministic continuous growth models. Other types of models will be introduced which may allow oscillations of the observed type. the trees might not immediately feel any noticable improvement. Let us describe one situation in which a population may oscillate about its carrying capacity. finally causing the tree population to tend back towards the carrying capacity. and since we know that due to limited nutrients (sunlight. yielding a few years later. the population could oscillate around its carrying capacity in a manner that we now explain. Before these young trees mature. Even as the number of seeds produced was decreased.) the population will not grow forever. More deaths than births could occur until the tree population diminished below the saturation level. The carrying capacity is not necessarily reached monotonically. water. etc.1. Having sufficient nutrients the trees would produce a large number of seeds. a large number of young trees. Two slightly different models of this situation are: For animals. Only in a few years would the mature trees notice the effects of their earlier overproduction.163 Sec. Again the delay in the trees response to its environment might prevent the population from monotonically approaching its equilibrium. we note that the basic idea is that the population does not respond immediately to its environment. the overcrowded ecosystem would prevent large number of seeds from being produced as well as developing into young trees for a number of years. we have indicated a possible cause of oscillations. 40 Growth Models with Time Delays manner around a stable equilibrium population. Then the death rate of trees might surpass the birth rate. The rate of change of the population is not a function of the present population N(t). the older trees would not feel their presence. Suppose that at a given time the population of trees was slightly less than its carrying capacity.

as discussed previously the constant growth model (using discrete time) is . (if td is sufficiently large). In this way.164 Population Dynamics—Mathematical Ecology fertile adult. or perhaps the delay would destabilize the equilibrium population as illustrated by Fig. Furthermore if the delay is sufficiently large the population might oscillate around its equilibrium population as illustrated by Fig. then the species would grow at a rate determined by the population at a previous time. In general. then population growth is modeled by We will attempt to illustrate how a delay can cause the population to oscillate around an equilibrium population. For example. We will not discuss their solution in this text. If we apply the ideas behind logistic growth to the delay mechanism.2a and 40. For example. the population would continue to grow at the rate for a population not near equilibrium Figure 40-2 Delayed growth. i. If there were a delay mechanism. If we measure population growth over a time interval Af.. 40-2 with a star). At a time when the population has almost reached equilibrium (indicated in Fig. for populations less than equilibrium. However. see exercise 40. 40-Ib.4. 40-1 a. the population grows. Then a linear delay-differential equation. In order to simplify our calculations we will return to the discrete time formulation. A? is sometimes called the discretization time. the population could go beyond its equilibrium value.e.3. but occurs with a delay td. 40. the discrete time over which changes are measured. delay-differential equations are considerably more difficult to solve than ordinary differential equations. N(t). suppose the growth rate is a constant RQ.1 for a simple example of the solution of a delay-differential equation. and vice versa. described for example by equation 40. Suppose a species grows in the logistic manner without a delay.2b are called delay-differential equations. then a general model of discrete population growth without a delay is where the growth rate measured over the time A/ may depend on the population. Equations 40.

We have assumed the change in population only depends on the previous year's population. Here. in which case this situation may be modeled by the discrete logistic equation: These equations automatically take into account a "delay" At.9 involves three levels of time.3. A much oversimplified model for this process is the discrete logistic equation with a time delay. The increase of deer over a year in a given area depends on the vegetation. An example of this will be given. We can directly solve this equation by explicit numerical calculations. the growth rate may decrease as the population increases. The solution to the discrete logistic equation is pursued in exercise 40.165 Sec. The discrete formalism is especially appropriate for perennial plants and animals which breed only during a specific time of the year. The change in population depends on the population At time ago. t + At. Equation 40. For a specific example. the delay in the growth mechanism is also a year (td = At = 1). the time delay td in this model must be an integral multiple of At. A discrete constant growth model with delay time td is while a discrete logistic growth model with delay time td is Since measurements are made every At time. if nutrients are limited. which in turn depends on how much vegetation was left uneaten by the deer during the previous year. In this way. The notation and . Suppose that we employ a discrete formulation of population growth. N(t). 40 Growth Models with Time Delays the growth rate is proportional to the population. consider a deer population. the change in population from one year to the next clearly depends not on the present population but only on the past population. the change does not depend explicitly on the population two years previous. In these cases. and t — At. but in addition growth occurs with a delay of td units of time. It is called a second-order difference equation since the times involved differ in integral units of At by at most two units of At. However. In particular. we would like to consider a slightly more complicated problem. measured every year (At = 1). t.

Using this notation the discrete logistic equation with a delay. but as the population increases the growth rate decreases proportionally such that b = Tfoj (for every 4 individuals.166 Population Dynamics—Mathematical Ecology is perhaps more convenient. that the growth rate without environmental influences is 20 percent per year (a = •£).3 #16 N17 N1S N19 N20 N2l N22 N23 = = = = = = = = 70.10 is appropriate with a delay in the growth process of 1 year (At = 1).7 64.9. knowing only an initial population is not sufficient to be able to use this equation for calculations! Suppose. the appropriate conditions to initiate a calculation are the population at two successive years.1 Ns N9 Nio Nil Nit 39.4 43.6 20.0 #i = 15. equation 40.9 48. the population is also needed one year later. becomes where a = a At and /? = bAt.0 72.7 56.2 67. as an example.0 75.6 N2* N25 N26 N27 N2» N29 N30 N3l = = = = = = = = 79.2 31. the growth rate is reduced 1 percent per year).9 and sketched in Fig. which we assume to be given as fifteen (Nt = 15).3 79.0 35.6 77.10 becomes Calculating the population for the following year.6 79. To solve the difference equation 40.0 N2 N3 N4 N5 N6 N7 = = = = = = = = = = = Nl3 = Ni4 = N15 = 17. .5 23.5 79. Note that since 3 levels of time are involved. For example.1 78.10 with At = 1.0 79.1 27. The initial population is known to be ten (N0 = 10).8 79.2 74.5 78. In summary Then equation 40. yields Additional results have been calculated from the difference equation using a computer No = 10.5 76.8 79. Let us calculate the population changes if equation 40.3 52.8 60.7 79. 40-3.

167 Sec. 37-39). a = % and ft = -^-Q. we might conclude that a different type of mathematical model is necessary. However. This is not surprising as we note that there exists an equilibrium population NE (Nm = NE for all m) if A nonzero equilibrium population is the same equilibrium population as in the case without a delay mechanism (see the logistic equation. Numerical solution of a discrete logistic equation with a In the previous calculation. We have just illustrated a disadvantage of explicit numerical calculations because actually we are not justified in assuming this same type of behavior occurs for all other delayed growth problems described . the population seems to be tending to 80. 40 Growth Models with Time Delays Figure 40-3 delay. At first. In this case the behavior of the discrete logistic equation with a delay is quite analogous to the continuous logistic equation. and consequently there is an equilibrium population of NE = ^ — 80. we introduced the discrete model with a delay in order to show an oscillation around the carrying capacity. Sees. For this example.

However. The linearization can be accom- . In a similar manner one can try to understand how the population may decrease even when it is less than the carrying capacity. Note that the population increases (i.. a//?. In fact. 40-4. to go beyond the equilibrium. Nm+l — Nm > 0) if Nm_! < a/0. It is perhaps possible that Nm might be greater than the carrying capacity but N m _j less than the carrying capacity. equation 40. Thus if the population one interval of time earlier is less than the carrying capacity. we will show that an oscillation can occur.e.10. To more accurately analyze the population satisfying a discrete logistic equation with a delay. However. #m+i . then the population increases. although always tending towards the equilibrium. It is the above ideas that form the basis for oscillations to be possible for solutions to the discrete logistic equation with a delay. Perhaps (as we will show) oscillation occurs in other cases. The population decreases to either less than or greater than the carrying capacity.10 predicts that although the population has grown beyond the carrying capacity.Nm > 0 if Nm > a/£ and Nm_l < a/jff). we do not even know if the equilibrium population will always be stable! Recall without a delay a population satisfying the logistic differential equation will approach the environment's carrying capacity without going beyond it. as illustrated by the numerical example. In the previous paragraph we have described a plausible circumstance.10. Figure 40-4 Discrete logistic equation with a delay: possible growth beyond the environment's carrying capacity. The discrete logistic equation with a delay.168 Population Dynamics—Mathematical Ecology by equation 40. as shown in Fig. We will now try to determine under what conditions this occurs. may sometimes have this property. we must linearize the nonlinear difference equation in the neighborhood of the equilibrium population. Then equation 40. that is the population increases if it is less than the carrying capacity and decreases if it is greater than the carrying capacity. it is still growing (that is. Introducing a discrete time enables the population. The population must decrease at the next interval since Nm+2 — Nm+l = Nm+l(ct — ftNm). it is the delay that allows the population once past equilibrium to continue to grow away from equilibrium (but this latter effect occurs only for the length of the delay).

| eym \ <C a//?. can be neglected (analogous to the linearization of a nonlinear continuous growth differential equation) since the population displacement is much less than the equilibrium. small displacements from an equilibrium population satisfy a linear constant coefficient difference equation.6. Thus. After doing so. in the next section we will analyze linear constant coefficient (second-order) difference equations.169 Sec. will be assumed to be much smaller than the equilibrium population: Substituting equation 40. —fiym-i(eym). Let or equivalently where the displacement from equilibrium. However only a = aAt and not ft = bAt determines the behavior of the population near the equilibrium. Thus as a good approximation As we could have suspected.10. in order to judge the effects of a delay. . 40 Growth Models with Time Delays plished by Taylor series methods as outlined in exercise 40. yields which when simplified becomes The nonlinear term. This analysis will not only show that oscillations are possible. but will also determine the conditions under which the equilibrium population is stable.1 Ib into the nonlinear difference equation. Here we will do the linearization using the perturbation procedure. we will be able to determine the behavior of a population near equilibrium with a delay present in the discrete growth mechanism. equation 40. €Nt(t) = €ym.

40. Determine the population for 0 < / < td. Suppose Determine a nonzero equilibrium population. (a) Explain (both ecologically and mathematically) why in order to solve this equation one needs to know not just the population at one time. . What linear delay-differential equation determines the stability of that equilibrium population ? (a) Briefly explain why is called a discrete logistic model. 40.770 Population Dynamics—Mathematical Ecology EXERCISES 40.3. (b) Suppose that the population N(t) is a constant N0 for — td < t < 0.3. and solve it.tMNW Briefly discuss the differences in the models.td)] (3) dNldt = N(t . (b) Using the discrete variable t = mAt + t0. a linear delay-differential equation.] (e) Explain the differences between the solutions of the continuous and discrete logistic equations. but the population for an interval of time of length td. Show that ym satisfies a linear difference equation.1. (A/ may be called a delay). How small must At be such that solutions of the discrete logistic equation have a behavior similar to solutions of the continuous logistic model ? Leslie proposed the following discrete growth process: 40. Can you determine the population for mtd < / < (m + l)f d ? 40. show that a nonlinear difference equation. 40. (d) Is Nm = a/b a stable equilibrium population ? [Hint: Let A^ = a\b + eym and neglect terms of O(€2). Assume the growth of a population is described by equation 40.td)R[N(t . Consider the following three general models for population growth with a delay: (1) dNldt = N(t .4.td)] (2) dNldt = N(t)R[N(t . the same as for the continuous logistic model.2.5. (c) Show that an equilibrium solution is Nm = a/b. What are the equilibrium populations? Compare this model to those discussed in Sec.

Suppose that 7V m+1 ./?Ar m _i). y0 and ylt are necessary for a unique solution. #„_. An equation of this kind can be put in the form We claim that the solution to this equation is quite analogous to the solution of constant coefficient second-order differential equations. Assume /?(0) > 0 and R(N0) = 0 with N0 > 0.171 Sec. 41 Linear Constant Coefficient Difference Equations 40. a constant coefficient difference equation. while for linear constant coefficient difference equations. For a constant coefficient differential equation. as it did in Sec. We look for solutions to equation 41. the solution usually exists in the form of exponentials. (a) How would you determine all possible equilibrium populations.7. 40 (see equation 40. show that equation 40. just as we did for first-order constant coefficient difference equations (see Sec. (c) If f(Nm. obtain the linear difference equation which determines the stability of an equilibrium population. show that N — N0 is a stable equilibrium solution. 33). the solution almost always is in the form of algebraic powers. Nm. 41. A^? (b) Using Taylor series methods.Nm = f(Nm. 40. for example the first two values. Linear Constant Coefficient Difference Equations The stability of an equilibrium solution of a nonlinear autonomous second order difference equation is determined from its linearization.i) = Nm(et . Again value(s) for r are determined by substitution: Dividing by rm~l yields a quadratic equation for r whose two roots.10).).6. Suppose that dNjdt = NR(N) with R(N) a steadily decreasing function (R'(N) < 0). Two initial conditions.12 follows using the result of part (b).1 in the form. NE. Without using the phase plane. .

However. . This case will only be discussed in exercise 41. pz =£ 4q. Introducing polar coordinates. and thus the polar form of a complex number is often much more convenient. that is p2 — 4q < 0. these two equations can be uniquely solved for ct and cz in terms of y0 and y^ (see exercise 41.1). it is possible to directly calculate powers of a complex number r = x + iy. The general solution should contain two arbitrary constants. this solution must be modified.172 population dynamcis -mathematical ecology we call /-j and r2. Thus we have a complete description of the solution. This result has assumed the two roots are distinct. we see Figure 41 -1 Polar representation of a complex number. just as a similar modification was necessary for constant coefficient differential equations.2. How is rm interpreted (for integer ni) when r is complex ? Since m is an integer. If the roots are equal. ^ r2. r = —p/2 ± i*/4q — p2/2. suppose the roots are complex. 41-1. Finally we arrive at the polar form of a complex number Thus or the latter result known as De Moivre's theorem. The arbitrary constants are uniquely determined from the initial conditions : Since r. For example. It is obtained by taking an arbitrary linear combination of the two solutions. for large integers this becomes unmanageable. The complex number is represented as a vector in Fig. However.

= — 0 2 (as illustrated in Fig. . Earlier in our study of differential equations arising for mechanical vibrations. it follows that the solution of the difference equation is Figure 41-2 Complex conjugate roots. a linear combination of imaginary exponentials was encountered. 41 Linear Constant Coefficient Difference Equations In our case of complex roots. Since | r t | = | r 2 1 and 0.173 Sec. one root is the complex conjugate of the other. There it was shown that an arbitrary linear combination of these functions is equivalent to an arbitrary linear combination of sines and cosines. Thus in the case of complex roots where Consequently. 41-2).

let us analyze the difference equation Solutions exist in the form ym = rm. at m = 0) to be /v/cf + £4. grows if \rl \ > 1 (i.e. Consequently. [Note that we have defined the amplitude of oscillation initially (i... while y0 = c3 from equation 41.e. where A — *Jc\ + cl. that |r. since \rl \ — 2l/m«.e.4) It is a growing oscillation with period m — 8/3. q < 1). | < 1 (i.. Suppose that we wish to solve equation 41. The amplitude of oscillation is 2m/2+/cl + cl. all that is unknown are . q > 1) and decays if |r.174 Population Dynamics—Mathematical Ecology The solution "oscillates" with an approximate period given by The amplitude of oscillation. As an example.5 subject to the initial conditions We already know the basic behavior of the solution. the amplitude has doubled when An alternate expression for the amplitude of oscillation is A2m/m". defined as *Jc\ + c\\rl \m (see Sees. the general solution of this difference equation is (see equation 41. where by substitution the roots r satisfy and thus are We see. | = ^2~ and 0j = 3n/4: Figure 41-3. 41-3. 5 and 12).] If the amplitude of oscillation grows.4. as sketched in Fig. the doubling interval of the amplitude of oscillation is m = 2 (slightly less than the period). let us compute the number of intervals md it takes for the amplitude to double: Thus.

m = f. These are not necessarily positions of the maximum of the oscillation since the "crests" may occur at nonintegral values of m. we could have more easily obtained them directly from the difference equation 41. 41 Linear Constant Coefficient Difference Equations the constants c3 and c4 (now determinable from equation 41. we can improve our estimate of the position of the crests. The theoretical amplitude is A2m/n" where A is an unknown constant and md the unknown doubling interval. it follows that The solution to the initial value problem is Although we may sketch a smooth curve from equation 41.175 Sec. To estimate the doubling time. However. it is a meaningful solution of equation 41.6): Using the trigonometric formulas. Thus from the sketch we can only roughly estimate the period as being equal to m = 3 (6 — 3 = 3). Using the two points m = 3 and 6 at which ym = 10 and 24 respectively. indicate the correct general behavior of the solution. The tabulated values are sketched in Fig. if we were only interested in the first few explicit values of ym. we obtain . not far different from the exact result. By sketching a smoothly growing oscillatory function from the data. we let the geometrically-doubling amplitude of oscillation curve go through our estimated maximum points rather than the actual maximum points (unknown from only the numerical values).7. The sketch also can be used to estimate the period of oscillation and estimate the doubling time of the amplitude of oscillation. Thus yz = — 2(y0 + y^ = —6 and similarly: m 0 2 i 1 2 -6 3 10 4 -8 5 -4 6 24 7 -40 8 32 9 16 10 -96 ym Let us show that these first few computed values.5 only at integral values of AM. 41-4.5. We observe maximums at m = 3 and 6. and thus improve our estimate of the period. when sketched. For example. y2 = 2(2 cos (37T/2) -f 3 sin (37r/2)) = — 6.

4 = 2 3/m ' and thus roughly estimate md as In 2.4 = 3/md In 2 or not far from the exact value of md = 2. Estimation of oscillation properties from data numerically Eliminating A (by dividing the two equations) we determine that 2. In the next section we will apply the results of this section in order to .176 Population Dynamics—Mathematical Ecology Figure 41 -4 obtained.

(c) Since xm = rm.3. take the derivative of (*) with respect to r and show that Thus show that is a second solution of equation 41. If r — x + iy. 41. Simple harmonic motion can be represented (as is discussed in Sec.5. (a) Show that xm = rm satisfies 41. we have shown where x = r cos 0 and y = r sin 0. Consider equation 41. 41.2.1 is given by equation 41. Assume that b is the complex conjugate of a (see exercise 5.777 Sec.5). 5) by the formula where a and b are arbitrary constants. For real r we know that For complex r. Use the polar representation of a to derive equation 5. 41 Linear Constant Coefficient Difference Equations investigate populations which are near the carrying capacity of the discrete logistic model with a delay. The right-hand side of (*) is zero only if r = —p/2. 41.1.1. Thus one solution isym =(-pl2)m.3.4. let us define In r such that rm = em ln T. Thus solve for In r for r complex.1 in the case in which />2 = 4q.. . (b) In order to obtain a second solution. Show that the constants Cj and c2 can be uniquely determined in terms ofjvo and>>. EXERCISES The general solution of the difference equation 41. using part (b) show that the general solution is (d) Verify the second solution by direct substitution.

Substituting. The results of Sec. i. 13). We will discuss what occurs in the discrete delay case. If 0 < a < £. in Sec. This is analogous to the overdamped oscillations of a spring-mass system (see Sec. m —» oo) for any initial conditions.. We show this by analyzing the solution. the population always tended monotonically to the environment's capacity. we obtain the quadratic equation whose roots are Thus the general solution is The equilibrium population is stable if this solution does not grow as / —> oo (i.178 Population Dynamics—Mathematical Ecology 42. ym = Wi(0 —* 0. and less than 1.e. Destabilizing Influence of Delays In order to investigate the behavior of a population near its carrying capacity. 41 imply that the solutions of this constant coefficient linear difference equation can be obtained in the form Although we could quote specific results from that section. are governed by the following linear difference equation: or where eNi(t) = eym is the displacement from the equilibrium and a = a A/.10). then as t —» oo. The equilibrium population is stable if 0 < a < £. if 0 < a < £.e. 0 < AI < r2 < 1. Recall that for the continuous logistic growth model. N = a/b. Populations near the equilibrium population. Although the population tends to the carrying capacity of the environment.. it may go beyond it once. 40 we discussed the discrete logistic equation (equation 40. The initial value problem implies .2.The population displacement vanishes and hence the population tends to the equilibrium population. then both roots are real. 0 < r t < 1 and 0 < r2 < 1. where we denote r2 as the larger root. the equilibrium population was always stable. Consequently. equation 42. let us instead only use the method. positive.

For the example of Sec. if Thus the population goes beyond the carrying capacity if Whether this occurs or not depends on the initial conditions. 40 in which a = £. j0 and y^ are any two initial displacements from equilibrium only when the linearization is valid. N0 = 10 and Nl = 15. The delay enables the population .779 Sec. From the computation in Sec. we see rl — (1 — A/I — 4a)/2 ^ .27. There are two important cases: (1) If ^ < a < 1. depending on \r\ = a1/2. ft = ^3. c2 has the opposite sign from j. 42-1. 40 it is seen that when the linearization is valid .27 y0 < y^ for any two successive populations. Thus we should not expect the population to go beyond the carrying capacity in that case. If a > £. 42 Destabilizing Influence of Delays and hence If y0 and j>i are of the same sign.4) where The solution grows or decays as it oscillates. as shown in Fig. However.. then the solution is a decaying oscillation (called a convergent oscillation). then the two roots are complex conjugates. The population oscillates around the equilibrium population. As we have shown the solution oscillates (see equation 41. Thus the sign of c2 must differ from the sign of yl. the population goes beyond the carrying capacity only if for large times (large m) the population has the opposite sign from its initial sign.

it can increase only until the next interval after which it must decrease (see exercise 42. Note in both cases (1) and (2).Figure 42-1 Decaying oscillation for moderate delays (| < a < 1). Let us discuss what occurs in this case. The special cases a = | and a = 1 are only briefly discussed in the exercise 42. Here.18. if the displacement population is greater than 0. but the population still tends to its equilibrium value if £ < a < 1. since for small populations this model. the species does not become extinct. Figure 42-2 Growing oscillation for large delays (a > 1). as shown in Fig. 42-2. 180 . Why are these cases of no particular importance? If a > 1. the large delay causes an oscillatory growth around the equilibrium population (a divergent oscillation). to go beyond its saturation level. that. the equilibrium population N = a/b is unstable. In the continuous model the equilibrium population is reached quickly. The population goes beyond the equilibrium population by larger and larger amounts. In other words. The population cannot tend to its equilibrium value.16). consider the case in which a > 1. (2) However. On the other hand. the delay tends to destabilize an otherwise stable population. for example.

0 Ar3 = 145. at the 6th time step. Equation 42.0 N! = 20. Note that the population increases beyond the carrying capacity only for one unit of time (the delay).7 AT5 = 24. As this is impossible it is clear that the mathematical model itself needs to be modified. A model which never has this difficulty is briefly discussed in exercise 42.6 (?) as sketched in Fig. but the delay time Af is varied. However. 42-3.1 N6 = -59. If the delay is moderate (£ < aAt < 1). and b. Perhaps the species being modeled should be considered to become extinct at this time.6 #4 = 273.13. the equilibrium population is stable in a manner quite similar to that which occurs for the continuous logistic equation. 42 Destabilizing Influence of Delays predicts growth. the environmental limiting factor. are the same for all populations. consider populations described by the discrete logistic model. Although the solution continues to vary for all time. In summary.5 yields: n0=10. We choose as initial conditions A straightforward computer calculation of equation 42. Suppose that a.5 allows in this case a growth rate less than — 1 measured in time At (or a loss of over 100 percent). the positive growth rate without environmental limitations.181 Sec. To illustrate this. Suppose Thus the theoretically unstable equilibrium population is N = a/b = 100. a negative population is predicted by this model. Thus the population must continually vary for all time due to the delays. let us consider a numerical experiment with this discrete-delay logistic model.0 N2 = 56. with time delay A?. for species . For species with extremely small delay times (aAt < ^). the population may exhibit convergent oscillations around the stable equilibrium population.

What is the approximate rate of growth ? (c) Show from part (b) that as A/ —> 0. the equilibrium population becomes unstable. The difference equation with R0 a constant describes a pure growth process with a delay. The delays destabilize an otherwise stable equilibrium population. 42. + Ro^tN^.1.2. Divergent oscillations occur.182 Population Dynamics—Mathematical Ecology Figure 42-3 delay. the growth rate approaches R0.3.3. Consider the following model of population growth: (a) What are the possible equilibrium populations? (b) Show that small displacements from the nonzero equilibrium popula- . EXERCISES 42. To be more precise. The effect of a discretization alone is discussed in exercise 40. (a) Show that Ni+i = N. we have shown that introducing a discrete measuring time and an equal time delay have destabilized the continuous logistic equation. Numerical solution of a discrete logistic equation with a large with larger delays (aAt > 1). The effect of the delay is probed in exercise 42. (b) Solve this difference equation.

183 Sec. 42. 42 Destabilizing Influence of Delays tion satisfy as do small displacements for the slightly different model in Sec. (c) Compare the model to that described in Sec.4-42. to the discrete logistic equation with a delay discussed in Sec. Show that r = —ae~rt". (b) If tm > 0. .5. Instead. Show that small populations grow in a manner which can be approximated by eat if At is small enough . Consider the following model of population growth with a large delay: Approximate the general solution if JR0 At is small but not zero (0 < RQ At <1). Show that the nonzero equilibrium population of (1) is stable if a At < 1 and unstable if aAt > 2.3. Also show the population "oscillates" around its equilibrium value if 1 < aAt < 2.7. In this problem we wish to illustrate the effect of a delay on the discrete logistic growth model. let us look for special solutions of the form of exponentials. describe the effect of a delay on the discrete logistic equation. (a) Determine nonzero equilibrium populations. 42. (b) Analyze the stability of the equilibrium population. show that the population oscillates around its equilibrium population with decreasing amplitude. What is the approximate "period" of this decaying oscillation ? (d) Show that zero population is an unstable equilibrium population. (c) If At is such that aAt = 1/2. 42. Consider (a) Show that small displacements from equilibrium satisfy the following linear delay-differential equation: 42. We will compare the discrete logistic equation without a delay. 42. Consider the following models: 42. 42. Using this result.6. we will not find the general solution. Show that it is possible for negative populations to develop.

but occurs if ym < 0. 42. (a) What is the expected behavior of the solution ? What is the period and amplitude of oscillation ? (b) If ^0 = 0 and y\ = \.5 be modified to prevent negative populations ? The following equation is a discrete growth model with environmentally limited growth which never predicts a negative population: (a) Show that a negative population can never occur.10 with « = 10.9. give an ecological interpretation of the parameters a and b. Using a computer.1. How might equation 42. how does the condition for destabilization depend on the time delay A/? 42. do some numerical experiments for a At > 1.12. 42. Give an ecological interpretation of this result. (b) Show that this equation has the desired qualitative properties of the logistic equation. then the displacement population can increase only until the next interval after which the displacement population must decrease. Ib. If a is a constant. (c) Compare periods of oscillation in parts (a) and (b). (c) If a < 1. (c) If A/ is very small.11. If 0 < a < £: (a) Show that the population may go beyond the equilibrium (at most once). what is the shortest period of oscillation that can occur? Consider equation 42. If a > 1.Ib. Consider equation 42. Explicitly compute the population as a function of time.15.8.10. 42.3 when a > ^. Describe how Figs. .1 with a = 2. 42-1 and 42-2 are consistent with this result. 42. (c) Show graphically there are two real solutions if the delay is small enough. (d) Compare doubling times of the amplitude of oscillation. how long does it take the amplitude of oscillation to double ? Consider equation 42. 42.16. (b) Give an example of an "initial" population that reaches equilibrium only after going beyond the equilibrium. 10. Reconsider the population growth model described in exercise 42. 42. Reconsider exercise 42.17. (b) Describe the result that exists which is analagous to part (a).184 Population Dynamics—Mathematical Ecology 42.13.13. Consider the oscillations implied by equation 42. (b) Show the amplitude exponentially increases if a > 1 and exponentially decreases if a < 1. explicitly compute ym for m <. 42. 42. but no real solutions for large delays. Consider equation 42. (a) Show that if ym > 0. at what time has the amplitude reached 1/e of its initial amplitude of oscillation ? (d) If a > 1. (d) Analyze the solutions of this difference equation near the nonzero equilibrium population. (a) What might be called the amplitude of oscillation.14.

In an attempt to understand large ecosystems. because of the war no observations were made at that time. What behavior of the population is expected near Nm = 41 43. in that case we were able to model the growth of the deer population in terms of the deer population in previous years. different models of the population growth of a single species were discussed. [Hint: See exercise 41. Describe the result. Soon thereafter. Consider equation 42. perhaps by the finiteness of a nutrient which could be another species. The increased number of sharks in turn devoured so . we will describe some specific observations that have motivated ecologists to seek models of population growth. Introduction to Two-Species Models In the previous sections. The fish population in the upper Adriatic Sea forms an interesting ecological system. the population of sharks increased since they had more than the usual food available. smaller fish who are eaten by the sharks. Ib. We have already discussed models that involve the interaction of different species. during World War I fishing was suspended. 42. In that way we were able to use a single species model. One such example was the small ecosystem formed by deer and the vegetation they consumed. resulting in more small fish than usual. To simplify our discussion of this ecological system. Consider (a) Give a brief ecological interpretation of the three terms on the righthand side of the above difference equation. the growth of a species is limited. we will consider more complex ecological models in which two species interact. Describe the result. Only small changes in populations were observed to occur from year to year.18.19. The fishermen's catch of small fish was not removed from the sea.2]. Prior to World War I. 43 Introduction to Two-Species Models 42. (b) Show that Nm = 4is the only (positive) equilibrium population. and the plentiful plankton upon which the smaller fish feed. we assume that the fish population consists of sharks (and other voracious species). In this section (and the ones to follow). (c) Linearize the difference equation for populations near Nm — 4. determine ym. determine^. However.185 Sec. we will study situations that involve the interaction of more than one species. man's massive fishing industry had resulted in the populations reaching a balance. However. In the logistic growth model. (b) If a = £. Before developing mathematical models. However. (a) If a = 1.

yielding three distinct types of interactions. has on itself as we have already discussed onespecies problems. Thus We have allowed the growth of one species to depend on the populations of both species. Another example of an interaction between two species occurs in forests which are dominated by two similar trees. by symmetry. —K However.186 Population Dynamics—Mathematical Ecology many of the fish. Other examples of predatorprey type interactions include plant-herbivore systems and a parasite-host pair. We will limit our attention to deterministic mathematical models without time lags. enhances the other. and vice versa? We are not as interested in what effects JV. Shortly. If both populations negatively affect each other ( ). The growth of one population was followed by its decline. not other environmental factors. Likewise species N2 can affect species Nt in two different ways. then the biological interaction is called mutualism or symbiosis. . First we discuss the kinds of interactions that can occur between two species. the simplest example of such an interaction being when two species compete for the same food source. Thus there are four possible types of interactions between two species. then we say the two species are in competition. the effect of species NI is to either increase or decrease the population cf the other species. the prey. The interaction between sharks and the small fish they eat is an example of the third type of interaction (H—) called predator-prey. one of these interactions is equivalent to another. In later sections we will develop a mathematical model that discusses this type of interaction. some specific models will be suggested. In general. We will formulate a theory modeling the competition between these trees for the limited area of sunlight. namely H— is equivalent to —h. The existence of one species. ++. . their respective populations being Nt and N2. Consider a small ecosystem with two species. If both populations enhance the other (+ +). represented by the four sets of symbols H—. while the predator might threaten the very existence of the prey. As in the mathematical models of single species ecosystems. very few small fish were immediately observed (contrary to what they expected). we assume that the rates of change of each species depend only on the populations of each species. that when the fishermen returned after the war. What type of effect can Nt have on N2.

and Linearization 44. were also put into the form of a system of two first-order differential equations: To understand the solution of this system. 44 Phase Plane. and Linearization Before investigating any particular model or type of interaction. discussed in the first part of this text.1 which satisfies any given initial values of both populations. let us discuss the general model for the interaction of two species: If we assume that there is no migration of either species. then As before many other assumptions are involved in the above formulation. the phase plane equation was considered. Equilibrium. (t0) and N2 (/„)• It is interesting to note that the second-order differential equations of mechanical vibrations. that is determine the solution of equation 44. Phase Plane. Analogous to the above equation for the two-species population models is the equation which results by eliminating the explicit dependence on t: . TV. Equilibrium. The mathematical model consists of a system of two first-order (possibly nonlinear) ordinary differential equations. We wish to solve the initial value problem.187 Sec.

dN2/dNl = 0/0. The vanishing of both populations is an equilibrium population. The births and deaths of species Nt must balance. since neither g(Ni. N{ — Nle and N2 = N2e. For any equilibrium population. N2) depend explicitly on time. Thus we will start by considering some essential qualitative features of these types of equations.. Frequently an explicit solution by either method will be lacking. Let . but our first task in understanding complex ecological models would seem to be to study those models for which all other environmental factors do not change in time. the population model we have formulated is an autonomous system. i. a so-called linear stability analysis. it may not be difficult to calculate possible equilibrium populations (there can be more than one set). as they were in the analysis of oscillation problems in mechanics. Thus an equilibrium population. This is the same question we would ask had this been a one-species population model or an equilibrium position of a nonlinear pendulum! We proceed by investigating what happens in time if both populations are near their respective equilibrium populations. The next question is whether a known equilibrium population is stable. In other words. is such that both and two equations in two unknowns. The solution of the phase plane equation is said to give the trajectories of the populations. For a specific two-species model. called a singular point of the phase plane equation. A^) nor f(Nv. The two populations are called the phases of the ecosystem. We define an equilibrium population as a possible population of both species such that both populations will not vary in time. autonomous systems.188 Population Dynamics—Mathematical Ecology which we also call the phase plane equation. and similarly those of N2 must balance. either the time-dependent system of equations or the phase plane equation must be analyzed.) To study the interactions of the two populations. Singular points of the phase plane equation are equivalent to equilibrium points of the time-dependent equation. The solution of the resulting first order equation can be sketched by the method of isoclines. the slope of the phase plane diagram is undefined.e. (This is not necessary in all biological problems.

Do not forget that the coefficients of the unknowns (N^ and N2l) on the righthand side of this equation are evaluated at the known equilibrium population (Nle and N2e). j>0). 44 Phase Plane. or equivalently where the first three terms expressed above correspond to the linearization of the function in the neighborhood of the point (JCD. then the following is obtained: This is a linear system of differential equations with constant coefficients.189 Sec. eN^ and eN2l may be interpreted as the differences between the total population and the equilibrium population (referred to again as the population displacement from equilibrium). If we neglect the O(e2) terms (corresponding to the nonlinear terms in the neighborhood of the equilibrium populations) and if we recall that Nlt and N2e is an equilibrium population.3). yields We use the Taylor expansion for a function of two variables. and Linearization where € is a small number 0 < \e\ <C 1 and where Nle and N2e are an equilibrium population (satisfying equation 44. In this manner. In the next section we will discuss the solutions of such a general system of two coupled first-order linear differential equations with . Equilibrium. Substituting this into the time dependent population equations.

6.1.3. If equation 44. Knowing and understanding the phase plane of all possible linear systems are necessary in the investigation of the phase plane corresponding to nonlinear systems. at first we study the phase plane equation in the neighborhood of an equilibrium population. the slopes of the phase plane are approximated in the vicinity of an equilibrium population by the slopes of the phase plane of the linear system. Again perturbations may be introduced. equation 44. Show that singular points of the phase plane diagram are equivalent to an equilibrium population. EXERCISES 44. the ecological models discussed in Sees. 45-47). Depending on the nature of the solution.2. Consider the following nonlinear systems: . we may instead use the phase plane equation. Shortly the phase plane of such linear systems will be studied. 44. the equilibrium population will be stable or unstable. then we derive Comparing this with equation 44.190 Population Dynamics—Mathematical Ecology constant coefficients. and the nonlinear terms are neglected. 44. 48-54 may be interspersed with the mathematical developments of the next sections (Sees. In order to describe population growth far away from an equilibrium population. For those who prefer examples before a general discussion (the author usually does). The method of isoclines may again assist in the sketching of the solution curves. Recall in the study of mechanical vibrations that the most important features of the solution in the phase plane occurred in the vicinity of the equilibrium positions. Knowing the solution of the linearized equation will only indicate the behavior of each species in the immediate vicinity of an equilibrium population. expanded via Taylor's theorem. the linear systems of differential equations themselves must be analyzed.2. Before doing so. Hence. Show that the vanishing of both populations in a two-species model without migration is an equilibrium population.4 is substituted into the phase plane equation.

Suppose that 44. (b) Linearize the nonlinear system in the vicinity of each equilibrium solution. (b) Suppose that Nie and N2e are both nonvanishing equilibrium populations. Two methods to solve these equations will be discussed. 45. Such a system is of interest in studying the stability of an equilibrium population.4.191 Sec. (a) Give an ecological interpretation of the variables g\(Ni.6). Here x and y are the displacements from an equilibrium population (see equation 44.1 via its Taylor series around the equilibrium population. 45 System of Two Constant Coefficient First-Order Differential Equations 44. Do this and show that equation 44. What linear system of differential equations determines the behavior of populations near to this equilibrium population ? An equivalent method to linearize equation 44. System of Two Constant Coefficient First-Order Differentia/ Equations We will discuss the solution of a system of two coupled first-order homogeneous linear differential equations with constant coefficients. (a) Determine all real equilibrium solutions. .1 in the neighborhood of an equilibrium population is to directly expand equation 44.6 results.N2).5.N2) and gz(Ni.

then x can be eliminated from equation 45. y = Be*1. METHOD OF ELIMINATION The first method is straightforward (but has the disadvantage of frequently being cumbersome with problems involving more than two equations). yields the characteristic polynomial the two roots being Thus This solution will be discussed in Sees. then equation 45.192 Population Dynamics—Mathematical Ecology A. or after some algebra This equation is easily solved since it has constant coefficients. Seeking solutions in the form ert. Thus x can be solved from equation 45. If c ^ 0*. The system of equations is reduced to one second-order differential equation.la: . 46 and 47.Ib For this reason this method is called elimination.la using equation 45. *If c = 0.Ib is directly solvable for y.

e. SYSTEMS METHOD (USING MATRIX THEORY) An alternative to the method of elimination is to solve the linear system. using the matrix and vector notation: or where This method has the advantage that it is easily generalized to solve a larger number of coupled first-order equations. v 0 ^ 0)exist only for certain values .6(i. equations 45.193 Sec. 45 System of Two Constant Coefficient First-Order Differential Equations B.5 into equation 45..Ib. The substitution of LVoJ equation 45. In what ecological circumstances might there occur a larger number of coupled first-order equations? Solutions are again sought in terms of exponentials or equivalently where v0 = \ ° is a constant vector to be determined.4 yields I TC I or equivalently Nontrivial solutions of equation 45.la and 45.

Vx ~\ If the roots are distinct. Equation 45. we obtain (a . there are nontrivial solutions to a homogeneous linear system if and only if the determinant of the coefficients equals zero. called the eigenvalues of the matrix A.194 Population Dynamics—Mathematical Ecology of r. then the nontrivial value of v0 =\ ° corre- spending to each eigenvalue (called the eigenvector corresponding to that eigenvalue) must be determined. This yields two values for the exponent the same result as obtained by elimination. Thus or Solving this determinant. = LvoJ -*• I yd r x~\ 'L then the solution is given by where ct are arbitrary constants or . If an eigenvalue is designated r{ and its corresponding eigenvector v.6 can be written as a homogeneous system (using the identity matrix I): or From linear algebra.r)(d ~r)-bc = Q.

dvjdt = Av. where We look for solutions of the form By substitution we see that For nontrivial solutions of matrix equation 45. agreeing with the values obtained directly from equation 45.195 Sec. It is more con- venient to introduce Ci as the arbitrary constant associated with the eigenvalue . From either equation.9. We write both equations as a check against a possible error in our calculation of the eigenvalue: Both equations are indeed equivalent. The eigenvector corresponding to the eigenvalue r — —1 is determined from equation 45. The two equations so implied must be equivalent. The two eigenvalues are r = — 1 and r = — 2.12. *0 = y0 and thus the eigenvector is the eigenvector is any multiple of the constant vector . we know that Evaluating this determinant yields or (r + 2)(r + 1) = 0.12. let us solve the following system: In the matrix notation. 45 System of Two Constant Coefficient First-Order Differential Equations As an example.

In the example to follow. a multiple of his man- ner we have obtained the general solution of equation 45. However. for example. where Euler's formulas were utilized: Consider the example .11: or equivalently where ct and c2 are arbitrary constants. Two equivalent equations follow from equation 45.12: The eigenvector corresponding to r = —2 is the eigenvector is any multiple of the constant vectI it usually is easier not to deal with fractions.11 is We now do a similar calculation corresponding to the eigenvalue r = —2. it follows from equation 45. 5 and 12. which is why we prefer the eigenvector to be a multiple oan.9 that any complex eigenvalues at least must be complex conjugates of each other. and d are real. if a.196 Population Dynamics—Mathematical Ecology r — —\. we will illustrate how to obtain real solutions to the system of differential equations 45. c. Thus a solution of the system of equations 45. b. The eigenvalues need not be real as in the previous example.1 when the eigenvalues are complex. The ideas are quite similar to those discussed earlier in Sees.

we note from equation 45. the eigenvector must also be the complex conjugate of the above computed eigenvector. However. 45 System of Two Constant Coefficient First-Order Differential Equations upon looking for solutions in the form follows that Thus the eigenvalues satisfy which yields r2 — 2r + 2 = 0. note that multiplying the second equation by (3 — /) yields or 2(3 — /)x0 — 10x0 = 0. we can do a similar calculation. the solution of equation 45.15 that since r = 1 -f i is the complex conjugate of the eigenvalue already calculated.197 Sec. The eigenvector corresponding to r = 1 + / satisfies the two equivalent equations: (To show these are equivalent. Thus the eigenvector corresponding to r = 1 — / may be written as In this manner.) The eigenvector may be written as an eigenvector corresponding to a complex eigenvalue often has complex entries. To find the eigenvector corresponding to r = 1 — /. which is exactly twice the first equation.14 is . or the eigenvalues are complex conjugates of each other.

A discussion of this case will be avoided in this text.5). For x and y to be real. where A and 8 are arbitrary real constants. Although this modification is not particularly difficult for systems of two equations in two unknowns. we introduce its polar representation. 41).16 then yields Since c. equation 45. Since the sum of a quantity and its complex conjugate is twice the real part of the quantity.14: where a and ft are arbitrary constants. equation 45.198 Population Dynamics—Mathematical Ecology where c. it follows that Letting a = c. + c2 and ft = /(c. (see exercise 5. (see Sec. Thus Evaluating the real part yields a relatively simple alternative form of the general solution of equation 45.13. Using Euler's formulas. Interested readers can consult books on differential equations which contain a substantial treatment of systems and/or matrix methods. . — c2) (as we did in Sec. is an arbitrary complex number. it may be more complicated in higherorder systems.16 that cz must be the complex conjugate of c. An alternate expression for the general solution can be derived if we use some knowledge of the algebra of complex variables. it follows from equation 45. 5) yields a more convenient form of the general solution of equation 45. and c2 are arbitrary constants.14: If the roots are not distinct the heretofore mentioned technique must be modified.

analyzed in the previous section: .l 6= 2 c=-2 ( l ) a = 2 6=-l c = l1 (m)a=3 6 = 2 c = 0 ( n ) a = l 6=00 c = 0 (o)a=4 6 = 2 c = 2 Determine jc and >> as functions of time. Stability of Two-Species Equilibrium Populations The behavior of populations near equilibrium are governed by a system of constant coefficient differential equations.3 6--2 c= 1 (e) 0 = 1 6 . [Hints: Show that the eigenvalues are 1. Consider for the following cases : (a) a = 0 b= 1 c = -4 (b) a= 1 b= 3 c= 1 (c) a = 2 6 = 1 c=l 1 (d) a = .] 46.2. Consider d= 0 d=-\ d= 2 d=-5 rf=-2 </= 2 d = -3 rf = -3 </= 1 rf=0 rf=-l r f = 0 < / = 4 d= -3 </=! Find the general solution.1. 2.199 Sec. 45.l 6= 3 c= 1 ( j ) a = 4 6=-3 c=l (k) a = . 46 Stability of Two-Species Equilibrium Populations EXERCISES 45.2 c=-l (f) 0 = 3 b= -1 c= 1 (g) a = -2 6 = 0 c= 0 (h) a = 1 6 = 0 c= 1 (i) a = . and 3.

e". one root zero q= 0 (a) other root positive p > 0 UNSTABLE (b) other root negative p < 0 NEUTRALLY STABLE *See p.3).2). . Since we are interested in these equations primarily in the context of the stability of equilibrium populations. while the roots are complex conjugates of each other if (a + d)2 — 4(ad ~ be) = 0. roots (1) real and unequal (2) real and equal (3) complex conjugates condition A> 0 A= 0 A < 0. The following table describes the many cases: (1) real and unequal A> 0 A. same sign q> 0 (a) both positive p > 0 UNSTABLE (b) both negative p < 0 STABLE B. It is usual to let Using these definitions.200 Population Dynamics—Mathematical Ecology It was shown that the solution consists of a linear combination of exponentials. different signs q < 0 UNSTABLE C. 102. the roots are real and equal if (a + d)2 — 4(ad — be) — 0. This is determined by the signs of the two values of the real part of the roots (see equation 45. The roots are real and unequal if (a + d)2 — 4(ad — be) > 0. 18) if the real parts of both roots are less than zero. we must discuss what happens as t —* oo. where are the roots of the characteristic polynomial (see equation 45. unstable if at least one is greater than zero. neutrally stable* if the real part of one root equals zero and the other is less than or equal to zero (except if both roots are identically zero in which case the solution is algebraically unstable). 58 and p. An equilibrium solution is said to be stable (see Sec.

A). *Except if a = b = c = d = 0. d (yielding small errors in the measurements of p. 27 referring to the stability of an equilibrium position of a spring-mass system with friction.3).201 Sec. . real part has same sign (a) real part positive p > 0 (b) real part negative p < 0 B. b. then certain of the above cases are subject to change. in which case we say this is NEUTRALLY STABLE (see exercise 46. both zero p= 0 (3) complex conjugates A< 0 A. A similar diagram appeared in Sec. 46-1. Examples are given in the exercises and also in later sections. In this manner we are able to determine if a given equilibrium population is stable or not. real part zero p=0 UNSTABLE STABLE ALGEBRAICALLY UNSTABLE* UNSTABLE STABLE NEUTRALLY STABLE These results are summarized in Fig. 46 Stability of Two-Species Equilibrium Populations (2) real and equal A —0 A. same sign (a) both positive p > 0 (b) both negative/? < 0 B. c. If there are small errors in the measurements of the quantities a. namely any case which contains an equal Figure 46-1 Stability diagram (the shaded region is stable). q.

Topics which are not covered in this text show that in the major cases the behavior of the linear system is an accurate approximation to the nonlinear system (at least in some immediate neighborhood of the equilibrium solution). Usually only the quadratic terms from the Taylor series are then necessary. The phase plane analyses for the major cases are described in the next section. some of the nonlinear terms neglected in the linearization process may be necessary in order to yield an accurate approximation in the vicinity of the equilibrium point. y = 0 is neutrally stable. Consequently the borderline case q = 0 will not be considered. but sometimes even additional terms are needed. 46. The major cases are (l)A.202 Population Dynamics—Mathematical Ecology sign. In considering population models we will concern ourselves only with cases having isolated equilibrium populations. y = 0 is not isolated. since there are other equilibrium points arbitrarily close.1 grow in some algebraic power of t. then the linear system (46. A < 0—case (3)B). If q = 0 (i. Show that in this latter case the equilibrium x = Q. while the others are referred to as borderline cases.1 is stable if Use the diagram in this section to determine the stability of the solution x = 0. The equilibrium x = 0..1 is said to be algebraically unstable if solutions to equation 46. y — 0 of the linear system of equations 45.e. For the stability of this latter case as well as the phase plane analysis for all borderline cases.2. Those cases in which small errors in measurements cannot induce changes are called the major cases. en. tx. In this case the equilibrium point x = 0. (1)B. . y = 0 for the examples in exercise 45.1.1. 46. and (3)A. Show that the solution x = 0. y = 0 of equation 46.1) has an equilibrium point not only at x — 0. Furthermore. y — 0 but at any point along the straight line ax + by = 0. the stability of a borderline equilibrium population for a nonlinear system is the same as the stability of its linearization except for the one borderline case that divides regions of stability from instability (p = 0. Show that in case (2)B the equilibrium is algebraically unstable except if a = b = c = d = 0. rather than exponentially in time. ad = be).3. Another distinguishing feature between the major and borderline cases is now described. EXERCISES 46. Consider the case (2)B in which p = 0 and A = 0. The study of linear systems was motivated by considering the linearization of a nonlinear system.

46). y = 2 is an equilibrium solution. (b) Linearize this system near this equilibrium. y = 2 stable? [Hint: Eliminate JCj or yt. The phase plane equation alone cannot determine stability or instability.] 47.4.203 46. What is /?? (c) Given that the linearization is of the form above. Why? [Answer: A term in the solution similar. Phase Plane of Linear Systems A. (a) Show that x = 1. but the resulting phase plane of the linear system will yield an approximation to the phase plane of the nonlinear system valid in the neighborhood of the equilibrium population (at least in the major cases as described in Sec. 47 Phase Plane of Linear Systems Consider where a is a constant that may be positive or negative. arrows are introduced via the time-dependent equations. for what values of a and ft is x = 1. To determine how the population changes along a trajectory. yi are respective displacements from equilibrium. to e~* . Reversing the role of time (replacing t by — /) may change a stable equilibrium point into an unstable equilibrium point. Show that where xj. GENERAL REMARKS Not only does analysis of the linear system determine the stability or instability of an equilibrium population. Sec. for example. The equation for the phase plane of the linear system is This equation does not explicitly depend on time.

the phase plane equation remains unaltered.] However. We reiterate. consider the following two systems: System (1) is unstable. As another example. but q and A do not.e. while the solution of system (2) is x = x e 2t o ~ > y = y*e~u-] However. while system (2) is stable.. Why ? [Answers: The solution of system (1) is x = x0e'. If we change all the signs of a. growing exponentially. a replaced by — a. b. . y = y0e'. c. the cases distinguished by p > 0 and p < 0 need not be considered separately. Arrows will indicate motion in opposite directions for increasing time and will be the only distinguishing feature. the phase plane equation remains the same. to determine stability the time dependence must be considered. and so on). Figure 47-1 Typical isoclines corresponding to zero and infinite slopes.204 Population Dynamics—Mathematical Ecology becomes e+t. d(i. it would be impossible to determine the stability or instability of an equilibrium population. Thus in discussing trajectories in the phase plane. Thus since p changes sign. b replaced by —b. the phase plane differential equation for both is the same: Thus given the phase plane equation.

For example. We will show that the result of this simple example is quite similar to the general case. These curves are sketched in Fig. Before discussing this case in general. (1)B (A > Q. The "roots of the characteristic equation" are —1 and +2 (having opposite signs) as follows from the exact solution The phase plane differential equation. namely along ex + dy — 0. The borderline cases are discussed in the exercises. q < 0)] is analyzed first. q < 0) The case in which the roots of the characteristic polynomial are of opposite signs [(1)B (A > 0. we propose to illustrate the general configuration for each of the three major classes. B. dy/dx = 0 and along ax + by = 0. 47 Phase Plane of Linear Systems We will show that to each different case there corresponds only one qualitative behavior in the phase plane. dy/dx = oo.205 Sec. yielding This equation for the trajectories can also be obtained by eliminating / from the time-dependent solution. SADDLE POINTS (A > 0. 47-1. We will consider only the three major cases: (1)A (A > 0. and (3)A (A < 0). first simple isoclines might be sketched. see Fig. Suppose that the differential equations were not coupled. 47-2 for a few different values of k: .? > 0). is separable and can be directly solved. the simplest possible example yielding roots of opposite signs will be solved. Instead of sketching additional isoclines. In sketching the phase plane of dy/dx = (ex + dy)/(ax + by).q < 0).

then certain regions correspond to high ground and others to low ground. For this reason this type of equilibrium point is called a saddle point. then the solution . 47-3 resembles the shape of a saddle with the "seat" at x = 0. If the roots are of opposite signs. Trajectories for a system with characteristic roots of opposite Arrows in time are introduced in Fig. Imagine the phase plane trajectories of this example as the paths of a person moving along some hill z = f(x. y = 0. y = 0) is unstable as most trajectories near that point eventually tend away from the equilibrium point. but it can be shown that they take an infinite length of time to get there. y) as viewed from above. 47-2 from the following facts: The equilibrium point (x = 0.206 Population Dynamics—Mathematical Ecology Figure 47-2 signs. For this case the equilibrium is always unstable. x > 0 and y = 0. The two special trajectories y = 0. jc < 0 tend towards the equilibrium point. The resulting hill as illustrated in Fig. If the person moves down the hill in some manner. We will show that whenever the two roots are of opposite signs-then in the phase plane the equilibrium point is a saddle point.

then We may assume that since the roots are real and of opposite signs. To sketch the trajectories in the phase plane. must be a linear combination of the two exponential solutions. A trajectory which starts at a finite point and goes in a straight line to infinity is called a .207 Sec. Instead. then and t is easily eliminated: This straight line represents two trajectories emanating from the origin (as t —> —oo. Referring to the time-dependent equations. x —* 0 and y —> 0 since r2 > 0) and going toward infinity in a particular direction depending on the signs of (rz — d)\c and c2. some specific trajectories are sketched. 45 that if c ^ 0. if ct = 0. This is not easy in general. 47 Phase Plane of Linear Systems Figure 47-3. it was shown in Sec. For example. / must be eliminated.

one of many different orientations of the four simple trajectories. then we have Fig. 47-4.208 Population Dynamics—Mathematical Ecology ray. We will show that the other trajectories (cl =£ 0 and c2 ^ 0) do not go through the origin. 47-4 (assuming that (r2 — d)\c > 0). since r 2 > 0 and /•i <0. As t—»+oo. it would violate a uniqueness theorem for the initial value problem. t is eliminated easily: Although this also represents two trajectories emanating from the origin. Thus if (rl — d)jc > 0 and (r2 — d)\c > 0. since then Thus again. . The direction of these trajectories are determined by the sign of Cj and the sign of (rt — d)/c. Figure 47-4 Some important trajectories. If they intersected the origin (the equilibrium point) at a finite time. Of more significance to the phase plane sketch is to notice that these curves do not tend towards the origin as / — > ±00. These two rays are sketched in Fig. these trajectories are tending to the origin as t —-> oo since rl < 0 (and tend towards infinity as / —> — oo). Two other simple trajectories occur if c2 = 0.

If cl ^ 0. in the direction xfy —* (rl — d)jc. Figure 47-5 Sketch of trajectories in the vicinity of a saddle point. In this manner we can sketch the phase plane of a saddle point. 47-2).e. 47-5. the only trajectories that intersect the equilibrium . tending towards being parallel to the outward going ray. as shown in Fig. assuming GI > 0 and c2 > 0) is sketched in Fig. this time parallel to the inward going ray. Similarly as /__oo. 47-4 in partially dotted lines (other cases will be sketched in an analogous fashion). The trajectory corresponding to one value of Cj and c2 (in addition.209 Sec. in the direction i. In order to sketch the phase plane quickly the important curves are the straight line trajectories. an unstable equilibrium point corresponding to one positive and one negative root. 47 Phase Plane of Linear Systems Thus all trajectories (c2 ^ 0) tend towards infinity as t —*• +00. The curve is dotted. indicating the general behavior the curve must have for values of t which are not either large and positive or large and negative. Note that the phase plane has the "similar" general shape as that previously sketched for the example of a simple system with a saddle point (see Fig. these trajectories also tend towards infinity as /—> — oo..

1 with a = \. Knowing the straight line trajectories and the simple isoclines (along which dy/dx = 0 and dy/dx = oo) is sufficient to quickly give a rough sketch of the phase plane when a saddle point occurs. . An easy method to determine all straight line trajectories (assuming a book with the answer is not readily available) is to substitute the assumed form of a straight line trajectory into the phase plane equation. As just discussed this will be a saddle point (since A > 0 and q < 0).q= —4. that is the above formula is equivalent to y/x = cf(r — d). A = 5. As an example. consider which is in the form of equation 45.b = — l.1 p = —\. gives two values It can easily be shown that these slopes are the same as obtained before. In some of the other cases a straight line trajectory is also of importance. which yields an equation to evaluate the slope Solving for m. Since the straight line goes through the equilibrium point. The phase plane equation is sketched in Fig.210 Population Dynamics—Mathematical Ecology point. 47-6 by first noting the isoclines.d= —2. Hence from equation 46.c = —2.

where arrows are introduced from the time dependent equation. Straight line trajectories. satisfy m = — 2[(1 + m)/(l — m)] or equivalently /M2 — 3m — 2 = 0. 47 Phase Plane of Linear Systems Figure 47-6. and thus finally Therefore we immediately have the reasonably accurate sketch shown in Fig. 47-7. Figure 47-7 Trajectories in the neighborhood of a saddle point (based on simple isoclines and straight line trajectories). y — mx.211 Sec. .

it is again convenient to first discuss a simple example. where the arrows are easily added since x = 0. NODES (A > 0. y — 0 is an unstable equilibrium point. Sketching these parabolas yields Fig. . for example. 47-8. q > 0) To analyze the case in which both roots have the same sign. This is a case in which both roots are positive. if Figure 47-8 Unstable node: trajectories.212 Population Dynamics—Mathematical Ecology C. In either way x — ky2. The exact solution is The trajectories are obtained by eliminating t from the above differential equations (dyjdx = y/2x) or by eliminating t directly from the solution. If both roots are negative.

Examples can be constructed in which the trajectories near the equilibrium are not parabolas. In these examples all trajectories enter the equilibrium point. 47 Phase Plane of Linear Systems then the trajectories are identical. a stable node or unstable node. the solution (if c ^ 0) is Both roots are now considered to have the same sign. In general. Consider The trajectories of this equation are given by the formula They appear similar to parabolas (as sketched in Fig. The equilibrium point is then stable.213 Sec. either . depending on whether both roots are negative or positive respectively. 47-9): Figure 47-9 Stable node: trajectories. only the direction of the arrows is different. This type of equilibrium point is called a node.

47-11. Since r 2 > r. 47-10 assuming both slopes are positive: Figure 47-10. the asymptotic behavior of the solution as t —-» ± oo helps to sketch the complete trajectories. both roots positive and distinct) or all tend towards the origin (stable. as f —> +00 . The time-dependent equations indicate that the arrows on these trajectories either all emanate from the origin (unstable. Once more these trajectories are straight lines with the same slopes as in the saddle point case: sketched in Fig. Figure 47-11. both roots negative and distinct) as shown in Fig.. Simple trajectories are again determined by considering those with c\ = 0 and those with c2 — 0.214 Population Dynamics—Mathematical Ecology or We will show a node always results. As in the case of a saddle point.

they do so tangent to the other special straight line trajectory. . This configuration is referred to as a stable node or an unstable node. the trajectories always have the type of sketch shown in Fig. namely As both x and y tend towards infinity.215 Sec. For example. then it is an unstable node and arrows are introduced emanating from the origin. as the trajectories approach the origin. Figure 47-12 Node: typical trajectories. 47 Phase Plane of Linear Systems Similarly as t —> — oo the opposite occurs. 47-12. the trajectories tend parallel to one of the special straight line trajectories. Furthermore. Thus. The simple isoclines along which dy/dx — 0 and dy/dx = oo can again be helpful in improving the accuracy of the sketch. if both roots are positive.

the alternate form (obtained by .216 Population Dynamics—Mathematical Ecology D. / 7 ^ 0 ) When the two roots of the characteristic equation are complex conjugates of each other. as we proceed to show. SPIRALS ( A < 0 . then we would expect an oscillation with the amplitude either growing or decaying. The phase plane for this example consists of closed curves (in general ellipses) encircling the equilibrium point called a center. if the roots have both nonzero real and imaginary parts. A simple example of equation 47. A simple example of a system of equations whose roots are complex numbers is The roots of the characteristic polynomial are determined by yielding or equivalently the complex conjugate pair The trajectories in the phase plane can be obtained from This first-order differential equation is not separable.2 corresponding to imaginary roots is the equations of a spring-mass system without friction. the analysis is more complicated than in the other cases. In the phase plane we should not be surprised that the resulting trajectories are spirals. then the solution exhibits a pure oscillation. Although the solution can be sketched using the method of isoclines. However. If the roots are purely imaginary (excluded by us as being a borderline case).

47 Phase Plane of Linear Systems suggests that the change of variables. Returning to the x. .* Since dyjdx = x(dvfdx) + v it follows that The equation is now separable. the roots are purely imaginary. the curves become the closed curves r = r0. circles. will enable its explicit integration.. i. in which case we see after exponentiating that Such curves are called exponential spirals (if a ^ 0). y variables (v = y/x) yields or This form suggests the introduction of polar coordinates. Each component of the solution (x or y) exhibits simple harmonic motion *In fact this method works in all cases (see equation 47. This is not surprising for if a = 0.1).217 Sec.e. What happens if a = 0 and why? [Answer: If a = 0. after some algebra Integration yields where c is the constant of integration. ±ib.

there is an unstable spiral if a > 0 and a stable spiral if a < 0. it is also conyeliient to introduce the polar angle 0 Thus Figure 47-13 Left-handed spiral . The stability or instability also follow from the exponential term ertt where r — a^t ib. In the general case of complex conjugate roots. The trajectories are circles. x decreases if by < 0 and x increases if by > 0. For example at x = 0. The time-dependent equations determine the temporal evolution of the solution. Similar curves can be sketched if b > 0 (see the exercises). then as 0 increases. For b < 0.218 Population Dynamics—Mathematical Ecology with circular frequency b. then as B increases. dx/dt = -\-by. 47-13. as / increases. These types of trajectories are also referred to as an unstable focus and stable focus respectively. Thus one trajectory is as shown in Fig. 47-15 (assuming b < 0). r increases. Thus at x — 0. perhaps most easily seen by the following calculation: If a/b < 0. If a/b > 0. see Fig. r decreases as shown in Fig. 47-14.

. Figure 47-15 Spiral trajectories. 47 Phase Plane of Linear Systems Figure 47-14 Right-handed spiral.219 Sec.

However. Then. or the sign of —b.the angle continually either increases or decreases. consider . for example. The solution either spirals into the origin or spirals out from the origin. the criteria for complex conjugate roots is or equivalently Thus there are no values of x and y for which d0/dt = 0. which must be the same (why?)). that is and In the example to follow. we show how these isoclines enable the sketching of the phase plane. Therefore d&/dt is always of one sign (the sign of c. by dividing by y2. recall the simple isoclines. Knowing that 0 increases in time (or decreases) is not sufficient to accurately sketch the trajectories. or However.220 Population Dynamics—Mathematical Ecology Eliminating dx/dt and dy/dt from the differential equations. yields We show that dO/dt ^ 0. For the roots to be complex. Suppose dQfdt = 0. A < 0 or equivalently For example.

dy/dx = 0. see Fig. Thus the trajectories must be right-handed unstable spirals. figure 47-17. 47 Phase Plane of Linear Systems Along x + y = 0. dd/dt < 0. . and along — 4x + y = 0. Since r = 1 ± 2/. yielding either Fig. Figure 47-16. 47-16. or equivalently 1 — r = =f 2i. the equilibrium point is unstable. 47-17a or b. The characteristic roots are determined as follows: Thus (1 — r)2 + 4 = 0. Since c is negative.221 Sec. An improved sketch is made in Fig. Thus 0 decreases in time. dy/dx = oo. 47-18 using the known simple isoclines.

Figure 47-19 Stability diagram including typical trajectory in the neighborhood of equilibrium (shaded region indicates a stable equilibrium point). 222 .Figure 47-18 Sketch of a spiral (using isoclines).

47. Consider the linear system Suppose that it has been shown that (a) Sketch the trajectories in the phase plane if a = 1 and b = 2. if k2 < 4Lg. Consider the systems Show that the phase plane analysis of both problems are the same although their solutions are entirely different. Consider equation 21. EXERCISES 47. Show that this linear system is a borderline case. (c) in the neighborhood of 0 = 0. (b) Sketch the solution in the phase plane. Reconsider exercise 45. then what would you conclude concerning the equilibrium population and the phase plane in its vicinity ? 47.2. y = 0. .3.1. Sketch the solution in the phase plane (dxfdt as a function of x). Sketch the solution in the phase plane (dO/dt as a function of 6): (a) in the neighborhood of 0 = n. 47.223 Sec. (a) Determine the stability of the equilibrium solution x = 0. SUMMARY The results of this section and the previous one are summarized in Fig. describing the nonlinear pendulum.1.1. if k2 > 4Lg. Consider equation 28.1. (b) Redo part (a) with a = — 1 and b =2.5. (c) If this system was the linearization of a nonlinear population model. 47 Phase Plane of Linear Systems E. describing a linear spring-mass system. 47-19. 47A. (b) in the neighborhood of 6 = 0.

If the fishermen refrained from fishing for the small plankton-eating fish for a couple of years (as occurred during World War I). close to what has just been described. and in a short time pose a severe threat to the small fish. Although this may seem somewhat unreasonable. if not impossible. frequently investigated in mathematics texts. Many other assumptions are involved in stating that Before formulating a specific model. certain observations (see Sec.224 Population Dynamics—Mathematical Ecology 48.) Let (number of a certain species of F = I fish (eaten by sharks) in a specific (region of the sea S = number of sharks in the same area. 43) indicate oscillatory fluctuations. let us again describe the effects we would like to model. This description of shark-fish interaction applies to other predator-prey situations. For another example of . in which case the ecosystem would consist of periodic population variations. It is possible that this process continues on indefinitely. describes the ecosystem of foxes and rabbits. Thus the population of sharks would increase. then these fish might be expected to increase in number. The area is assumed bounded in a way such that migration across the boundary is. (Another predator-prey situation. Once the small fish have diminished in number. the sharks would have enough food to sustain a larger population of sharks. Once having increased. We will study the interaction in the sea between sharks and the small fish consumed by the sharks. Eventually the population of the plankton-eating fish would diminish. Predator-Prey Models One of the first two-species ecosystems which has been mathematically modeled involves a predator and its prey. This in turn allows the small fish to return towards their original population. Thus the model to be developed may have validity to other ecosystems besides the one involving sharks and fish. the sharks can no longer sustain their enlarged population and must decrease in number. at least unlikely to be a major factor.

Derivation of the Lotka-Volterra Equations In developing a model of an ecosystem representing the interaction of its two species. it is advantageous to first model the population growths ignoring the interactions between the species. 49 Derivation of the Lotka-Volterra Equations two-species interactions. In considering the fish without sharks. 1953). This assumes that the plankton is unlimited. Fundamentals of Ecology. 48-1.225 Sec. The birth rate is larger tian the death rate. Saunders. P. yielding an exponential growth. Thus we ask what equation should the population of sharks satisfy if there were no fish and vice versa. B. Observational data of these populations have been available for many years. the assumptions concerning the birth and death processes of the fish must be delineated. keeping yearly records of the numbers caught (presumably reflecting the total populations of lynx and hare). Philadelphia: W. consider the lynx (a type of wildcat) and hare populations in Canada. Figure 48-1 Oscillation observed in Canada of populations of lynx and hare (data from E. The data reveals remarkably periodic fluctuations of the populations of lynx and hare as illustrated in Fig. Since the fish eat plankton which are presumed to be very abundant. Odum. 49. On the other hand. it is suspected that the growth rate of the fish (without sharks) is constant. if the . The Hudson Bay Company trapped both the lynx and its primary food the hare.

where c is this positive proportionality constant. then the food source of the sharks is nonexistent. Thus To model the growth of fish. S) = — kS. c. where /z(0. becomes a — bF — cS with sharks. which was a — bF without sharks. b. we assume that the growth rate of sharks is increased proportional to the number of fish. 0) = aF or g(F. then a logistic growth model might be proposed. We again assume this interaction effect on the growth rate is proportional to the population. Hence in the absence of fish we assume that With no fish. If there are no fish. is modeled as —k + AF with fish. The latter equation incorporates the first. is such that g(F. F = a/b. the sharks would be an endangered species eventually vanishing. dF/dt = g(F. The growth rate of the sharks. 0) = aF — bF2. The existence of fish enhances the shark population. However. 0) is given by either g(F. We now will model the complex interaction between fish and sharks. which was —k without fish. If b ^ 0. Thus we have determined dS/dt = h(F. Thus we have described the background to the set of differential equations developed independently by Lotka and Volterra in the 1920s. Thus. thus fish will cause an increase in the growth rate of the sharks. In this case. S). where A is the positive proportionality constant. then there exists an equilibrium population offish. For the simplest mathematical model of this process.226 Population Dynamics—Mathematical Ecology population growth of the fish ceases at some large population. the death rate of the sharks is expected to exceed the birth rate. Consequently the growth rate of the fish. where the constants a. in this case the existence of sharks will decrease the growth rate offish. The sharks behave in an entirely different manner. k describing the different processes are all . A. S1). if there are no sharks. the rate of population change of the fish. we use similar ideas.

show that (e) Solve this difference equation and determine the behavior of the solution as time increases. ym = 2 satisfy the system of linear difference equations (d) If xm is eliminated from this system. (c) Show that small perturbations to the equilibrium population xm = 3. (You should be able to quickly explain what effect each term in these equations represents.) This is not the only possible model of a predatorprey ecosystem. A two-species ecological system is described by the equations where xm and ym are the populations at time mAf. Consider the following three-species ecosystem: 49. Other models are suggested in the exercises in later sections. (b) Find all equilibrium populations. Assume that the coefficients are positive constants.1. 49 Derivation of the Lotka-Volterra Equations positive (except for b which may = 0 if an unlimited amount of plankton is permitted). but it is one of the simplest. The question you should now be asking is: Does the qualitative behavior of solutions of this nonlinear system of differential equations seem consistent with the observed oscillatory behavior? EXERCISES 49. Describe the role each species plays in this ecological system.227 Sec.2. . (a) Briefly describe the interaction.

228 Population Dynamics—Mathematical Ecology 50. If the growth rate offish increases. If at some time the sharks number exactly S — a/c. The equation describing the growth of the shark population can be analyzed in a similar manner. We will discuss the case in which the food source (plankton) is unlimited. let us briefly explain how the sharks limit the growth of the fish. reasonable? Solutions must be calculated and compared to observations. the growth rate of the fish in the absence of sharks and c. Before attempting to solve this equation. then at that time the fish population does not vary. this level of sharks decreases.. These equations imply that for a sufficiently large population of sharks (S > a/c) the fish population diminishes. the inhibiting effect of the sharks on the fish. This level of sharks depends on a. Equivalently if the sharks become more effective in attacking the fish (larger c). Suppose initially that there are very few . Qualitative Solution of the Lotka-Volterra Equations Is the model proposed for the population changes of a predator-prey system. Then the sea environment can sustain an infinite population of prey (fish) if there were no predators (sharks). then this level of sharks increases. Similarly if c is increased. then less sharks are necessary to prevent the fish from growing. in which case the ecosystem is governed by the following system of coupled nonlinear ordinary differential equations: The case in which b ^ 0 is posed as an exercise. This is equivalent to assuming that b = 0.

then dF/dS = 0 and if S = 0. which also must be equilibrium populations. equation 50. since to do so it must cross either axis. 50 Qualitative Solution of the Lotka-Volterra Equations sharks. Consequently. the trajectories of the solution in the phase plane will be sketched: Unlike the phase plane of mechanical systems. though knowing the . only the first quadrant is necessary since both the populations F and S must be positive. From the differential equations it is clear that the fish will grow in number. the number of fish will cease to increase. Intersections of isoclines corresponding to different slopes of the solution must be singular points. As the number offish increases. as shown in Fig. namely The case of zero population (2) is not as important. At that time. Eventually the number of sharks will become so large that the growth rate of fish will become negative. We note that if F = 0. which is an impossibility. unlike the axes. Thus the lines F = 0 and S = 0 are both isoclines as well as solution curves.229 Sec. There are only two possible singular points.2. Instead let us investigate mathematical properties of the special case of the Lotka-Volterra equations. Hence in order to understand at least the qualitative behavior of the solution (and also certain quantitative behavior). the number of sharks must increase. 50-1. We should first verify that the model does not have the indignity of predicting negative populations. No simple explicit solution to these equations in terms of elementary functions has been obtained. then dF/dS = oo. a solution once positive can never become negative. Other simple isoclines are: and These isoclines are not solution curves. We could continue the qualitative description of what should happen based on the differential equations.

then the equilibrium population offish remains unchanged. corresponding to improving the sharks' ability to kill fish.230 Population Dynamics—Mathematical Ecology Figure 50-1 Simple isoclines for a predator-prey ecosystem. k. The following paradoxical behavior of this system is only understood with some care. only the sharks are affected. but their increased number consumes more fish exactly offsetting the increased birth offish. a. giving the sharks more food. the sharks grow in number. if there are no sharks then the fish will . Hence. results in a decrease in sharks. but in decreasing the fish. then dF/dt = aF. solution curves in the neighborhood of this equilibrium population assists in the sketching of the phase plane. then the predator-prey ecosystem has reached a dynamic balance. The explanation of this last phenomena is that it takes fewer fish to balance a hardier shark population. The equilibrium populations depend on all four parameters. This results not in increasing the sharks. On the other hand. Similarly increasing c. If the growth rate offish. Can you explain this last phenomena? If the populations are not at equilibrium. (Ecologically. If F = k/A. then not only does this not effect the equilibrium population of sharks but more unusually the equilibrium number of fish is actually decreased. increasing a) results in more fish being hatched.e. If S = 0. then the phase plane is employed to determine temporal population changes. also behave in seemingly surprising ways. The efficiencies of the interactions. but only the appropriate two ratios are significant. An increased efficiency of the predator results in the decrease of the equilibrium number of prey.. that is F always increases. We will show that such an equilibrium population is stable. >l and c. is decreased. and 5" = a/c. In other words increasing the egg-laying capacity of the fish (i. is increased. An increased number of sharks is necessary to balance the growth of the fish. if the death rate of the sharks. Increasing A corresponds to the fish containing additional nutrients for the sharks.

a similar statement about sharks is Figure 50-3 Qualitative behavior of predatorprey trajectories. . Why is this reasonable? Also the fish decrease if the number of sharks is greater than its equilibrium value (S > a/c).2. grow. if F = 0. A similar analysis can be done near the equilibrium shark population. Furthermore. 50-2. If the fish are at the number necessary to balance the growth of sharks. then dS/dt = —kS and S decreases (the sharks tend towards extinction). as shown in Fig.) Similarly. 50-3. In fact for all population of fish. and follow from equation 50. the fish are increasing in number if the sharks are fewer than their equilibrium population and vice versa. The fish increase from that value if the number of sharks is less than its equilibrium value (S < ajc). F = k/Ji.231 Sec. The resulting variation in shark population yields Fig. Mathematically. S = a/c. 50 Qualitative Solution of the Lotka-Volterra Equations Figure 50-2 Single species trajectories. then dF/dt = kfJi (a — cS). These results determine the population trajectories if either species initially is not present.

The phase plane diagram indicates in general a clockwise pattern. For example. etc. As in mechanics.) that appear in the phase plane diagram. there are at least three possible kinds of "orbits" of the phase plane. then a "little common sense" would suggest that Figure 50-4 Possible orbits of the phase plane. If the solution curves are sketched. and also part of another solution curve is spiralling outwards (as though the fish and sharks increase in numbers after some time). Figure 50-5. then it is apparent that the populations may oscillate around the equilibrium values of the respective populations. The isoclines along which dFjdS = 0 or oo are seen to be quite useful as these curves separate regions for which a species is increasing from regions for which the same species is decreasing. This explains the additional arrows (<7. 50-3. If this is valid. 50-4. . As drawn. 50-5. suppose that the trajectories are as shown in Fig. part of one solution curve is spiralling inwards (as though the populations of fish and sharks after an oscillation approach closer to their equilibrium values). ^.232 Population Dynamics—Mathematical Ecology valid regardless what the population of sharks is. Fig. Figure 50-6 Limit cycle (darkened curve). as shown in Fig.

2 yields Neglecting the nonlinear terms results in This system of linear differential equations can be most easily solved by elimination. 44. F = k/A and S = a/c.233 Sec. such that for that solution the population returns to the same value. for example. 50 Qualitative Solution of the Lotka-Volterra Equations there exists a solution curve "in between" the two drawn. 50-6 does not occur for equation 50. The differential equations 50. To elucidate the behavior of this fish-shark predator-prey ecosystem. yielding a periodic oscillation of the population. 50-6. We will show that this phenomena as sketched in Fig. eliminating F1 from equation 50.2. it can occur in other population models. we should investigate both species when they are near their respective equilibrium values.6 yields or equivalently . However. called a limit cycle as sketched in Fig. We do this by employing a linearized analysis as suggested in Sec.2 may be linearized in the neighborhood of the equilibrium population by using perturbation methods as follows: where eFl and eSl represent population differences from the equilibrium population. Substituting this into the differential equations 50.

In the vicinity of the equilibrium population. If at t = 0. But we must be exceedingly careful because this small change may affect the entire qualitative behavior of the solution. As the solution completes a "cycle" in the phase plane." an "energy" integral. This result can also be obtained from the phase plane differential equation. where the circular frequency co = (ak)l/2. In other words. The results of a linearized analysis predict that populations oscillate with constant amplitude in the neighborhood of the equilibrium population. The period of oscillation. In terms of St and Fj which is an ellipse in the phase plane. this conclusion is not always justified. Note that S10 is the displacement from equilibrium population of sharks at t = 0. Thus we know what is demonstrated in Fig. depends only on the growth rates a and k. Fl = F10. However. by separation and integration. The equilibrium population appears to be stable. instead of its . The small nonlinear terms (neglected in the linearization) probably will make only a small change. This period is only valid for small oscillations around the equilibrium population.6.7 and a springmass system or a linearized pendulum.234 Population Dynamics—Mathematical Ecology Thus. there is a "constant of motion. 50-7. the number of sharks oscillates around its equilibrium population. readily determined from the analogy between equation 50. since from equation 50. then Flo = (cjatycoQ. The fish also oscillate.

However. as discussed at the end of Sec. This borderline situation is characterized by the solution oscillating with constant amplitude. perhaps the nonlinear terms may cause it to oscillate with smaller and smaller amplitude. In this case the nonlinear terms may make the equilibrium solution more stable than predicted by the linearized stability analysis.255 Sec. If this occurs. In this case the equilibrium population would be unstable even though the linearized stability analysis predicted stability. the opposite can occur. In summary. 50 Qualitative Solution of the Lotka-Volterra Equations Figure 50-7 lation. y = S^ and . 47. It is possible for the neglected nonlinear terms to cause the solution to oscillate with increasing amplitude. the stability of an equilibrium population is always determined by a linearized stability analysis except when the linearized stability analysis predicts the equilibrium population is on the borderline between being stable and being unstable.6 are put into the form where x = Fl.5-50. our original prediction of stability is unaltered. The nonlinear problem for a predator and its prey must be investigated! If you have already studied Sec. Predator-prey phase plane oscil- coming back exactly where it started (yielding a periodic solution). In this case the linearized stability analysis is inconclusive. Equations 50. let us briefly show that the same conclusion can be reached by quoting the general results for linear systems. 47.

we . plane in the neighborhood of the other equilibrium population. Let us now reconsider the system of nonlinear differential equations describing the interaction between a predator and its prey. 5 = 0. Figure 50-8. using information about the population near this equilibrium population. A neutrally stable equilibrium population (called a center) is a borderline case. In the phase plane. we see Fig. The equilibrium population is what we called a saddle point (see Sec. the graph and table at the end of Sec. To prove this. 47). 21). we should not ignore the phase. F = 0 is accomplished by just neglecting the nonlinear terms in which case The fish exponentially grow and the sharks exponentially decrease. F = 0. Before analyzing the exact nonlinear problem. In fact we will show that almost every solution curve is a closed curve and represents a periodic solution (see Sec. We will show the trajectories in the neighborhood of the equilibrium population F = &/A and S = a/c are closed. 46 indicate that the equilibrium population is neutrally stable. and hence the linear analysis does not necessarily yield the same results as that of a complete nonlinear analysis.236 Thus Population Dynamics—Mathematical Ecology Since p = 0. Linearizing the differential equations 50.2 in the vicinity of S = 0. 50-8. Thus.

For a curve which is somewhere spiral-like. To do so it is convenient to exponentiate the above equation. 50-10. For small F. F(tQ) — F0. while for large F. This is an implicit equation relating the population of fish and sharks. Nevertheless we will show these equations represent closed curves. Although we cannot easily sketch F directly as a function of 5". there are values of S for which there are more than two values of F. Z exponentially decays to 0. 50-9. Figure 50-9. and the initial population of sharks. We might guess the completion of both sketches (as marked in the figure in dotted lines). Thus roughly we obtain Fig. We can easily verify these are indeed correct by calculating the first derivatives dZ/dF and dZ/dS and showing each derivative is zero at only one place. Similarly for small S.237 Sec. Let in which case also Let us sketch Z as a function of S and also as a function of F. Z tends towards 0 and for large S. the functional relationship is obtained readily using an auxiliary variable Z. equation 50. This equation is separable: and can be directly integrated. yielding where E is a constant determined from the initial population of fish. A possible difference between closed and open curves is illustrated in Fig. letting eE — EQ>0. . Z algebraically tends towards +°o.3. 50 Qualitative Solution of the Lotka-Volterra Equations return to the study of the phase plane. S(t0) — S0. For each value of E there corresponds one solution curve. Z exponentially tends towards +00. We will show all such curves are closed.

Other values of S yield either 0. Specifically. which yields 2 values of F (also marked •). or 2 values of F as illustrated in the figure. 1. for values of S sufficiently small. Figure 50-12. . there are no values of F. In particular.238 Population Dynamics—Mathematical Ecology Figure 50-10. 50-12 with •. These curves have zero slope at the respective equilibrium populations F = kjX and S — a/c. This determines Z. 50-11. Take one specific value of S marked in Fig. there are still no values of F until S Figure 50-11. As S is increased. Thus we derive Fig.

Then there is only one value of F. no matter which value of E0 is chosen (as long as £"„ is large enough to insure a solution). 50-13. 50-15. Thus for various values of E0 we obtain Fig. The population of fish and sharks are periodic functions of time. the process reverses itself. Figure 50-14 Trajectories of sharks (S). an oscillation similar to the one observed in nature. as we continue to increase S. . Figure 50-15 Lotka-Volterra predator-prey oscillation.239 Sec. 50 Qualitative Solution of the Lotka-Volterra Equations reaches a value at which the first intersection occurs. Figure 50-13. The increase in sharks (predator) lags behind the increase in fish (prey). is sketched in Fig.fish (F) predator-prey ecosystem. F as a function of S. Thus the solution curve. It is a closed curve. there are always exactly two values off. 50-14. As S is increased further. However. Can you show the period is finite? The populations fluctuate periodically around their equilibrium values in a rather complex fashion as roughly sketched in Fig.

50. 2. [Hint: Explain ajb > £/A]. Do not attempt to solve the resulting system. In the phase plane. Reconsider exercise 50. Consider the predator-prey model with 6 = 0 . where FI and Si are the displacements from the equilibrium populations of the fish and sharks respectively.1. 50. . Using the Taylor series for a function of two variables (equation 44. 50. Compare these populations to the ones which occur if b = 0.2.2.] 50. Analyze that equation and determine the conditions under which the equilibrium population is stable.6. Briefly explain the qualitative and quantitative differences between the two cases. Eliminate Ft (from the second equation) to derive a second-order constant coefficient differential equation for Si.3. Suppose that the value of k was increased. 50. equation 50. (If needed. b = 0 and b ^ 0. Show that the two are equivalent.5). [Hint: See exercise 50. Calculate all possible equilibrium solutions.4. Do this in two ways: 1. Also sketch the isocline(s) along which dFjdS = oo.2 (b ^ 0). (a) What does increasing the value of k correspond to ecologically? (b) How does this aifect the nonzero equilibrium populations of fish and sharks ? (c) Briefly explain this effect ecologically. Determine the constant coefficient linear system of differential equations which governs the small displacements from the equilibrium populations.1.240 Population Dynamics—Mathematical Ecology EXERCISES 50.5 for the answer. Indicate arrows as in Fig. Give an ecological interpretation of the inequality in exercise 50. sketch the isocline(s) along which dFjdS = 0.2.4. equation 50. Using perturbation methods as described in Sec.4 can be put in the form: 50. Explain on your diagram where all possible equilibrium populations are.) Reconsider the predator-prey model of exercise 50.5. sketch all possible cases. Show that your answer to exercise 50. Consider the predator-prey model with bj±Q. 50-3.

1. F = 0. assume that one exists (with both species nonzero) and analyze its linear stability. Sketch the phase plane for the predator-prey ecosystem described in exercise 50. F = a/b. Describe the predator-prey interaction. On the other hand if there are many prey for each predator. what do you expect to happen? (b) Sketch the solution curves in the phase plane. [Hint: Use the behavior of the phase plane in the neighborhood of all equilibrium populations. Answer the same questions.11.10. then the predator will find them and increase. 50.8. (What is the ecological significance of this population?) (c) Sketch the phase plane in the neighborhood of S = 0.] An alternate predator-prey model was suggested by Leslie: 50. Suppose that the equilibrium number of prey without predators ajb is smaller than the prey necessary to sustain the predators kjL Sketch the solution in the phase plane. Reconsider exercise 50. 50. 50.1. 50 Qualitative Solution of the Lotka-Volterra Equations 50.12. Reconsider exercise 49. with c = 0. Do you have any objections to this model ? Compare this model to the Lotka-Volterra model. (c) Describe the agreement between parts (a) and (b). However.2.14. (a) Sketch the phase plane in the neighborhood of the equilibrium population in which both populations are nonzero.9. Reconsider exercise 50. Without explicitly determining the equilibrium population. (b) Sketch the phase plane in the neighborhood of S = 0. Consider the predator-prey model.13. Show that if there are many predators for each prey.12. equation 50.4.7.241 Sec. Refer to exercise 50. 50. the predators change in a different manner. (d) Use the information gained from parts (a)-(c) to sketch the entire phase plane. 50. What are all possible equilibrium populations? Consider the following two-species population growth model: How does this model differ from equation 50.9 with A = 0 instead of c = 0.2. if alb = k/L . The equation for the prey Fis the same. (a) Without solving the differential equations. Sketch typical time dependence of predators and prey. 50. then the predators cannot cope with the excessive competition for their prey and die off.

Describe the fluctuations of both species when they are near the equilibrium population in which both species are nonvanishing.17.16. yields If this equation is integrated from the initial time 10 to some arbitrary time. that is let t = tlt where ..242 Population Dynamics—Mathematical Ecology 50. 47 may be of some help. [Hint: The results of Sec. Compare your results to the predictions in Sec. 51. [For a thorough discussion of this problem. then the result is Suppose we integrate over a complete period of oscillation.5. Formulate a discrete time model (i. 50. Consider the time-dependent equations for the predator S.15. Sketch the trajectories in the phase plane in the neighborhood of Consider the two qualitatively different cases. Average Populations of Predators and Preys For the mathematical model of predators and preys analyzed in Sec.e. Briefly explain the mathematical reasoning you use to conclude that the natural position of the nonlinear pendulum is a stable equilibrium position. involving difference equations) for a predator-prey interaction. Determine all possible equilibrium populations. the populations of both species periodically oscillate around their equilibrium values. Dividing by S.] 50. Maynard Smith (see p.] 50. 256 for date and publisher). Refer to exercise 50. see Mathematical Ideas in Biology by J. It is interesting to note that the average value of the populations of the predators can be easily calculated (as can the average value of the preys). 50.

2.1 at t = tl that Thus. Thus Show that for the predator-prey system. a/c. it follows by evaluating equation 51. independent of initial conditions). the average value of the prey population is k/Ji. the same as the equilibrium population of the prey. No matter which trajectory this ecological system traverses (i. the average populations remain the same. For the ecosystem described by equation 50.243 Sec. Note that the left-hand side is an average value. EXERCISES 51. 51 Average Populations of Predators and Preys Figure 51 -1 Period 7" of oscillation of a predator S(t). it could be shown that the average population of the predator is identical with its equilibrium population. The notation of a bar over a quantity often denotes the average. The period of oscillation Tis indicated in Fig.1. show that the average population of the predator is the same as the equilibrium population of the predator.e. Consequently. equation 50.2..2. 51. 51-1. Since the population is periodic. Similarly. .

Alternatively. If this were a fox-rabbit ecosystem the suggestion might involve using animal traps.244 Population Dynamics—Mathematical Ecology 52. The large arrow represents the instantaneous addition of more predators. Suppose that "x" marks in Fig. 51). the result of this may be to increase the magnitude of the oscillation.2). Can the situation be improved? Someone suggests that introducing more predators 5" will reduce the number of prey. the average populations of the predator and prey equal their equilibrium populations. for two species of insects which interact as predator and prey. As indicated. Man's Influence on Predator-Prey Ecosystems Suppose we observe a trajectory (see Fig. If we consider a predator-prey ecosystem in which the prey is considered to be undesirable. In any event. and hence actually remain the same (see Sec. On the other hand. Thus the new mathematical model would be . Figure 52-1 Predator-prey ecosystem: effect of instantaneously introducing additional predators. the use of an insecticide would have the same effect. let us analyze that suggestion using the mathematical model. Although at first that seems reasonable. 52-1 the position in the fluctuation cycle at which more predators S are introduced into the ecological model. Mathematically we can formulate this new problem by supposing that both predator and prey are eliminated in proportion to their number. if the prey were rabbits and the predator foxes. 50-7) of the predator-prey ecosystem (equation 50. then it is possible that the number of prey Fis so large as to upset some people (for example. it is suggested to gradually eliminate some of both the predator and prey. then farmers may be dissatisfied with the large number of rabbits). No improvement has occurred.

Thus this model predicts that an observed oscillation of many periods duration is due to some earlier event. The equivalent parameter k is increased. Since the average population of the predator is a'/c and the average of the prey is fc'/A. for example. Although this seems reasonable for the sharks and fish of the Adriatic. 53 Limitations of the Lotka-Volterra Equation where al and <72 represent the possibly different additional death rates because of. the animal traps. Thus a predator-prey model is sought which perhaps indicates oscillation of a type inherent to the system rather than determined by initial conditions (see a limit cycle. as it is equivalent to the original predator-prey system with slightly alterred parameters. It is not necessary to solve this system of equations. Let Therefore.245 Sec. We have not done so here. setting traps will only yield more rabbits on the average. for example. of a pesticide is to decrease both species' interaction-free growth rates. as one should compare detailed observational data with the predictions of this predator-prey model. it is entirely unbelievable that lynx and hares should oscillate in a manner determined by events a hundred (or more) years ago. Thus the effect. Does this verify our model? Not really. for example. A predator-prey model which results . it is concluded that the predators are decreased and the preys increase Can you explain why this happens ? Thus. Limitations of the Lotka-Volterra Equation We have now indicated that the predator-prey model has certain qualitative features that have been observed in nature. 50-6). There are some difficulties with the previously developed predator-prey model. Try telling some farmer that! A predatorprey ecosystem illustrates the very delicate balance of nature! 53. One objection is based on the fact that the model predicts an oscillation whose amplitude strongly depends on the initial conditions (similar to a linear problem). Fig.

sometimes ecologists attempt to reproduce in the laboratory small scale ecosystems. The deterministic predator-prey population growth models do not result in any probability that one (or both) species becomes extinct. as though nature seeks to restore predators and preys to an ecological balance. Such stochastic predator-prey models will not be discussed in this text. there is a Figure 53-1 Wild predator-prey oscillation.1. Predator-prey laboratory experiments usually eventually result in the predator eating all the prey. During a wildly oscillatory predator-prey fluctuation.2. The process of mathematical modeling continues as more is learned about nature. what eventually happens to the predators? . They attempt to model situations using laboratory species in a way similar to which the mathematician models using mathematics. More realistic models incorporate random processes.246 Population Dynamics—Mathematical Ecology in such a feature is described in an exercise in Sec. 53. EXERCISES 53. More realistic mathematical models of predator-prey ecosystems involve both phenomena. Neither will predator-prey models which exhibit delays be discussed. 50. Then (of course) the predators also die out. 53-1. This is best indicated by briefly mentioning some experimental results. If the prey become extinct. It is known that predators do not instantaneously react to an increasing or decreasing of the prey population. There are other objections to the predator-prey models as developed here. Both species become extinct. As opposed to direct observation of nature. it will become extinct. finite probability that when the population of prey is low. Formulate predator-prey equations which incorporate the above delay. indicated by the trajectory in the phase plane shown in Fig. Other predator-prey models are also discussed in the exercises in which the amplitudes of oscillation decay as time goes on. Thus it is possible that the prey does not recover to dominate in numbers over the predator at some later time.

We assume the effect of competition is to reduce each species' growth rate by an amount proportional to the other species' population.1 to 50. indicating that the respective growth rates are inhibited by both populations in a linear manner. To incorporate their competing for the same food. In the phase plane (especially useful since the above system does not have an explicit solution involving elementary functions). In this section we will investigate a two-species ecosystem in which both species compete for the same limited source of nutrients. These equations of two competing species have been derived using the same modeling techniques as we used for the predator-prey interaction. a logistic growth model is proposed for each in the absence of the other. Two Competing Species Not all species form predator-prey relationships. where it is convenient to call one species x and the other y. Since the food source (or other nutrient) is limited. we will show that the solutions are entirely different. we again first specify the birth and death processes that take place for each species in the absence of the other. Consequently.247 Sec.1. noting that only certain signs change). The resulting equations appear quite similar (compare equations 54. factors must be included which retard the growth rate of each due to the presence of the other. . 54 Two Competing Species 54. In deriving the governing system of ordinary differential equations. However.

From these diagrams or directly from the time-dependent differential equations. the phase plane differential equation is not separable. 54-1. the axes are again both isoclines and solution curves. The other simple isoclines are and Both of these isoclines are straight lines with positive x and y intercepts. the only equilibrium populations correspond to Figure 54-1 Two competing-species models: simple isoclines. In sketching the phase plane. there exist either three or four equilibrium populations (marked • on diagram). depending on the parameters in the competing-species model. illustrated in Fig. . Geometrically. In the cases sketched on top.248 Population Dynamics—Mathematical Ecology Unlike the predator-prey model. there are four types of configurations.

In these cases it is especially important to investigate the stability of the coexistent equilibrium population in order to see if this is a viable result of two competing species. Instead we will discuss just one case. where only those cases in which the equilibrium populations are both positive are pertinent. the time-dependent equations are and the phase plane equation is In addition to x being the stronger. are greater (for equal populations) than the "self- . then a > k since then if x = y. we assume that one of the two species is more suited for competition. 54 Two Competing Species the extinction of at least one of the populations. Thus. dx/dt > dy/dt. we assume that the interaction between the species is strongly competitive in the sense that the interaction terms. it is possible for there to be an equilibrium population in which both species coexist. For example. These equilibrium populations are thus Note that these populations are less than the equilibrium populations that exist without any interactions. For example. in the cases sketched on the bottom. This equilibrium population is given by the intersection of the two straight lines. Most of the analysis of competingspecies models is left to the exercises. Let us assume both behave in the same manner in the absence of the other. All four cases will not be discussed. XE < a/b is most easily seen geometrically. But.249 Sec. However. if x is the stronger. Consider two competing species that are virtually identical. That is the growth rates and logistic limiting factors are identical. — kxy and — axy.

Alternatively. it could also occur if each species psychologically affected the other adversely. You should be able to do these calculations. this problem corresponds to the inequality in which case an equilibrium population can exist. as shown in Fig. Arrows are introduced to indicate changes in time. This might occur if each species emitted a substance toxic to the other. From the time-dependent equation: (3) In general.250 Population Dynamics—Mathematical Ecology interaction" terms. Let us investigate the stability of that equilibrium population. 54-2. — bx2 and — by2. Since a > k. . Thus a > b and k > b. let us sketch the trajectories. Rather than the calculations of the linearized stability. As a review the steps will be outlined. * Figure 54-2 Equilibrium for the interaction of two nearly identical competing species.

" it is possible for it to become extinct. 54 Two Competing Species separated by an isocline on which separated by an isocline on which Thus. The sketch of the solution curves is improved by noting that the equilibrium points can be classified by the techniques of Sec. indicating that the equilibrium population involving coexistence is unstable for this problem.251 Sec. we have Fig. This is an example of the principle of competitive exclusion. 47 as follows: 1. Only one species can survive. Thus following the arrows.9 it is shown that the trajectories in the neighborhood of the node x = ajb. The straight line trajectories in the neighborhood of the equilibrium population are determined in exercise 54. note that the two equilibrium populations corresponding to the extinction of one of the species are both stable.8. A species eliminating its competitor is a stable node. Thus although species x is "stronger. In exercise 54. one or the other of the populations will become extinct. y = 0 can be of two different (but similar) forms depending on whether . Figure 54-4 sketches solution curves that more clearly illustrate this result. 54-3. Coexistent equilibrium population is a saddle point (always unstable). depending on the initial population of each species. Figure 54-3 Qualitative behavior of trajectories of two competing species. However. 2.

Zero population is an unstable node (isoclines in the neighborhood of the origin are straight lines since a — c—see exercise 54.10. y = mx -\.Q. As a specific example of the competition of two species. indicating that since x is stronger.252 Population Dynamics—Mathematical Ecology Figure 54-4 stronger). In that case. 3. as an approximation. Yeasts produce alcohol such that the resulting alcoholic content is approximately proportional to the total number of yeast .4 becomes It is not too difficult to see that approximate solutions (to leading order) for x and y large are straight lines. The slope is determined by substitution: where in this calculation the intercept Q is ignored since both x and y are large. Two competing species: sketch of trajectories (x is the a\b > 2 or a/b < 2. equation 54. it is "more likely" that x will be the surviving species. consider a famous experiment involving the growth of two types of yeast (x and y) in a medium (in the absence of oxygen). the slope m is certainly greater than 1. Similar behavior in the neighborhood of x — 0.1). y = a/b is described in exercise 54. Furthermore we might wish to more accurately sketch the solution curves if both populations are large. It follows that Since a > k > b.

253 Sec. Thus This is an example of the model we have developed of competing species with k = b and a — d. [Hint: It is separable.1. (e) Sketch the solution in the phase plane. EXERCISES 54. equation 54. 54 Two Competing Species (x + y).2. However. [Hint: dyjdx — ?] (c) Solve the resulting phase plane differential equation. In this problem we wish to approximate the phase plane (for the competing species model) when both species' populations are small. and after solving. Consider the competing species model. Show that the solution curves appear differently for the three cases: 54. by eliminating /. . determine the phase plane differential equation. Sketch the phase plane and the trajectories of both populations if Describe the interaction between the two species in each case.1. Exercise 54. (a) Show that for small populations (b) Without solving this simple system of differential equations. the growth rates of yeast are increasingly retarded as the content of alcohol is increased.6 describes the competition between these two types of yeast.] (d) Show that you can obtain the same result by solving the time-dependent differential equations of part (a).

1. explain in ecological terms. Consider the three-species ecosystem 54.254 Population Dynamics—Mathematical Ecology 54. Suppose b = d = 0.) Consider the competing species model. (a) Give an ecological interpretation to the quantities . (Which species is the predator. (b) Both species compete with each other for the same nutrients. ft > 0.5.7. What are a and ft! Show that a > 0. y = ajk) is 54. show that this represents a predator-prey situation. which yeast has the highest tolerance of alcohol ? Consider equation 54.3. determine whether this equilibrium solution is stable or unstable. (Do NOT solve any differential equations. the interactions of species G with species R and F.6.1. Can you predict the outcome of the competition. y = ajk ?] Formulate one system of differential equations describing all of the following interactions in some region between species x and species y: (a) The nutrients for both species are limited. (b) Show that the system of differential equations describing populations near the equilibrium (x = c/0. (a) If G = 0.1 with b = d — 0. (c) By eliminating either Xi or y\. 54. Consider the competition between two types of yeast described by equation 54. and which is the prey?) (b) If G ^ 0. assuming all coefficients are positive constants.6.4. [Hint: Are there any approximately straight line solutions in the neighborhood of the equilibrium solution x — c/0. (d) Sketch the phase plane solution of equation 54. equation 54. If c\d > ajb. where xt and yt are the displacements from equilibrium. (a) What are all possible equilibrium solutions ? Show that x = c/0. (c) There is a migration (from somewhere else) of species x into the region of interest at the rate W per unit time. 54. y = a/k is an equilibrium population.

4.9.3 and 54. Consider equations 54. What differential equations describe this ecological system ? 55. 54.4.> 0) in the phase plane. Without any additional calculations. and . 54. Two species. 54. Show that the equilibrium population x = ajb. Do not do a linearized stability analysis. (d) Introduce arrows indicating the direction of trajectories (time changes) of this ecosystem.9. Consider (a) Give a brief explanation of each species' ecological behavior. delays.) (b) Determine all possible equilibrium populations. 55 Further Reading in Mathematical Ecology (b) Only using a rough phase plane analysis. The qualitative time changes should be indicated everywhere (F > 0. 47B). (c) In the phase plane. Briefly explain this result using the terminology introduced in part (a). Show that the coexistent equilibrium population given by equation 54. (Account for each term on the right-hand side of the above differential equations. Determine the straight line trajectories in the neighborhood of this equilibrium population. stochastic fluctuations. Show that most trajectories approach the line y = 0 as t —* co if a/b > 2.255 Sec. Determine what line most trajectories approach as / —» oo if a{b < 2. y = a/b if k/b > 2 and if klb < 2.12. show that the coexistent equilibrium population is unstable if alb > cja and eld > afk and stable if a/b < c/a and c/d < a/k.10. Further Reading in Mathematical Ecology Our study of mathematical ecology and population dynamics has been limited to some relatively simple models.5 is a saddle point in the phase plane (see Sec. y = 0 is a stable node (see Sec. are in competition and are the prey of a third species C. 54. spatial dependence. Problems involving more than two species. migration. 47C). Consider exercise 54. draw the isoclines corresponding to the slope of the solution being 0 and co. (e) From the phase plane in part (d). describe the trajectories in the neighborhood of x = 0.11. A and B. 54.3 and 54. briefly explain which (if any) of the equilibrium populations is stable and which unstable.8. Consider equations 54. G .

. N. Mathematical Ideas in Biology. R.256 Population Dynamics—Mathematical Ecology age dependence. PIELOU. E. Princeton. Cambridge: Cambridge University Press. Comparisons of our mathematical models to experiments and observations have been essentially of a qualitative nature only. . Cambridge: Cambridge University Press. Models in Ecology. C. New York: John Wiley and Sons. MAYNARD SMITH.. POOLE. Niw York: McGraw-Hill Book Company. 1973. MAYNARD SMITH. have been treated only briefly. R. I suggest the following books: MAY. if at all. 1974. 1968.J. M. W.. J.. An Introduction to Quantitative Ecology. For those interested in studying further in this fascinating area.. J. Stability and Complexity in Model Ecosystems. An Introduction to Mathematical Ecology. 1974. 1969.: Princeton University Press.

Traffic Flow .

This page intentionally left blank .

where to construct entrances. In particular we will discuss what happens to a line of stopped traffic after a light turns green. here we will only formulate mathematical models that are deterministic. exits. However. Instead we will study some simple problems which have recently received a mathematical formulation. Difficulties in this theory occur when light traffic catches up to heavy traffic.of individual cars. and overpasses. We will attempt to predict these quantities if they are known initially. and other desirable ends. eliminate accidents. However. 71-75). how to develop a progressive traffic light system.56. Thus we will study some methods to solve partial differential equations. density. minimize automobile exhaust pollution. Statistical theories can be developed. how traffic flows along a unidirectional road. 76-77). how long the cycle of traffic lights should be. necessitating the analysis of traffic shocks. We will begin our investigation of traffic problems by discussing the fundamental traffic variables: velocity. we will primarily study traffic situations resulting from the complex interaction of many vehicles. and flow (Sees. In this text we do not propose to formulate (no less solve) all these kinds of traffic problems. Nearly uniform traffic flow is first discussed. the ultimate aim is to understand traffic phenomena in order to eventually make decisions which may alleviate congestion. Traffic problems which may be amenable to a scientific analysis include: where to install traffic lights or stop signs. whether to change a two-way street to a one-way street. 60) and experimental relationships between car velocity and traffic density (Sees. (Sees. Rather than analyzing the behavior. 78-82). Conservation of cars (Sec. enabling the introduction of the concept of a traffic density wave (Sees. In particular. traffic congestion has become especially acute. 66-70). The method of characteristics is developed for nonuniform traffic problems and applied to some examples (Sees. 57-59). maximize flow of traffic. in recent years. 61-64) give a formulation of traffic problems in terms of a nonlinear partial differential equation (Sec. These include the traffic pattern formed by a uniform flow of traffic being 259 . how many lanes to build for a new highway. discontinuities in density. trains or buses). A number of examples with traffic shocks are described (Sees. whether to build the highway or to develop alternate forms of transportation (for example. 65). rather than ordinary differential equations as previously analyzed in the sections on mechanical vibrations and population dynamics. Introduction to Traffic Flow Transportation problems have plagued man long before the advent of the automobile.

This would be the velocity measured at time t by an observer fixed at position x. of each car. for example.h.h.260 Traffic Flow stopped by a red light and the effect of a temporary delay of traffic caused. (60 m.h. for example. With N cars there are N different velocities. each depending on time. (72 k. as shown in Fig. t) implies that at each x and / there is one velocity. This velocity (at jc. Automobile Velocities and a Velocity Field Let us imagine a car moving along a highway. Then highway problems are briefly discussed which incorporate previously neglected exits and entrances (Sees. 57-1.p. = miles per hour. dx0(t)/dt and its acceleration dzx0(t)/dt2. The existence of a velocity field u(x. The most common is to measure the velocity «. In many situations the number of cars is so large that it is difficult to keep track of each car. by an accident. . Thus this model does not allow cars to pass each other (since at the point of passing there simultaneously must be two different velocities). N. 57-2.p. ut(i) i= 1 . Suppose that car 1 moves at 45 m.p. we associate to each point in space (at each time) a unique velocity.h. 57. i).then its velocity is.p. . . Expressing this statement in mathematical terms. the velocity field u(x.p. ut = dxjdt. In a highway situation with many cars each is designated by an xt(i). called a velocity field. . = 88 feet per second % 96£ k. Figure 57-1 Highway (position of cars denoted by xt). f) at the car's position xt(t) must be the car's velocity ut(t).h. % 27 meters per second). . labeled car 1 and car 2.p. at time f) is the velocity of a car at that place (if a car is there at that time).h. Instead of recording the velocity of each individual car. There are two ways to measure velocity. k. The position of the car might refer. to the center of the car. 83-85). = kilometers per hour.)* and *m. as shown in Fig. of course. u(x. If the position of the car is designated x0(t). As an example consider two cars on a highway.

car 2 at 30 m.).p. 57-3 yields each car's path. the velocity u is undefined at most times at a fixed position along the highway. An example of that might be (This expression is dimensionally consistent if the numbers 15 and 30 above are velocities.2 u = 30. u is a function of jc and t. and . on a highway with two cars. Thus Integrating these equations yields each car's position as a function of time. Also assume that car 1 is at x = L > 0 at t = 0. It is not always convenient to use a velocity field unless there are many cars. Sketching these motions on a space-time diagram as in Fig. However. 57 Automobile Velocities and a Velocity Field Figure 57-2. (48 k.) Note that when x = 30/.p. Figure 57-3 A vertical highway is sometimes convenient for sketches. In this way a velocity field can be formed.h. then from equation 57. dx/dt. Suppose that a continuous velocity field defined everywhere (for t > 0 and x > 0) existed.261 Sec. because then the slope of a car's trajectory. is its velocity.h. while car 2 is at Jt = 0 at / = 0.

If the car labeled 0 moves at the velocity 30 + 15 ft.2. Clearly. as illustrated in Fig. The velocity field of equation 57. as indicated in Fig. This velocity field is one of many with this property. . To emphasize this concept Figure 57-5 Constant velocity field. the velocity field we use to approximate this situation is the same constant V0. i) = V0.262 Traffic Flow when x = 45t + L. then u = 45. dx/dt = 30 + 15 ft. Figure 57-4. and the initial positions range continuously (and linearly) from 0 to L. then the cars' velocities as ft tends from 0 to 1 range continuously (and linearly) from 30 to 45. The concept of fields is common in many areas. let us assume that there are an infinite number of cars (of zero length) each labeled with a number ft. A simpler example of a velocity field occurs if each car in a stream of traffic moves along at a constant velocity K 0 .1 we show that the solution of this system of differential equations yields the velocity field given above in equation 57. 57-5. x(Q) = ftL. and starts at t = 0 at position ftL.2 was developed in the following manner. u(x. 57-4. As a simple model. In exercise 57. Let ft = 0 correspond to the first car on the left and ft = 1 correspond to the first car on the right.

plotted) is sketched in Fig. (Let ft = 0 correspond to the first car on the left and ft = 1 to the first car on the . Assume that there are an infinite number of cars (of infinitesimal length) on a roadway each labeled with a number ft ranging from ft = 0 to ft = 1. y. t). 57 Automobile Velocities and a Velocity Field we briefly discuss one example that is familiar to you. EXERCISES 57. However. Temperature is often given as a function of position and time. Depending on the particular traffic application.1. we could be interested in either the velocity field or the velocities of individual cars. rather than to associate a temperature (possibly changing in time) with each air particle moving around in the atmosphere. We are normally interested in the temperature field since we do not move about with air particles. A typical weather map (with isotherms. constant values of temperature. T(x. We repeat that these velocities are equal in the sense that both concepts of velocity are used in discussing traffic flow. It is usual to specify the temperature at fixed positions and at a fixed time.263 Sec. 57-6. a person aboard a freely floating balloon might be more interested in the temperature at a moving position! Figure 57-6 Two-dimensional temperature field.

Determine a velocity field satisfying all the following properties: (a) at x = 30?. It is separable. Assume the car labeled ft starts from x = ft (ft > 0) with zero velocity.) Suppose that the car labeled ft moves at the velocity 30 + 15ft (dxfdt = 30 + 15ft) and starts at / = 0 at the position ftL (x(0) = ftL). (a) Sketch curves in x-t space along which u(x. but the car does not move at a constant velocity. (a) Determine the position and velocity of each car as a function of time.5.2.u = 45. which starts (at t = 0) at x = LI 57. (c) Consider the car labeled ft. (a) Show that the cars' velocities steadily increase from 30 to 45 as ft ranges from 0 to 1. . 57. and also assume it has a constant acceleration ft. Suppose a velocity field is given: (a) Determine the motion of a car which starts at x = L/2 at t = 0. 57. [Hint: See exercise 57. Suppose that the velocity field u(x. t) is known. t) is known. (c) the velocity field varies continuously (and steadily) from 30 to 45 as x ranges from 30/ to 45/. t) is a constant.] (d) u(x. (d) Sketch curves along which u(x. (c) Determine the velocity field u(x. t). What mathematical problem needs to be solved in order to determine the position of a car at later times.1. (b) atx = 45t+L. each designated by a number ft.4. [Hint: Why does dx/dt = (30* + 30L)/(15/ + L)? Solve this differential equation. show that the trajectory of a car (the position as a function of time) can be sketched by the method of isoclines. If the velocity field u(x. (b) Sketch the path of a typical car. Suppose that u(x. t) is constant along straight lines in the x-t plane. /) = e~'. Show that (d) Eliminate ft from these two equations to "derive" the velocity field 57. t) 96 (15x + 30L)/(15/ + L). (c) In the same x-t space used in part (a).264 Traffic Flow right. t) is constant.3. 57. (b) Determine the time dependence of the position of any car. u = 30. sketch various different car paths. Consider an infinite number of cars.7.] (b) Show that u(x. (b) Show that the cars' initial positions range from 0 to L as ft ranges from Otol.6. 57.

Figure 58-1 hour).9:00 9:00.M. the flow might be different. This quantity is called the traffic flow.M.7:30 7:30-8:00 8:00. Number of cars passing Number of cars passing per hour 7:00. . For example. The observer could compute. we are unable to distinguish variations in the flow that occur over shorter lengths of time. the average number of cars passing per hour (per lane). However. 58-1.9:30 9:30-10:00 10:00-10:30 433 652 594 551 280 141 167 1304 1188 1102 866 560 282 334 In this example. At different positions along the road.265 Sec. Thus the flow also depends on x. and we write q(x.M. Measurements of traffic flow could have been taken over even shorter time intervals. Traffic Flow and Traffic Density What traffic variables could an observer easily measure in addition to car velocities ? An observer fixed at a certain position along the highway could measure the number of cars that passed in a given length of time. Traffic flow. q(t). Thus the flow q depends on time. for example. as a function of time (measured every half By measuring the traffic flow over half-hour intervals. the largest flow of traffic occurred during the period 7: 308:00 in the morning. q. we cannot tell that the period from 7: 45-8 : 00 A. q(t). as shown in Fig. may have had considerably "heavier" traffic than from 7: 30-7: 45 A. 58 Traffic Flow and Traffic Density 58. Suppose the following measurements were taken at one place over half-hour intervals: Time A. t).8:30 8:30.

note that the computed flow fluctuates wildly as a function of time. can be approximated by a continuous function of time as illustrated in Fig. we assume that there exists a measuring interval such that 1. The number of cars (at a fixed time) Figure 58-3. To remedy these difficulties. Another standard traffic measurement occurs at fixed times (rather than at fixed positions as for traffic flow). Figure 58-2 Traffic flow modeled as a continuous function of time. it is short enough so that the variations in the traffic flow are not smoothed over by averaging over too long a period of time.266 Traffic Flow if measurements were made on an extremely short interval of time. 2. for example over 10-second intervals.9 10-19 20-29 30-39 40-49 50-59 0 2 1 4 1 4 Number of cars per hour 0 720 360 1440 360 1440 In these measurements. If such a measuring time exists. . 58-1. Fig. it is long enough so that many cars pass the observer in the measuring interval (eliminating the wild fluctuations). then the step-like curve for the traffic flow. 58-2. then the following data might be observed: ( in Number of\ cars afterseconds 7:00/ 0.

4 kilometers) of roadway at a fixed time. For convenience it is now assumed (as throughout the discussion of traffic flow) that all vehicles have the same length. 58-3. These measurements yield the number of cars in a given length of roadway. the number of cars per mile (kilometer). 58-4. 58 Traffic Flow and Traffic Density between two positions can be counted. number of cars per mile Number of cars i-H iHi if-i* lf-2 23 16 22 8 92 64 88 32 As another example. then the density. then a plausible traffic situation is depicted in Fig. Suppose the distance used in measuring density is very short (y^ mile «e 10 feet «s? 3 meters). A systematic procedure could be used to take into account cars not completely in a given region at a fixed time. is (This result is most easily obtained by considering one mile (kilometer) of cars in this configuration). the word "car" is used loosely to represent any vehicle. for example. as illustrated in Fig. In order to use one unit of length in traffic problems. Figure 58-4 Traffic density equals the inverse of the spacing [p = W + </)]• As with traffic flow. then a typical measurement might be: Distance along road (in miles) Traffic density. Here. 58-5. L is measured in miles (kilometers) rather than feet (meters). Perhaps estimates of fractional cars could be used or perhaps a car is counted only if its center is in the region.267 Sec. L. The measured data (using approximate fractions of cars) would be: . all vehicles are treated the same. by photography. which might be converted into the number of cars per mile (per lane). imagine a situation in which cars are equally spaced. there are difficulties with traffic density if measurements are made on too short an interval. If the distance between cars is d (the distance d + L is called the spacing). a sketch appears in Fig. a quantity called the density of cars p. If traffic density is measured over £ of a mile (.

002-.014 .016 .020 .006-.008 . This measured density is an extremely discontinuous function.004-.002 .010 . Figure 58-6 Spatially discontinuous measurement of traffic density.014-.268 Traffic Flow Figure 58-5 Measuring traffic density using an extremely short interval. 58-6. the real local variability of traffic density would be smoothed out.018-.006 .010-.004 .012 . If we expect to approximate density as a continuous function of x. On the other hand. then we obtain Fig.012-.016-. t) is sketched as a function of position (at a fixed time). if measurements of density are taken only over large distances (for example. then only average densities are computed. we obtain Fig. If the measuring distance is too .022 Number of cars Numbers of cars per mile 0 i 1 i 0 0 0 0 0 4 \ 0 83 333 83 0 0 0 0 333 167 0 If the traffic density p(x. every 5 miles).018 . Distance 0-. 58-7 and then densities must be measured over intervals of distance that are not too small nor too large.008-.020-.

56.1.1.8. 51. 47.9.3. 53. 63. then the density at 50 is defined to be the number of cars between 50 — m/2 and 50 + m/2 divided by the length m (converted to cars per mile). 99. 15.8. 3. then the large variability of the traffic data disguises the desired smooth nature of density. 59.4. with boundaries labeled from 0 to 100.9. 78.1.4.0. then an average density is computed that does not properly take into account variations of density. 58.269 Sec.9. since each mark is -^ of -fa of a mile (or TtfVn °f a m^e). 34.8.4.7. Let us illustrate by an example the significance of the measuring interval. The measuring distance must be large enough so that many cars are contained therein. 6.6. 28. 43. 62. Consider a ^j of a mile segment of a (one-lane) highway. if the measuring distance is too small. If we use a length of highway of m units centered around the 50 mark. 66. 12. let us imagine the cars to have zero length. 61. further subdivided into one hundred smaller intervals of equal length.1. 37. 40. 95.5.2. 14.3.To be more precise let us see how the calculation of density at the 50 mark depends on the measuring interval.5.3. each car illustrated in Fig. 90. 81. 9. 44. For simplicity.46.1.6. 20. 69. 58 Traffic Flow and Traffic Density Figure 58-7.6. 18. 76.6. the traffic density is slightly less than 333 cars per mile. 25.9. Suppose a photograph was taken and from it we determined that cars were located at the following positions: 1. Equivalently. 58-8 as a '•': Figure 58-8 The traffic density "appears" as the density of ink dots.7.9. 87. 16. We will calculate the density at the 50 mark along the highway.6. 54.8.0.0.6. From the data (or diagram) we see that near the 20 mark cars are approximately 1^-2 intervals apart.9.3.84. Cars near the 50 mark are slightly greater than 1£ units apart and hence the traffic density is slightly less than 1 car per 1£ units of length. 31. 65.9. On the other hand.6.3.1. 72. Further along the roadway cars spread out to about 3 units apart (near the 30 mark) before again becoming close together between the 45 and 65 mark (the distances apart being about l£-2 units apart). large.7.5.2. but small enough so that variations in densities can be measured. 21.1. 23. 49.8. On this basis we set up a chart recording the measured density: . 60.

m Interval of interest Number of cars Number of cars per interval Traffic density: number of cars per mile (number of cars per interval x 500) .5 2 3 4 5 6 7 8 10 12 14 16 18 20 30 40 50 60 70 80 90 100 49| 49i 49* 49 48^ 48 4?i 47 46£ 46 45 44 43 42 41 40 35 30 25 20 15 10 5 0 50* 50i 50| 51 51i 52 52i 53 53i 54 55 56 57 58 59 60 65 70 75 80 85 90 95 100 0 0 1 1 1 1 1 2 3 3 4 5 6 7 9 9 10 12 18 22 25 30 35 38 41 45 1 i i i 1 i * ! ft i7! ft 'ft tf ii i! n ft *9 W il *i iVu 0 500 333 250 167 250 300 250 280 312 300 292 321 281 278 300 300 275 250 250 250 237 228 225 .5 1.0 1.Length of interval.

many cars should be contained in the measuring interval. Three cars is too few. but to get 9 cars requires a fairly long distance. yielding a measured density between 275 and 325 cars per mile. 58 Traffic Flow and Traffic Density This is more strikingly illustrated by sketching the traffic density as a function of the measuring interval. The density then gradually decreases (as m increases) again until the next car is located. as in Fig. eventually a car is found and the average density dramatically increases. For short measuring intervals.271 Sec. For a meaningful calculation. 58-9. The amplitude of the fluctuation decreases as the measuring interval gets longer. As m is increased. violent fluctuations occur. . An interval between 10 and 20 units long would seem appropriate in this problem. but not so many that the local average seems to be lost. For extremely large measuring distances the Figure 58-9 Rapid variation of density's dependence on measuring length.

density and flow).. [Hint: Consider Pa(x 4. From this example. If at lOm. P = AAx. We will assume measuring intervals exist (in space and time) such that the traffic density and traffic flow are smooth functions of position and time without the local variations being lost as a result of an extremely long measuring interval. an amount much less than that associated with the local traffic conditions near the 50 mark. (a) Show that the probability of there being exactly n vehicles along a highway of length x.h. In this text we will primarily be concerned with traffic situations involving a large number of vehicles. Pn(x). then what is the density of traffic? [You may assume that the average length of a car is approximately 16 feet (5 meters)].272 Traffic Flow average density is approaching 225 cars per mile.] (b) Evaluate and interpret the following quantities: <l)P n (0).A*) and form a differential equation for Pn(x\ (see Sec. . this situation gets worse for less dense traffic conditions). Alternatively. a statistical theory may be formulated. Can you guess where the continuum hypothesis arises in investigating motions of liquids and gases ?) EXERCISES 58. Also assume that the probability of two or more cars in that segment of highway is negligible. 36). in this text we will assume that a local traffic density is a reasonable variable. (16 k. 58. However. We will mostly be interested in the collective behavior of many vehicles (a macroscopic theory). we see this is not always entirely the case (in particular.e.2. Neglecting the possibly rapid variation from car to car of car-following distance and thus utilizing the concept of a smooth traffic density will yield a description of traffic flow problems which hopefully will be a good model.p.h. « ^ 0 . especially when an appropriate measuring interval does not exist. For a reasonable local value of density to exist we must assume that the distance between cars stays approximately the same over a distance that includes many cars.p. Assume that the probability P of exactly one car being located in any fixed short segment of highway of length A* is approximately proportional to the length. satisfies the Poisson distribution. For this reason we will be satisfied with investigating the behavior of average traffic variables (i. We will limit our investigation to deterministic models.) cars are one car-length behind each other.1. This is called the continuum hypothesis. (It is also used in fluid dynamics.

(a) What is the jump in density (i. 59-2.3. as shown in Fig. In T hours each car has moved u0r distance (moving at a constant velocity. What is the flow of cars ? To answer that. travels in r . 59. We will show that there is a close relationship among these three. Since each car Figure 59-1 Constant flow of cars. Distance a car. The density has a discontinuity at distances that first include an additional car. Hence the traffic density does not change. Flow Equals Density Times Velocity In the past sections we have briefly discussed the three fundamental traffic variables: velocity. 58. how large must the measuring interval be such that discontinuities in density are less than 5 percent of the density? How many cars are contained in that measuring distance? (c) Generalize your result to a roadway with a constant density of p0 per mile. 59 Flow Equals Density Times Velocity (c) Calculate the expected number of cars on a highway of length x. and thus the number of cars that pass the observer in i hours is the number of cars in UOT distance. We first consider one of the simplest possible traffic situations. traffic is moving at a constant velocity UQ with a constant density p0. see Fig. 59-1. density. moves at the same speed. consider an observer measuring the traffic flow (the number of cars per hour that pass the observer).273 Sec. Interpret your answer. the distance between cars remains constant. the distance travelled equals the velocity multiplied by the time).e. and flow. Since p0 is the number of cars Figure 59-2 hours. moving at constant velocity u0. Suppose that on some road. the magnitude of the discontinuity) ? (b) If cars are equally spaced (100 per mile). Consider the example discussed in this section in which the density is calculated as a function of the measuring interval..

t). 59. we will show that this is a fundamental law. q(x. q(x0.) such that cars are one car length behind each other.1. between t = 0 and t = tQ. t). what is the traffic flow? Suppose that at position XQ the traffic flow is known. Thus the results for constant u and p do not need modification for nonuniform u(x. then we will still show that An easy way to show this is to consider the number of cars that pass x — x0 in a very small time Af. between t0 and t0 + A?. i. u(x. are related by equation 59. p(x. t).p.1. 59-3. is Although this has been derived from an oversimplified case. The number of cars passing is approximately u(x. will pass the observer. Thus the number of cars per hour which we have called the traffic flow. i)&t p(x. Approximate distance a car travels in At EXERCISES 59.. The traffic flow is given by equation 59. their values at x = x0 and t = t0.e. approximately u(x. For traffic moving at 10 m. velocity u(x.p.. (16 k. /). t) and p(x. Consequently. Calculate the number of cars that pass x0. as shown in Fig. t). and varies with time. density p(x. i. then /JOMOT is the number of cars passing the observer in T hours. Figure 59-3 hours. t) can be approximated by constants. t). . the If the traffic variables depend on x and t. i) and p(x.1.h.e.h. /). t). In that small time the cars have not moved far and hence (if p and u are continuous functions of x and /) u(x. q.274 Traffic Flow per mile and there are MOT miles. the cars that occupy a short space.2. t) and flow q(x. In a small time A/. the three fundamental traffic variables. f)Af.

but do so using these traffic field variables. between x = a and x = b. /). 0)) and the traffic velocity field for all time (u(x. Conservation of the Number of Cars In this section. is measured as a function of time. the traffic variables density. In an experiment the total number of cars that pass a position x0 after / = 0. tj). if a traffic light turns red and shortly later green. Then the motion of each car satisfies the following first order differential equation: Solving this equation (which at least can be accomplished numerically using a computer) would determine the position of each car at later times. suppose we knew the initial traffic density (p(x. By following each car we insisted that the number of cars stays the same. Suppose that the density and the velocity field are known initially for a highway of infinite length.3. Thus. (a) Briefly explain why the curve M(x0. we formulate a deterministic model for traffic flow. then can the pattern of traffic be predicted ? We can consider the two fundamental traffic variables to be p(x. t) is increasing as / increases.275 Sec. as shown in Fig. we could calculate the density (although this calculation may be difficult. However. On some interval of roadway. Let us now instead try to "conserve" cars. velocity. t) and u(x. Consequently at later times. However. the number of cars N is the integral of the traffic density: . Assume this series of points has been smoothed to make a continuous curve. t) (since q — pu was demonstrated in the previous section). M(x0. and flow were introduced so that we wouldn't have to follow individual cars. 60 Conservation of the Number of Cars 59. it would involve deciding what measuring interval to use). the traffic density at future times can be calculated knowing the traffic velocity (and the initial density). 60-1. Can we predict the densities and velocities at future times? For example. We want to determine how the density can be calculated easily if the velocity is known (later we will solve problems in which the velocity also isn't known). (b) What is the traffic flow at t = T ? 6O.

t) and q(b. t)At. t) At.276 Traffic Flow Figure 60-1 Cars entering and leaving a segment of roadway. then f ' q(b. t = tv was near t = t0 and this integral was . and the number increases as a result of cars entering the region at x = a. We can generalize this result to situations in which the number of cars crossing each boundary (the traffic flow q(a. If there are no entrances nor exits on this road. /) is the number crossing at x = b per unit time. dNjdt. equals the number per unit time crossing at x = a (moving to the right) minus the number of cars per unit time crossing (again moving to the right) at x = b. An exact expression is needed for the number of cars crossing at x = b between t = t0 and t = f. which for At small is approximately q(b. The number decreases due to cars leaving the region at x = b. In the approximate derivation. Dividing by Af and taking the limit as At —> 0 again yields equation 60. minus the number crossing at x = b between / + At and /. We improve this last derivation by eliminating the need for using an approximation. Consider the difference between the number of cars in the region at t = t0 and t = ti (these times do not need to be near each other). then clearly the number of cars between x = a and x = b is increasing by 25 cars per hour. N(t + At) — N(i). f)) is not constant in time. or since the number of cars per unit time is the flow q(x. then the changes in the number of cars result from crossings at x = a and x = b only. which for At small is approximately q(a. The rate of change of the number of cars. Thus. If cars are flowing at the rate of 300 cars per hour at x = a. then the number of cars between x = a and x = b might still change in time. f). t) dt is the number crossing at x = b between t = t0 and t = t l f Jt.2. Since q(b. equals the number crossing at x = a between t + At and t. Perhaps this derivation of this important result was not clear to some of you. but flowing at the rate of 275 cars per hour at x = b.. Assuming that no cars are created or destroyed in between. An alternate derivation of this result follows. The difference in number of cars between times t + At and t.

Combining equations 60. it follows that Ja p(x.1 and 60. However. (but slightly more elegantly) take the derivative with respect to tv. it follows that Since tv could be any arbitrary time.3. the number of cars is conserved. without an approximation Divide this expression by ti — t0 and take the limit as tl tends to t0. Equivalently. we obtain From the Fundamental Theorem of Calculus (the theorem that implies that the derivative of the integral of f(x) is f(x) itself).277 Sec. (they are two independent times). Let us assume that the flow of cars approaches zero as x approaches both ±°°.2. Since t0 does not depend on /. consider an extremely long highway which we model by a highway of infinite length. From equation 60. No cars are created or destroyed. This does not mean the number of cars between x = a and x — b is constant (if that were true then (djdt) fb q(a. tl is replaced in notation by t. f0). This law expresses a property of traffic over a finite length of roadway a < x < b. As an example. i) dx = 0 or Integrating this yields . 60 Conservation of the Number of Cars approximated by A/ q(b. /) = q(b.2 is rederived. yields This equation expresses the fact that changes in the number of cars are due only to the flow across the boundary. ?))• Equation 60. an integral conservation law. more concisely.3 is called a conservation law in integralform or. and thus the previously stated result equation 60.

4b is exactly the definition of the derivative of q(a. the full derivative with respect to time in equation 60. Divide by —Aa and take the limit as Aa —> 0: The right-hand side of equation 60. Thus.4b the limit can be performed in two equivalent ways: (a) The integral is the area under the curve p(x.3.3 must be replaced by a partial derivative.278 Traffic Flow which states that the total number of cars is constant for all time. Since Aa is small. 0) were known: The integral conservation law. as shown in Fig. although equation 60.* In all three. In later sections (see Sec.3 assumed the positions x — a and x — b are fixed in time. (d/da)q(a. On the left-hand side of equation 60. (1) Consider the integral conservation of cars over a small interval of highway from x = atox = a + Aa. 77) we will find it necessary to relax these assumptions. x = a and x = b. . t) and u(x. We will do so in three equivalent ways. valid at each position of the roadway. The constant could be evaluated if either the initial number of cars N0 or the initial density p(x. Rough approximations are made. In the first derivation. will be expressed as a local conservation law. t). later results of this section are valid only in regions in which the traffic variables are continuous functions of x and /. which yield the correct result. d/dt means d/dt holding a and b fixed in time. t) between x — a and x = a + Aa. 60-2. as we will shortly improve the derivation. t). t) are continuous functions of x and /. t) with respect to a (properly a partial derivative should be used. equation 60. since the derivation of equation 60. are considered as additional independent variables. since t is fixed). We will find that. The number of cars between a *In these derivations we will assume q(x.4a. p(x. those who are not convinced by our first approximation should be patient. the integral can be approximated by one rectangle. Thus from equation 60. In other words. the endpoints of the segment of the roadway. we investigate a small segment of the roadway. However.2 is always valid.

Thus. Using the definition of N(a.279 Sec. t). t). p(a. t)jda. 60 Conservation of the Number of Cars Figure 60-2. Then. the average number of cars per mile between a and a + Aa is In the limit as Aa —> 0. Consequently we derive from equation 60.5a holds for all values of a. By either method. introduce the function N(x. in which case .t).4b that (b) On the other hand. (2).4b again equals —(d/dt)p(a. equation 60.5a follows. Since equation 60. In a subsequent derivation. again from the Fundamental Theorem of Calculus. /). and a + Aa can be approximated by the length of roadway Aa times the traffic density at x = a. (a) or (b). we show that the error vanishes as Aa —> 0. the number of cars on the roadway between any fixed position . the right-hand side is —dN(a. Thus the left-hand side of equation 60. it is more appropriate to replace a by x.t0 and the variable position jc.

we have shown that must be valid if there are no entrances nor exits along a roadway. dividing by Aa. as was done in derivation (2). It expresses a relationship between traffic density and traffic flow derived by assuming that the number of cars is conserved. let us briefly discuss a different and powerful argument. yielding the equation of conservation of cars. By three equivalent methods. Thus. It is called the equation of conservation of cars. The only function whose integral is zero for all intervals is the zero function.280 Traffic Flow or simply This is a partial differential equation (an equation involving partial derivatives). again b is replaced by x. Equation 60. Equation 60. expressing . that is. and taking the limit as Aa —> 0).6 states that the definite integral of some quantity is always zero for all values of the independently varying limits of the integral. equation 60.5 follows. cars are not created nor destroyed. a < x < b.4a. Now take the partial derivative with respect to b. equation 60. Hence.4a. (2) The equation of conservation of cars can be derived more expeditiously. (This is equivalent to letting b = a + Aa.5 immediately follows by taking the derivative with respect to b.4a. Consider the intergral conservation law. Since b represents any position on the road.5. (3) An alternate derivation for a roadway of finite length (a < x < b) is based on noting that the following relation is clearly valid for the righthand side of equation 60. from equation 60. again equation 60. for any finite segment of highway. Thus. However. It is valid everywhere (all x) and for all time.

60 Conservation of the Number of Cars the conservation of cars. t).281 Sec. one fixed at x = a and the other moving in some prescribed manner.1. we know from Sec. Equation 60. (You may assume that p(x. also depends on t. where q is the vector flow of the quantity whose density is p. Show that the number of cars on the highway at time / is where N0 is the number of cars on the highway at / = 0. .) Show this result *In two or three dimensions using the divergence theorem equation 60. EXERCISES 60. that equation 60. for traffic problems. if p is any local quantity that is conserved (in one spatial dimension) and if q is the flow of that quantity across a boundary (often in physics called the flux).2. then it can be shown. Consider a semi-infinite highway 0 < x < co (with no entrances or exits other than at x = 0). p(x.) Suppose that we are interested in the change in the number of cars N(t) between two observers. In general.7 is valid!* Sometimes experiments must be performed to determine how the flow q depends on other quantities in the problem. using the same arguments we have just developed.7 is valid in many situations having nothing to do with traffic. (Note that the integrand.7 must be replaced by dp/dt + V-q = 0. x = b(t): (a) The derivative of an integral with a variable limit is 60. However. t) —» 0 as x —»• oo. 59 that and thus conservation of cars can be written as a partial differential equation relating the traffic density and the velocity field.

explain whyJaW p(x. Considering the cars as particles. there is no equivalent to Newton's law prescribing the manner in which cars must move. 61. If this was a mechanical system.1 reduces to a partial differential equation for the unknown traffic density. traffic density and car velocity. (a) Without using any mathematics.382 Traffic Flow either by considering lim [N(t + A/) — N(t)]{At or by using the chain rule for derivatives.1 can be used to predict the future traffic density if the initial traffic density is known. We could pursue the solving of this partial differential equation. equation 61. are related by only one equation.4. verify mathematically that part (a) is valid (exercise 60. it is the decisions of individual drivers. If the traffic flow is increasing as x increases (dqjdx > 0).4a. we need to know the particles' velocities. equation 61. equation 61. (c) Interpret the result of part (b) if the moving observer is in a car moving with the traffic.2 may be helpful). 60. t) dx is constant if a and b (not equal to each other) both move with the traffic. rederive equation 60. In this case. conservation of cars. If the velocity field is known.3. (b) Using part (a). However. A Velocity-Density Relationship The two variables. What factors influence individual car velocities? . As an initial value problem. then we would investigate the forces in the system and use Newton's law to study the motion of the particles. The unknown velocity field must be investigated. this appears senseless since we do not know the velocity field. explain physically why the density must be decreasing in time (dpldt < 0). However. (b) Using dpldt = -d/dx (pu) show that A»-»0 60. It is not forces which cause cars to move.1 is analogous to the ordinary differential equation for the position of a mass or the ordinary differential equations developed in population dynamics. (c) Assuming dpldt = —(dldx)(pu).

"On kinematic waves II. 229. J. On the basis of these types of observations. M. and hence the drivers' average speed is not appreciably less than the desired speed. 64 on car-following models). B. speed limits if the driver is law abiding. As the density increases further . A theory of traffic flow on long crowded roads.the velocities of the cars would continue to diminish. as the density increases (that is as there are more and more cars per mile) eventually the presence of the other cars would slow the car down. However. we make a basic simplifying assumption that at any point along the road the velocity of a car depends only on the density of cars. Soc. A... let us describe some observed traffic phenomena with which you are no doubt familar. 317-345 (1955). then the driver of each car has the freedom to do as he or she wishes within certain limitations (i.283 Sec.42-51 (1956). If there are no other cars on the highway (corresponding to very low traffic densities). "Shock waves on the highway." Operations Research 4." Proc. 61 A Velocity-Density Relationship A way to investigate the velocity field is to conduct experiments to understand an individual's response to traffic stimuli.. and thus *LIGHTHILL. As traffic increases to a more moderate level. P. G. Roy. and WHITHAM. Why does a particular driver drive the way he or she does ? We will postpone a discussion of such experiments and the resulting mathematical models (see Sec. . However. Only occasionally will each driver slow down because of the presence of other vehicles. If traffic is sufficiently light. then the car would travel at the maximum speed wmax. I. Lighthill and Whitham* and independently Richardst in the mid-1950s proposed this type of mathematical model of traffic flow. and consequently the average speed of the traffic is lowered. tRiCHARDS. encounters with slower moving vehicles are more numerous. "max is sometimes referred to as the "mean free speed" corresponding to the velocity cars would move at if they were free from interference from other cars. in heavy traffic changing lanes becomes difficult. It is still not difficult to pass slower-moving cars.e. Instead. or certainly the technological limits of car performance).

no intersections. Another such example is reduced speed limits near some schools during school hours.) Consequently. then. An exception occurs when the speed limit at night differs from that during the day. We ignore the possibly erratic behavior of individual drivers. the speed limit need not be constant as it might differ in various sections of the roadway. the speed limit is the same. This maximum density. (Usually. Perhaps it would be more realistic to introduce a stochastic model in which it is proposed that at a certain density some percentage of drivers drive at certain speeds and others . wmax = USL. for examples. where L is the average length of a vehicle. There are many assumptions involved in this hypothesis. usually corresponds to what is called bumper-to-bumper traffic. If a speed limit USL is obeyed. x.284 Traffic Flow At a certain density cars stand still. that is u'(p) < 0. It is not being suggested that the dependence of the velocity on the density is the same for all road conditions. (Cars are observed to come to a stop in dense traffic before cars touch each other. u = u(p). because of a different road curvature and banking). the general type of curve shown in Fig. and so on). t). we will primarily investigate a given stretch of highway with approximately constant properties (same number of lanes. As stated earlier. Figure 61-1 Car velocity diminishes as traffic density increases. Furthermore there may be time dependence in the relationship. though it may not be a very bad approximation. 61-1. a different segment of the same highway might have a different relationship (for example. Thus we will assume u = u(p). It states that every driver drives at the same velocity given the same spacing (density). the curve is steadily decreasing. />mtx. However. u(p. of the effects a police car (either moving or stationary) or weather conditions have on car velocities. but not always.) However. as a result. This is clearly not valid. for example. Thus /jmtx < 1/L. at fixed positions along the highway. relating the two traffic variables (velocity and density) is reasonable.

then u changes instantaneously.) of speed you should stay at least one car length behind the car in front of you. let us nevertheless investigate its implications. Now that we have enumerated a number of objections to the model.p. as for examples. EXERCISES 61.2.p. We will not discuss such stochastic models although some traffic researchers have developed mathematical models along these lines. Find the flow of cars as a function of density. This theory does not take into account that on multilane highways passing is not only permitted but is a quite frequent event.1. if . on one-lane roads. If u = u(p). Perhaps this is an important consideration because your own experience as a passenger or driver tells you to slow down if you observe trouble ahead.p.. but the velocity-density curve is different for those two circumstances. the effects of car passing must be small. In addition u = u(p.e. This model represents a possible first step in the development of a mathematical theory of traffic flow.p.h. for example. 61 A Velocity-Density Relationship drive at slightly different speeds. There is another law that gives a maximum'speed limit (assume this is 50 m.p. Another assumption is that the velocity only depends on the density. then a high speed car as it approaches a slower line of traffic must itself slow down. (80 k.)). travel. These effects can also be incorporated into a more refined mathematical model. Some experiments have indicated that u = u(p) is reasonable while traffic is accelerating and also reasonable while traffic is decelerating.h.1.h.285 Sec. How far would the driver following at u m. exactly one length). (a) Assume the car immediately ahead stops instantaneously. you do not wait to slow down until you actually reach the trouble. In order for this theory to be a good approximation. not. The state laws on following distances. briefly discussed in exercise 61. Thus the model as formulated will not take into account the finite driver reaction time nor the finite response time it takes an automobile engine to change velocity (accelerate or decelerate). x. were developed in order to prescribe spacing between cars such that rear-end collisions could be avoided. on the traffic density a few cars ahead (which may be quite different). or extremely crowded highways.h. /) implies that if p changes. u = u(p). tunnels. (16 k. 61. determine the density of cars as a function of speed (you may assume the average length of a car is 16 feet (5 meters)). Assuming that people obey this law (i. Modifications in the mathematical model to include these kinds of effects also have been attempted.h. Many state laws say that for each 10 m.

such as the number of lanes and the smoothness of the road. do not ignore the length of a car.) (b) The calculation in part (a) may seem somewhat conservative.p.The distance between cars was also *GREENBERG H." Operations Research 7. the driver decelerated at a constant maximum deceleration a? (Do not ignore the length of a car. The data from the Lincoln Tunnel were obtained by two observers (with a machine's assistance) a short distance apart. . assume the first car also decelerates at the same maximum rate a. the results are demonstrated in table form and sketched in Fig. Experimental Observations In this section. we will discuss some experimental evidence which reinforces the qualitative arguments of the previous section. 79-85 (1959).3. are constant. which we assume is valid on a stretch of road for which road variables.* Traffic was measured in the Lincoln Tunnel (a heavily traveled approximately two-mile long tunnel under the Hudson River connecting New Jersey and New York City) and also on the Merritt Parkway (a major limited-access divided highway in Connecticut). The velocity is a function of density u — «(/?). Show that the acceleration of an individual car is given by 62. L.286 Traffic Flow (1) the driver's reaction time was T. Perhaps the best place to observe traffic flow without inhomogeneities would be in tunnels. u(x. Assume that a velocity field. 61.1 corresponds to part (b) if human reaction time is about 1 second and the length of a car is about 16 feet (5 meters). L.) (c) Show that the law described in exercise 61. as is shown in Fig. The following traffic data was obtained by Greenberg. and (2) after then. exists.. Instead. "An Analysis of Traffic Flow. but the driver following still takes time T to react. since cars rarely stop instantaneously.h. 62-2. t). The velocity of each vehicle was obtained by recording the times at which they passed each observer (v = J/(/2 — *i)). bridges. How far back does a car have to be traveling at u m. in order to prevent a rear-end collision ? (Again. or highways. 62-1.

8 26.2 109.h.9 121.1 7.5 10.8 10.2 97.h.9 144.0 106.7 106.5 30. note that four times are known /„„ tatt /*„ /*.0 90.7 8.5 11.) (cars/mile) 32 28 25 23 20 19 17 16 15 14 13 12 11 10 9 8 7 6 34 44 53 60 74 82 88 94 94 96 103 112 108 129 132 139 160 165 38. The distance between the two cars (front of one car to the front of the next) varies from vj.* The data for all cars moving between 14. the distance between cars when the second car passes the first observer is the velocity calculated for the first car multiplied by the time interval measured by the first observer.1 8.3 130.1 123.1 11. Generally the data indicates that increasing the traffic density results in lower car velocities (although there is an exception in the data).2 29. For example.1 106.4 106.2 11.6 9.2 10.2 80. (a and b represent the two cars).0 28. Looking at the raw data (not given here) would suggest how much variation To measure the distance between cars.4 141.) (cars/mile) Merritt Parkway Velocity Density (m. for example. 62 Experimental Observations Lincoln Tunnel Density Velocity (m.4 11.9 28.p.7 30.8 7.3 11.5 35.8 31. In this manner an average density for cars moving at 15 miles per hour can be calculated.8 31.5 101.5 12. . and the average of the distance to the preceding car was calculated.4 20.5 130.1 calculated.5 and 15.3 13.p.5 miles per hour was collected.7 99.287 Sec.6 34.6 16.(tai — t^) to va(tai — /*s). It is assumed that the cars have maintained their speed over the distance d.4 27. and two velocities have been calculated va and v/.3 8.8 31.3 107. The data is recorded in a manner such that the velocity can only depend on the density.

Figure 62-2. velocities and densities were obtained in five-minute intervals and averaged (yielding an average velocity as a function of an average density). . The Merritt Parkway data was accumulated in a different manner.288 Traffic Flow Figure 62-1 Actual velocity-density data. in velocity existed between drivers driving under approximately the same density conditions. There.

Figure 63-1 fie density. We will assume that the road is homogeneous such that the car velocity depends on traffic density and not on the time or position along the road. 63 Traffic Flow In the work to follow u(p) is assumed to be obtained by any similiar type of observation. Calculate a velocity-density curve. and so on) in order to maximize the flow on a given roadway.289 Sec. (b) With the help of a friend. number of toll booths. lane width. 63-1. we have hypothesized (based on many observations) that man's driving habits are such that if p = />max.1. Traffic Flow A traffic engineer might well request traffic-control mechanisms (traffic lights. speed limit. 63. see Fig. The largest flow q = pu would occur if cars were bumper to bumper (p = />mtx) moving at the speed limit (u = «max). stop signs. the flow also only depends on the density. namely zero. EXERCISES 62. We will postpone any attempts to explain in more detail the processes by which drivers travel. do an experiment on a road during light and heavy traffic situations. but furthermore. yielding a minimum traffic flow. then cars would be bumper to bumper and would not move. (a) Briefly explain the two different experimental methods to obtain a velocity-density curve. Clearly this is not safe (is that clear?). Since the traffic flow (number of cars per hour) equals density times velocity. Car velocity depends only on traf- . Such a theory might result in some explanations of the specific shape of the velocity-density curve.

Fig. This shows that a maximum of traffic flow occurs at some density (with a corresponding velocity). Figure 63-3. as demonstrated in Fig. Thus. moving at a velocity of about 19 miles per hour. or 2. if there is no traffic (p = 0). the traffic flow's dependence on density is as illustrated in Fig. The absolute maximum of the flow occurs at the only local maximum. 63-4. The data from the Merritt Parkway. the traffic flow must be positive. 63-2. if the traffic is not moving (M = 0 and thus p — />max). We assume that the flowdensity relationship is as sketched in Fig. The flow may be zero in two significant ways: 1. is inconclusive as the velocity range in which the maximum occurs was not observed. In other words we assume that dq/dp decreases as p increases.290 Traffic Flow The flow thus has certain general properties. . Traffic engineers call the maximum traffic flow the capacity of the road. concave downwards. (d2q/dp2) < 0. in general. 63-2. 63-3. For other values of density (0 < p < /?max). occuring at a density of about 82 cars per mile. This flow-density relationship is sometimes called the Fundamental Diagram of Road Traffic. Figure 63-2 Fundamental Diagram of Road Traffic (flow-density curve). The data from the Lincoln Tunnel indicate a maximum traffic flow of about 1600 vehicles per hour.

and GARDELS. December 1963. as was suggested by the previous theoretical discussion. In addition. this idea resulted in increased traffic performance in the Holland Tunnel as reported in Scientific American. However.e. what might be called bottlenecks. as the density increases and the flow correspondingly increases. These measurements have shown that the maximum traffic flow (the capacity) and the corresponding velocity are both lower in the older tunnels. traffic should somehow be forced to move at a density (and speed) corresponding to maximum traffic flow." Scientific American 209 No. Thus. This indicates that the newer tunnels. permit drivers to go faster at the same traffic density. are moving at a speed less than that associated with maximum flow (i. Assume that the cars. "Vehicular Traffic Flow. with greater lane width..Figure 63-4 Merritt Parkway flow-density data. 291 .. flow-density measurements have been made in different segments of the same tunnel. the cars' velocity remains nearly wmax. wmax. Can you suggest why they occur there ? To improve traffic efficiency. improved lighting. Thus for small densities the flow q can be approximated by umtxP. To illustrate an application of this concept. consider a situation in which many cars are waiting to enter a tunnel. R. K. then the traffic usually travels at the maximum speed. Measurements in the Lincoln Tunnel and in the Holland and Queens-Midtown Tunnels (other tunnels into New York City) show nearly identical linear flow-density relationships for small traffic densities. If the density is almost zero.* "HERMAN. differences among the three tunnels develop in the flow-density curves. 6. pp. A signal which literally stops traffic and then permits it to go (in intervals yielding the density corresponding to maximum flow) would result in an increased flow of cars through that tunnel. momentarily stopping traffic would actually result in an increased flow! Although we have simplified the problem somewhat. These experiments have shown that the capacity of the roadway varies. Typically in tunnels these bottlenecks occur on the upgrade of the tunnel. There are places in which the capacity is lower than elsewhere. and so on. increasing linearly with the density. Even as the density is somewhat increased. as often occurs. 35-43. more cars per mile than is most efficient!).

Some traffic data was compared to a flow-density relationship of the following from: The best fit (in a least-squares sense) occurred for (a) What is the maximum density? (b) What is the maximum velocity? (c) What is the maximum flow? (d) Guess what type of road this is.5.)? At what density and velocity does this maximum flow occur? Will increasing the speed limit increase the road's capacity? 63. 0 (and «'(/>) <. sketch the flow as a function of density (the Fundamental Diagram of Road Traffic).e. (80 k. Assume that u = u(p).h. show that (1) «(/>m. (a) Briefly explain the four basic properties of a velocity-density relationship: (2) «(0) = wmax (3) dujdp < 0 (4) dqfdp decreases as p increases (i. a = «max and ft = —umttx/pmtrL. (b) Assume that the velocity depends on the density in a linear way. d2qldp2 < 0).1. Suppose that drivers accelerate such thai where a is a positive constant. 63. u(p) = a + ftp.. (c) What is the flow as a function of density? Sketch the Fundamental Diagram of Road Traffic. (d) At what density is the flow maximum? What is the corresponding velocity? What is the maximum flow? 63. Show that if u"(p) <.1).292 Traffic Flow EXERCISES 63.x) = 0 63. 0). Show that in order for this to satisfy the properties in part (a).p.p.3. what is a road's capacity if the speed limit is 50 m. If a equals a car's acceleration. Is the minus sign reasonable? Consider exercise 61.7.4.6. then the flow-density curve has only one local maximum. Using the Lincoln Tunnel data of Sec. If cars obey state laws on following distances (refer to exercise 61. 62. . 63.3. 63.h.2.

xn(t). However. This process is modeled by specifying the acceleration at a slightly later time. this equation represents a sys- . let us allow some time delay before the driver reacts to changes in the relative velocity. In order to explain the velocity-density curve. We postulate that an individual car's motion only depends on the car ahead. (d) Show that a is the velocity which corresponds to the road's capacity. then the car following will slow down (and thus A > 0). equation 64. 63. the more the car behind accelerates or decelerates. The resulting flow-density curve fits quite well to the Lincoln Tunnel data. Consider the wth car on a highway. If the car following is going faster than the preceding one. we assume cars cannot pass each other (frequently a rather rash assumption). we assume that a car's acceleration is proportional to the relative velocity. subject to the condition that tf(/*m«x) = 0. what is the capacity of the road ? 64. A measures the sensitivity of the two-car interaction. Steady-State Car-Following Models In this section we will suggest a method for determining the specific observed velocity-density relationships. The larger the relative velocity. Mathematically.1 suggests that acceleration or deceleration occurs instantaneously. Theories of this sort are called car-following models. we can carefully analyze actual drivers making their driving decisions. show that (c) Solve the differential equation in part (b). 64 Steady-State Car-Following Models (a) Physically interpret this equation. where Tis the reaction time. Instead. (b) If u only depends on p and the equation of conservation of cars is valid. (e) Discuss objections to this theory for small densities. In the simplest such model.8.293 Sec. The mathematical model suggested here is motivated by both its being reasonable and by the results of some experiments run for the purposes of developing such a model. If u = «mtx(l — />3//>m»x). As before.

near p = /> max . In this way the following velocity-density relationship is derived. this carfollowing model yields a velocity-density relationship We choose the one arbitrary constant d. i. such that at maximum density (bumper-to-bumper traffic) u = 0. Imagine a steady-state situation in which all cars are equidistant apart. it predicts an infinite velocity at zero Figure 64-1 Steady-state car-following model: velocity-density relationship. yields an equation relating the velocity of cars at a later time to the distance between cars. sketched in Fig. 64-1. Thus and hence letting dn — d Since is a reasonable definition of traffic density (see equation 58.2. called a system of delaydifferential equations. . In other words.. How does this compare with experimental observations of velocity-density relationships ? Equation 64. Integrating equation 64.e.4 appears reasonable for large densities. However. and hence moving at the same velocity.1).294 Traffic Flow tern of ordinary differential equations with a delay.

2 which is linear.p.295 Sec. Thus We choose the critical density pc such that the velocity is a continuous function of density. perhaps u is only limited by the speed limit. The flow q = pu is thus A comparison with the experimental data of Sec. We can eliminate this problem. Figure 64-2 Speed limit cut-off of car-following velocity-density curve. one 500 feet (160 meters) behind but the other only 25 feet (8 meters) behind.) faster than the car ahead.p.h. The simplest way to model this is to let the sensitivity be inversely proportional to the distance Thus the revised car-following model is a nonlinear car-following model. 64 Steady-State Car-Following Models density. w = w max . Imagine two drivers each driving 20 m. . Thus we can hypothesize that equation 64. For small densities. At small densities. Let us attempt one improvement. the more likely the driver is to respond strongly to an observed relative velocity. 64-2. We know the drivers will decelerate quite differently. as shown in Fig.h. The closer the driver is. Instead it is more likely that the speed limit influences a car's velocity (and acceleration) at small densities. A driver's acceleration or deceleration also depends on the distance to the preceding car. as opposed to equation 64. (32 k. 62 shows only a moderate agreement. by noting that this model is not appropriate for small densities for the following reasons.4 is valid only for large densities. the changes in speed of a car are not due to the car in front.

2. In that way sketched in Fig. . 64-3. This formula can be shown to agree quite well with the observed data. the velocity is zero. equation 64. with a response time T (a delay). Many other types of similar car-following theories have been formulated by traffic researchers. in which case Once more the integration constant is chosen such that at maximum density. They help to explain the relationship between the individual action of single drivers and their collective behavior described by the velocity-density curve.296 Traffic Flow Again this equation can be integrated yielding Let us consider a steady-state situation. The constant c has a simple interpretation. Difficulties as p —» 0 are again avoided by assuming Figure 64-3 Nonlinear steady-state car-following model: velocity-density relationship. The constant c is chosen (perhaps by a least-squares fit) so that the formula agrees well with observed data on a given highway. Consider the linear car-following model. u = wmax. that for low densities.1. EXERCISES 64. namely we will now show it is the velocity corresponding to the maximum flow: imply that the maximum traffic flow occurs at p = pmllje. in which case the velocity at the maximum flow is c.

equation 64. decreases with n if 64. 64 Steady-State Car-Following Models (a) Solve for the velocity of the nth car. (d) Show that the above inequality holds for all ft) only if XT < \.297 Sec. corresponds to m — 0.2.5. (e) Conclude that if the product of the sensitivity and the time lag is greater than ^. it is possible for following cars to drive much more erratically than the leader.. m — 0. . (This conclusion can be reached more expeditiously through the use of Laplace transforms. vn = dxjdt. equation 64. Show that (b) Assume the lead driver's velocity varies periodically Also assume the wth driver's velocity varies periodically where fn measures the amplification or decay which occurs.3. / = 1. Show that these solutions are exponentially damped if 1/e > A r > 0. General car-following models of the following form can be considered: Note that the linear model. Look for homogeneous solutions (v0(t) = 0) of the form en.2. with a sufficiently long time lag). (c) Show the magnitude of the amplification factor/.1. (a) Solve for the velocity of the nth car. Show that where/0 = 1. decaying. equation 64.e. / = 0 and the inverse-spacing model. va = dxjdt. or constant amplitude? As with exercise 64. the use of Laplace transforms simplifies the above calculations.) Consider the linear car-following model. Thus show that (b) Consider two cars only. Thus 64. (c) For what values of Ardo solutions exist of the form en with r complex. such that the solutions are oscillatory with growing.2. with a response time T (a delay). the leader and the follower. In this case we say the model predicts instability if A r > ^ (i.

Show that the two car-following models described in this section satisfy dqldp < 0. the drivers would adjust their velocities immediately. Different cars will move at different velocities (since the spacing is nonuniform).8) implies This is a partial differential equation in one unknown variable p. then conservation of cars (equation 60. However. Partial Differential Equations For a given segment of a highway. Show that the assumption that w(/?max) = 0 does not yield a reasonable velocitydensity curve. Suppose a nonconstant initial traffic density existed. and.1.4. experiments can be run to analyze the density dependence of the velocity. Instead. or equivalently Figure 65-1 Nonuniform traffic. under the steady-state hypothesis ? (b) Consider the case proposed by Edie.5. and briefly discuss the resulting model. . equation 65.298 Traffic Flow (a) What velocity-density curves are implied by the above general model. assume w(0) = w max . If we were interested in the density of cars at a later time we would "just" need to solve the partial differential equation. / = 2.] 65. if you wish for simplicity you may assume that pmaK = 225 cars/mile and also assume that the theoretical curve exactly satisfies the first data point. Thus the density will change immediately. 64. under our assumptions. Which theory best fits the data ? [Hint: The parameters of the model are usually chosen by making a least-squares fit to the data.h. m — 1. determined by u = u(p). as shown in Fig. 65-1. M = 32 m.p. 62). Compare the two theoretical models of this section to the Lincoln Tunnel data (see Sec. If we assume that under all circumstances the driver's velocity is a known function of p. 64. This process would continue. The traffic problem has been formulated in terms of one partial differential equation. when p = 34 cars/mile.

To illustrate this difference. This last expression is often easier to use since by the chain rule and thus One appropriate condition in order to solve uniquely the partial differential equation is an initial condition. then the first two would be ordinary differential equations. With an «th order ordinary differential equation. q can be considered a function of/? only. if X°) = Po>then However. .3 only one initial condition is needed since the partial differential equation only involves one time derivative. for example. However. there are some major differences between ordinary and partial differential equations due to the additional independent variable. n initial conditions are needed. one initial condition is needed. The number of conditions are correspondingly the same for partial differential equations.299 Sec. the general solutions being: To have a unique solution. If p only depends on t. now we assume (as originally proposed) that p depends on both x and /. Thus for equation 65. 65 Partial Differential Equations since q — pu. let us consider three extremely simple first order partial differential equations: These are called partial differential equations because p is assumed to depend on x and t (even though there is no explicit appearance of x in either of the first two equations).

0) = f(x) ? In this case it is quite simple as c0(x) = f(x). The arbitrary constant now depends on x in an arbitrary way. p(x. 0). However. the constant can depend on x in an arbitrary way.1. p is a constant for each fixed x. where c0(x) is an arbitrary function of x As a check. Hence the constant is an arbitrary function of x. and hence the solution of problem (2) satisfying the given initial condition is For example (3). 0) — /(x). To determine the arbitrary function. in which case we quickly can verify that p —• c0(x) is the solution. t) = c0(x). one initial condition is needed (corresponding to the one initial condition for the ordinary differential equation). 0) being prescribed. Can the partial differential equation be solved for any given initial condition. can the arbitrary function. The partial differential equation can again be integrated yielding (for each fixed x) As before. p = c0(x) is substituted into the partial differential equation. We now consider example (2). p(x. for different x's different constants could result. Hence The initial condition is satisfied if f(x) = c^x) -f. be determined such that initially p(x. dp/dt = 0. dpfdt = 0. Thus if x is kept fixed. the initial traffic density p(x. 0) = f(x) ? Equivalently. that is for p(x. . In general arbitrary constants of integration become arbitrary functions when the integration is of a partial derivative. Thus is the general solution of dpfdt = 0. Again we will satisfy the initial condition p(x. keeping x fixed (as implied by dfdt) yields the solution of the ordinary differential equation. By integrating.300 Traffic Flow In (1). The partial derivative means keeping x fixed. t). Thus solves the partial differential equation and simultaneously satisfies the initial condition. p doesn't change with t. The initial condition is the initial value of p(x.

Is there a solution of dpldt = — xzp. t) = cos / for t > 0? 65. 0) = sin x. Although we will be able to solve this partial differ- . Determine the solution of dpldt = p* which satisfies p(x. 65. 0) = /(*). which satisfies p(x.3.301 Sec. Determine the solution of dpldt = xtp which satisfies p(x. 0) = cos x for x > 0 and p(0. 65. In summary we have been able to solve partial differential equations in the case in which they can be integrated. This partial differential equation cannot be directly integrated as could the simple examples in the previous section. Linearization The partial differential equation which was formulated to mathematically model traffic flow is or equivalently One possible initial condition is to prescribe the initial traffic density We will solve this problem. /) = 1 + sin x along x = — 2/. 66 Linearization For other values of x the constant may vary.5. EXERCISES 65. since both dpldt and dp/dx appear in the equation. 65. The arbitrary constants that appear are replaced by arbitrary functions of the "other" independent variable.2.4. is satisfied if c3(x) = f(x). 0) = cos x. and hence the solution of the partial differential equation is The initial condition. p(x.1. 66. that is determine the traffic density at all future times. Determine the solution of dpldt = p. Determine the solution of dpldt — (sin x)p which satisfies p(x. yielding the solution of the initial value problem. 0) = f ( x ) . such that both p(x.

satisfies the partial differential equation 66. . 0) = f(x). This is verified by noting that a constant density. namely one in which the density is nearly constant. What does your experience tell you? What kinds of phenomena do you observe? Does the density seem to stay constant? Some of you have probably had the experience of driving at a steady speed and all of a sudden. You must slow down. Let us imagine driving in a traffic situation in which the density is approximately constant. yields Thus to leading order (that is neglecting the small terms) the following equation is obtained: *An equilibrium solution of a partial differential equation is a solution which does not depend on time. then there should be an approximate solution to the partial differential equation such that where \€pl\<^p0. and so on.t) is called the perturbed traffic density (or the displacement from the constant density pQ). If the density is nearly uniform. t). Expanding that expression via a Taylor series. Let us investigate that situation. for no apparent reason. Any constant density is an equilibrium* density. If the initial traffic density is a constant.2 into the second form of the partial differential equation 66. for now let us first discuss a simpler problem. nearly equaling the constant p0. €pi(x.1. Substituting equation 66. Ib yields where a power of 6 has been cancelled. Thus the perturbed traffic density is also known initially. then the density should remain constant (since all cars must move at the same speed). independent of A:. the car in front slows down. The derivative dqjdp is evaluated at the total traffic density p0 -f epi(x. Assume that the initial density is a known function of x. then the car behind slows down.302 Traffic Flow ential equation. p^x.

equation 66.3 ? 67.t) satisfy? Is it the same as equation 66. 67 A Linear Partial Differential Equation This partial differential equation governs the perturbed traffic density.1. (dq/dp)(p0). (d) By substituting into the partial differential equation of part (a) and by neglecting nonlinear terms. what equation does Pi(x.3 is a linear partial differential equation while the exact traffic equation 66. where c = (dq/dp)(po). However. The coefficient that appears in equation 66.3. A Linear Part/a/ Differentia/ Equation In this section. Assume that dqjdp = a + bp (a) What nonlinear partial differential equation describes conservation of cars? (b) Describe ill possible (time-independent) traffic densities p(x) which satisfy the equation of part (a).303 Sec. This resulting partial differential equation is the simplest kind involving both partial derivatives. . EXERCISES 66. is a constant (the slope of the flow as a function of density evaluated at the constant density). we will solve the partial differential equation corresponding to the linearization of the traffic flow problem.1 is a nonlinear partial differential equation. (c) Show that p = pQ (any constant) is a time-independent solution.

Although there seems no motivation to do so. it will depend on x' and t. distance ct (since it is moving at a constant velocity c) as shown in Fig.304 Traffic Flow where c is a constant. it is more convenient . let us introduce a new spatial coordinate x' moving with the constant velocity c. if x' — 0. However. then x = ct. c = (dq/dp)(po). the moving coordinate system has moved a Figure 67-1. At a later time t. x = ct + x' or We will first investigate what happens to the partial differential equation describing linearized traffic flow in this moving coordinate system. c has the dimensions of a velocity (why?). Furthermore at x'. Instead of the solution depending on x and t. 67-1. 67-2. Thus. in making changes of variables involving partial derivatives. Figure 67-2 Frame of reference x' moving at velocity c. Let us assume that the two spatial coordinate systems x and x' have the same origin at t = 0: See Fig. The initial condition is that the initial perturbed traffic is known. It will be shown to be a very important velocity.

In this manner. The reason for this is clear from the definitions of these two partial derivatives. we use the variables x' and t'. in a stationary coordinate system). If x' is fixed. For different values of *'. Thus for the moving coordinate system. 67 A Linear Partial Differential Equation to introduce different notations for each set of variables.305 Sec. Consequently. while d/dt' means the time derivative keeping x' fixed (that is. the chain rule of partial derivatives is used: and thus Note that even though t' = t. /?. is constant. pl may be a different . 65). in a coordinate system moving with velocity c). it enables us to have a notational distinction between keeping x fixed and keeping x' fixed. The changes in time may be different in the two systems. This emphasizes the importance of introducing a new time variable t'. where t' = t. this partial differential equation in a coordinate system moving with velocity c becomes which simplifies and becomes This partial differential equation can be directly integrated (see Sec. d/dt means the time derivative keeping x fixed (that is. that is p^ is constant in time in a coordinate system moving with velocity c. (djdt') ^ (d/dt). the change of variables we use is In order to express the partial differential equation in terms of the new variables.

it can be integrated (in a coordinate system moving with velocity c). Note that this velocity may be different from the velocity at which an individual car moves. the density stays the same. In the original variables. The general solution to equation 67. 83).1 is satisfied by equation 67. 0) = /(. Consequently.306 Traffic Flow constant. p^ is a function of x'. To again verify that this really is the solution. These lines are called characteristics* of the partial differential equation. The density is said to propagate as a wave (called a density wave) with wave speed c. Even though equation 67. In Sec. where g(x') is an arbitrary function of x'.2. the density stays the same.x). Using the chain rule and Thus it is verified that equation 67.1 involves both partial derivatives with respect to x and t. The essential property of characteristics is that along those curves the partial differential equation reduces to an ordinary differential equation. . the solution of the partial differential equation satisfying the initial condition is or equivalently If we move with the velocity c. we will not find it necessary to use the change of variables formula for partial derivatives. Curves along which p is constant are not always characteristics (see Sec.1 contains an arbitrary function.1. 69 we will derive this result in an easier way. Along the curves x — ct = constant. 65. we substitute it back into the partial differential equation 67. Can the arbitrary function be determined in order to solve the initial condition ? The general solution is but initially Pi(x. just like the examples of Sec. Thus f(x) = g(x).

67-5. but dpjdt and dpjdx = c. 67 A Linear Partial Differential Equation In this case the characteristics are all straight lines with velocity c. the density equals the value it has at t = 0. The sign of the slope is significant as we have indicated that small perturbations to a uniform density move at the . see Fig.307 Sec. Note may Figure 67-3 Characteristics of dp\ldt + c dpi/dx = 0. Likewise dpjdx is not necessarily zero since pl may change keeping t fixed. yields Fig. dx/dt Sketching various characteristics in a space-time diagram. Figure 67-4 Density variations. 67-3 and 67-4 we have assumed c> 0. that pl stays constant along the characteristic. Along each characteristic. What is the sign of c? Recall The Fundamental Diagram of Road Traffic is of the form shown in Fig. 67-4. not be zero. In Figs. Thus the slope is positive for densities less than that corresponding to the capacity of the road and the slope is negative for densities greater than that corresponding to the road capacity. 67-3. As illustrated there dpjdt may not equal zero since keeping x fixed PI may vary.

Consider dp\jdt + c(dp\ldx) = 0.7.5. 67. (b) Sketch p(x) at t = 2. what it means for the velocity of a density wave to be negative! Figure 67-5 Flow-density curve: road capacity. 67. Show that 67.2. where |e| < pQ. . 67-3.4). Suppose that 67. (a) What is the density wave velocity if the traffic density is nearly p0 everywhere? (b) Show that this density wave velocity is less than the car's velocity. EXERCISES 67. Show that c = dqfdp (p0) has the dimensions of a velocity. Suppose initially (/ = 0) that the traffic density is p = p0 + € sin x.1.308 Traffic Flow velocity equal to that slope (equation 67. t). a constant. would you say p = p0. is a stable or unstable equilibrium solution of equation 66.6. 67.4. What is the slope of the straight line characteristics sketched in Fig. Determine p(x. Suppose we observe p\ in a coordinate system moving at velocity v. Does p stay constant moving at the car velocity ? Based on a linear analysis. Suppose initially (/ = 0) that the traffic density is (a) Sketch p(x) at / *= 0. The wave velocity can thus be positive or negative! In the next section we will attempt to describe what a density wave is. in particular.1? 67.3.

68-3. and • indicates a relative minimum or relative maximum of the density. figure68-1 light and haeavy Let us imagine a situation involving heavy traffic of a nearly uniform density. 68 Traffic Density Waves 68. 68-1. 68-2. Using these definitions and the Fundamental Diagram of Road Traffic. see Fig. The previous analysis has shown that the density remains constant if an observer moves with velocity c which is negative. Figure 68-3 Characteristics of dp\ldt + c dpi/dx = 0. Fig. .309 Sec. Thus density is constant along characteristics. while in light traffic the opposite is true. where the dotted line is the approximate constant initial density. we conclude that in heavy traffic the perturbed densities move with a negative velocity. Positions of the relative maxima Figure 68-2 Nearly uniform heavy traffic. Traffic Density Waves For convenience heavy traffic is defined as traffic such that the density is greater than the optimal density (corresponding to the road's capacity) and light traffic such that the density is less than the optimal density. Suppose that the initial density is sketched in Fig. sketched in a space-time diagram.

no car is actually moving backwards. Figure 68-5 Backwards moving density wave. The cars trailing an excess of traffic may slow down before they reach the position of the original bulge of cars due to the "chain reaction. (As an exercise. moves backwards at a constant speed c. The variations in the traffic density appear to be moving backwards. 68-5. Sketching the density p as a function of x and / requires a three-dimensional sketch. t. Of course. . the density wave moves backwards. p). As time continues to increase the density Figure 68-4 Initial traffic density. where c = (dq/dp)(p0). Initially it is as shown in Fig. p vertical and t into the paper) we obtain Fig. A short time T thereafter the density has moved backwards a distance |CT|. Let us sketch the density for different times. 68-6. and hence is not always easy to draw. The cars behind thus move faster than the cars in the bulge and hence it is the car's behind which will move into a region of larger density. How does this occur ? The cars in a bulge of traffic have a larger density and slow down. For this example (with x horizontal. The density has been indicated as staying the same along paths with velocity c. as shown in Fig." Have you ever observed this phenomena? What happens to cars at the front of a bulge of traffic? Do they remain in the bulge? Let us attempt to explain why for heavy traffic of nearly uniform density. 68-4. you should be able to use Figure 68-6 Traffic density: three-dimensional sketch (x.310 Traffic Flow are marked with heavy lines and minima in dotted lines. (c < 0). This density wave travels without a change of shape.

we see that for heavy traffic. illustrated by Fig. than at x = b. and thus the density will be closer to that originally further ahead. the density appears to be moving backwards! A similar analysis would yield the same conclusion for regions in which the density is increasing further ahead along the roadway. 68-7 on the Fundamental Diagram of Road Traffic: Figure 68-7 Flow-density relationship: p0 is heavy traffic. as shown in Fig. Figure 68-8 Traffic is locally decreasing. At which point is the flow larger ? From the Fundamental Diagram of Road Traffic. 68 Traffic Density Waves the arguments that follow to explain why the density wave moves forward for light traffic!) Consider traffic which is approximately the constant p0 sketched in Fig. Since the density at x = a is larger than at x = b. A short time later the density between a and b will decrease. Consider a segment of traffic in which the density is decreasing ahead on the roadway. 68-10. circled in Fig. the traffic flow at x — a is less than at x — b (in general for heavy traffic the flow decreases as the density increases). Thus the integral conservation of cars implies that the number of cars between a and b is decreasing. 68-9. cars move slightly slower at jc = a. Figure 68-9 Traffic is locally increasing. A different common example of this type of illusion is the waves present in a "jumping" rope. 68-8. In some sense it could be said that there is an optical illusion of backward motion. It is possible for the density waves to move backwards while all cars are moving forwards. .311 Sec.

since u — q/p. then the slope of the straight line characteristics are the same as the slope of the appropriate tangent to the Fundamental Diagram of Road Traffic. that is straight lines with velocity c. all the characteristics are straight lines parallel to this tangent. see Fig. We know pl is constant along curves in which Jt — CMS constant. 68-12. there are two important velocities: one the velocity of individual cars and the other the velocity at which a density wave travels. A wave appears to move to the right. equals the density wave velocity corresponding to p near p0. just as there are car velocity and density wave velocity for traffic flow problems. We can clearly show these lines are not the trajectories of cars because cars move at velocity w. Fig. The slope of the straight line from the origin (q = 0. When the rope is vertically moved by the person illustrated above. and hence a car path . as we will show. In traffic flow problems. but we all know that the particles of the jumping rope only move up and down! For a jumping rope there are two different types of velocities. Thus if the traffic density is nearly p0. the wave velocity and the rope velocity.312 Traffic Flow Figure 68-10 Wave on a "jumping" rope. If these characteristics are sketched in a space-time diagram such that slopes have the units of velocity. p = 0) to the point on the flow-density curve representing the constant density p0 is the car velocity u. 68-11. 68-11 can be quite useful in sketching solutions to the partial differential equation Figure 68-11 Fundamental Diagram of Road Traffic: traffic density wave velocity c and car velocity u. a "disturbance" is propagated at first to the right. Straight lines with these two slopes are sketched in Fig. The slope of the Fundamental Diagram of Road Traffic (q(p) versus p)atp = p0.

»/2? Figure 68-13.p. What is the estimated velocity of the density wave corresponding to P= An. 68-12. which equals the slope of the other straight line drawn on the Fundamental Diagram of Road Traffic. 68-13. Consider both cases in which the traffic is getting heavier down the road and lighter.) how fast a car moves at p = An. has slope dx/dt = u. For example.h.x/2.1. Later we may wish to improve this approximation to take into account the effects of the variation of the traffic density. Suppose that q as a function of p has been experimentally determined. The car moves forward. 68 Traffic Density Waves Figure 68-12 Characteristics and a car path. 68. EXERCISES 68. In this situation we have approximated the car's velocity as a constant because the density is approximately constant. . determined from flow-density curve.313 Sec.2. the approximate path of a car moving at velocity w(/?0) starting from x = L is sketched in Fig. Explain why a density wave moves forward for light traffic. while the density wave moves backward. as illustrated in Fig. Estimate (in m.

what is the velocity of a density wave? 69. that is if then Thus />i is constant.2. the same conclusion we reached in Sec. while (dxjdt^dp^dx) represents the change due to the fact that the observer moves into a region of possibly different traffic density. Let us consider the traffic density measured by a moving observer (not necessarily in a car moving in traffic). equation 69. t). in an alternate way. since the chain rule of partial derivatives implies that The first term dpjdt represents the change in the traffic density at the fixed position. equation 69. with the partial differential equation for the perturbed traffic density.314 Traffic Flow 68. In this case. An Interpretation of Traffic Waves In this section.3. the solution to the partial differential equation for the perturbed traffic density. For very light traffic.1. We will also find this concept useful in our study of the fully nonlinear traffic flow . The traffic density measured by the observer depends on time. The rate of change of this density depends both on the variation of the traffic and on the motion of the observer. we know that increased traffic density does not appreciably reduce a car's velocity. Let the position of the observer be prescribed by x = x(t). Pi(x(t). It is apparent that if the observer moves with velocity c. An observer moving with this special speed c would measure no changes in density. 67. Compare this expression for the change in density moving with an observer. we obtain.

Show that for an observer moving with the traffic. However.1. Suppose for graphical purposes that we replace t by y. where a and ft are constants. an algebraic solution is easily obtained. By integrating equation 69. pl can be a different constant. ft = /(a).3. equation 67. Thus the constant ft depends on the constant a.2. and hence equation 69. 70-1. ft is an arbitrary function of a. where u is the car's velocity. explain why the curves along which pl is constant must be parallel to #. pl = ft along x — ct + a. From equation 69. a different constant a). 67-69.1 becomes (a) For what two-dimensional vectors g does g-Vpi = 0? (b) Briefly explain why g is perpendicular to Vpi. Suppose that the initial traffic density is constant for the semi-infinite expressway illustrated in Fig.3. (d) Show that the results of part (c) are in agreement with the results of Sees. How many cars per hour would have to continually enter in order for the traffic flow to remain . obtained by transforming the partial differential equation to a coordinate system moving with the velocity c. 70 A Nearly Uniform Traffic Flow Example equation. For a different straight line (i. (c) Using part (b).e. EXERCISES 69. or which is identical to the result. the rate of change of the measured density is 69. we see that pl is a constant only if x — ct is a constant..3. A Nearly Uniform Traffic Flow Example In this section another type of traffic problem involving a nearly uniform traffic density will be solved.315 Sec. 70.

t) = g(0." as the number of cars in between stays the same assuming constant density. the entrance to the expressway at x = 0. t).) Note that the initial condition is only valid for x > 0 (rather than in the previous sections in which — oo < x < oo). Semi-infinite highway (only entrance at uniform ? The traffic flow at the entrance must be p0u(po). The general solution to the partial differential equation has already been obtained . Thus the flow at x — a must be the same as the flow at the entrance q(a. Since the traffic density is prescribed to be constant. the left hand side is zero. What is the resulting traffic density ? The partial differential equation is the same as before: being derived from The traffic is assumed initially to be uniform. But the flow at x = a is p0u(po). Using the integral conservation of cars.316 Traffic Flow Figure 70-1 x = 0). equation 70. Thus q(Q. t) = />0w(/>0)-In other words. the flow "in" must equal the flow "out.1. suppose that the flow in of cars is slightly different (and varies in time) from that flow necessary for a uniform density. so that the initial condition is (This could be generalized to also include initial densities that vary slightly from the uniform case. The initial condition must be supplemented by the flow condition. the flow corresponding to the uniform density p0. To prove this statement (though to many of you a mathematical proof of this should not be necessary). called a boundary condition since it occurs at the boundary of the roadway. However. with #. consider the interval of roadway between the entrance and the point x = a.(/) known.

The traffic flow is approximated by since c = q'(po). equation 70. i. 70-2. t) is the same as the traffic density at the entrance at a time x/c earlier. Since the perturbed traffic flow is known at the entrance. 70-2. The traffic density at the entrance can be determined since the traffic flow is prescribed there (use equation 70. 70-2.317 Sec. 70 A Nearly Uniform Traffic Flow Example or equivalently Let us use the concepts of characteristics assuming light traffic. the density pl = 0 or the total density p = p0. sketched in Fig.e. in the shaded region in Fig. may be expressed using Taylor series methods. c > 0 (heavy traffic is discussed in the exercises). then . The density p{ is constant along Figure 70-2 Characteristics along which the density is constant. these lines. In this region the traffic density only differs slightly from a uniform density. q(p) = q(p0 + €g). Hence.. Thus the density at the entrance at time t — (x/c) yields the density x miles along the roadway at time t. The characteristics are the lines x — ct = constant. qv(t). What is the density of cars if the density remains the same moving at speed cl From the diagram in Fig. The unshaded region is where on the highway it is noticed that cars are entering at a nonuniform rate. xfc is the time it takes a wave to move a distance x at speed c. Thus the perturbed traffic flow is simply c times the perturbed density. the traffic density at (x. since p = p0 at t = 0.3.1 assuming p is near />„)• The traffic flow.

where 0 < €g(f) < p0. 0) = p0. Suppose initially the traffic gradually thins out from p = p0 + € to p = p0. 70. In this situation what might happen to cars waiting to enter the highway? 70. and hence at position x the information has taken time x/c to travel. but the traffic density was prescribed at the entrance. Assuming nearly uniform.318 Traffic Flow and thus by letting z = — ct Consequently the total car density is given by equation 70. where 0 < € <C Po. EXERCISES 70. being slightly different from p0. but heavy traffic. Assume there is nearly uniform light traffic on a semi-infinite highway. Suppose initially the traffic was exactly uniform p(x. Assume that 0 < ea < Pou(p0)(a) What are the dimensions of € and a? (b) Determine the traffic density at later times. in the following manner: Suppose that the traffic flow is prescribed at the entrance (x = 0).2.3.1.3 as In summary This solution clearly indicates that information (that the traffic is entering at x = 0) is propagated at a velocity c. Determine the traffic density at later times. . show that in general it is impossible to prescribe the traffic flow at the entrance to a semi-infinite highway. Assume that there is nearly uniform light traffic on a semi-infinite highway.

as shown in Fig. it is seen that the density will remain constant from the observer's viewpoint. the velocity at which nearly uniform traffic density waves propagate. if For this to occur the observer must move at the velocity q'(p). this velocity is called the local wave velocity. Nonuniform Traffic— The Method of Characteristics The nonlinear first-order partial differential equation derived from conservation of cars and the Fundamental Diagram of Road Traffic is In the previous sections we considered approximate solutions to this equation in cases in which the density is nearly uniform. Since . 71-1. Since this velocity depends on the density (which may dramatically vary from one section of roadway to another). The density of traffic at the observer changes in time as the observer moves about. If the observer moves at the local wave velocity.319 Sec. 71 Nonuniform Traffic—The Method of Characteristics 71. Again consider an observer moving in some prescribed fashion x(f). or p is a constant. The traffic was shown to vary via density waves. We will find the techniques of nearly uniform traffic density to be of great assistance.1 to equation 71. Thus there exist certain motions for which an observer will measure a constant traffic density.2. By comparing equation 71. then the traffic density will appear constant to that observer.

4 are ordinary differential equations. the slopes (related to the velocities) of different characteristics may be different. equations 71. and thus all the curves (characteristics) were parallel straight lines. However.. but different observers may move at different constant velocities.e. the density is the same density as it is at the position at which the characteristic intersects the initial data. Consider the characteristic which is initially at the position x = a on the highway. the traffic density remains the same. Thus along this one characteristic. p is constant. dp\dt = 0 or p is constant.320 Traffic Flow Figure 71 -1 Path along which the traffic density stays the same. Each characteristic is thus a straight line as in the case of nearly uniform flow. which is a known constant.3 and 71. since they may start with different initial traffic densities. as shown in Fig. Initially p equals the value at x = a (i. at t = 0). In the case of nearly uniform flow. this characteristic Figure 71 -2 Characteristic initially at x = a. In nonuniform traffic flow. Consequently. . Each moves at its own local wave velocity. 71-2. The local wave velocity which determines the characteristic is a constant. and therefore the local wave velocity for this observer remains the same! The velocity at which each observer moves is constant! Each observer moves at a constant velocity. The characteristics may not be parallel straight lines. these curves are again called characteristics. For each observer. the observer moves at the local wave velocity. dx/dt = q '(/>«). Along the curve dx/dt = q'(p). Along a characteristic.

x = a. . If we are able to determine such a characteristic. the traffic density p is a constant. for the characteristic initially emanating from x = ft. 63-3).Thus. 71-3. Thus the equation for this one characteristic is Along this straight line. This technique is called the method of characteristics. also a straight line characteristic. Let us describe some properties of this density wave velocity. At this velocity the traffic density stays the same. the x-intercept of this characteristic. the characteristic that goes through that space-time point must be obtained (see Fig. Figure 71-4 Using characteristics to determine the future traffic density. we have Fig. the density wave velocity decreases as the traffic becomes denser. the density of the desired point is given by the density at the appropriate jc-intercept. The density wave velocity. is extremely important. then since the density is constant along the characteristic. equals a since at / = 0.321 Sec. 71 Nonuniform Traffic—The Method of Characteristics is a straight line. but with a different slope (and corresponding different velocity) if #'(/>«) =£ q'(pp). Figure 71 -3 Possibly nonparallel straight line characteristics. dqjdp. To determine the density at some later time t = t* at a particular place x = x#. for example. Similarly. where k. 71-4). In this manner the density of cars at a future time can be predicted. We have assumed dqfdp decreases as p increases (see Fig.

71. rather than the interaction with other cars. 71-5.2. where does P = /»m«/2? Figure 71 -6. that is the density of automobiles (or density wave) always moves at a slower velocity than the cars themselves! Figure 71 -5 du/dp < 0. density wave velocity and car velocity.1. Since we know q = pu(p). . Two hours later.7) suggest that the traffic flow is approximately (where a and />max are known constants). Referring to the theoretical flow-density relationship of exercise 71. Suppose the initial density p(x.) Consequently. dqfdp < u. du/dp < 0 by the original hypothesis that cars slow down as the traffic density increases.322 Traffic Flow Furthermore. Experiments in the Lincoln Tunnel (combined with the theoretical work discussed in exercise 63. To do so the characteristic velocity is conveniently expressed in terms of the traffic velocity and density. show that the density wave velocity relative to a moving car is the same constant no matter what the density. 0) varies linearly from bumper-to-bumper traffic (behind x = —*0) to no traffic (ahead of x = 0) as sketched in Fig.1. (Equality above is valid only in very light traffic when speed limits. control an auto's velocity. EXERCISES 71. 71-6. we will now show a relationship between the two velocities. see Fig.

71. In this section we will formulate and solve one such interesting problem. After a Traffic Light Turns Green In the past sections the intent has been to develop in each reader a sufficient understanding of the assumptions under which we have formulated a mathematical model of traffic. Briefly explain how this solution was obtained.7.4.3.2). 72 After a Traffic Light Turns Green 71.323 Sec. dx\\dt = dx2/dt = u0.5. (In reality the line is finite. If dxjdt ^ dq/dp.8. with a train stopping traffic). Consider two observers. 71. In this section we have shown that if dx/dt = dq/dp. determine how x and / should change so that the traffic density remains the same. Show that ct + ccx = 0. how would you determine where to look to observe the density p0 ? Let c = dqfdp. (a) Show that V = kqj&p.) If the light stops traffic long enough. moving at the same velocity u0. 71. by a different method. p = 0 for x > 0. equation 71. then dp/dt = 0.6. (Hint: See exercise 60. 72.3. both moving at the same velocity F. Suppose that the flow-density relationship was known only as a specific sketched curve. is it possible for dp/dt = 01 Briefly explain. Since the cars are bumper to bumper behind the traffic light. Assume that the cars are lined up indefinitely and. Show that the rate of change of the number of cars between xz and Xi equals the flow relative to the observer x2 minus the flow relative to the observer at x\. We call the position of the traffic light x = 0. Show that p = f(x — q'(p)t) satisfies equation 71.1 for any function /. 71. are not moving. Consider two moving observers (possibly far apart). 71. Note that initially p = f(x). p = /?max for x < 0. The time has come to solve some problems and explain what kinds of qualitative and quantitative information the model yields. (b) Show that the average density between the two observers stays constant. Our analysis is limited then to times and places at which the effects of a thinning of the waiting line can be ignored. Suppose that traffic is lined up behind a red traffic light (or behind a railroad crossing. 71. . 72-1. Let us. If the initial traffic density was known. Xi and x2. Let us insist that Using equation 71.1 and the Taylor series of two variables. then we may also assume that there is no traffic ahead of the light. such that the number of cars the first observer passes is the same as the number passed by the second observer. rederive the fundamental result. Thus the initial traffic density distribution is as sketched in Fig.9. of course. but could be very long.

then dpjdt = (dp/dt) + (dx/df)(dp/dx) = 0. the traffic starts to thin out. the traffic light turns from red to green. Suppose that at / = 0. What is the density of cars for all later times? The partial differential equation describing conservation of cars. t) is constant along the characteristics. note that ifdx/dt = dq/dp. must be solved with the initial condition Note the initial condition is a discontinuous function. traffic hasn't started to move even after the light changes. Traffic is less dense further ahead on the road. As a brief review. the density is becoming thinner or rarefied and the corresponding solution will be called a rarefactive wave. can we guess what happens from our own observations of this type of traffic situation ? We know that as soon as the light turns green. Figure 72-2 turns green. Before solving this problem. but sufficiently far behind the light. Thus we expect the density to be as illustrated in Fig.1 may be solved by the method of characteristics as discussed in Sec.324 Traffic Flow Figure 72-1 Traffic density due to an extremely long red light. 71. Thus the traffic density p(x. Traffic density: expected qualitative behavior after red light We will show the solution of our mathematical model yields this type of result. 72-2. which are given by . Partial differential equation 72.

There p(x. that is no cars have reached that region. . In fact imagine you are in the first car. 72 After a Traffic Light Turns Green The density propagates at the velocity dqjdp. Now we analyze the characteristics that intersect the initial data for x < 0. the density moves at a constant velocity. and therefore in this model you accelerate instantaneously to the speed Wmax. the density is zero. the car velocity for zero density. The characteristics are straight lines. At a fixed time if one is sufficiently far from the traffic light. Thus p = 0 along all lines such that where this velocity has been evaluated using equation 72. Hence the characteristic which emanates from x = x0 (x0 > 0) at t = 0 is given by Various of these characteristics are sketched in Fig. Figure 72-3 Characteristics corresponding to no traffic. In the x-t plane where each characteristic may have a different integration constant k. and thus there would be no cars at x for t < #/wmax. The characteristic velocity for zero density is always wmax. The characteristic curves which intersect the *-axis for x > 0 are all straight lines with velocity t/max. Let us analyze all characteristics that intersect the initial data at x > 0. where the cars are standing still being at maximum density. As soon as the light changes you observe zero density ahead of you. 72-3. The first characteristic in this region starts at x = 0 and hence x = «maxf. then no cars have yet arrived and hence the density is zero. p = />m«.You would not reach the point x until t — Jt/wmax.325 Sec. Thus below in the lined region (x > wmaxO. P = />ma* along these characteristics determined from equation 72. Since p remains constant.2.2. 0) = 0.

72-4. Consider the «th car in line at the light. Thus these characteristics are all parallel straight lines with the appropriate negative velocity that intersect the negative x-axis. Does w'(/>max) significantly vary for different road situations ? Data roughly extrapolated from the Lincoln Tunnel experiments (see sec.e.. You can perform this experiment. the maximum density is certainly in the region of "heavy" traffic. (Note u'(pm^ < 0. how far back). This velocity is negative since «'(/>maJ < 0. After the light changes to green the cars start moving such that it takes a finite amount of time before each car moves. wouldn't you say ?) Figure 72-4 Method of characteristics: regions of no traffic and bumperto-bumper traffic. Is the waiting time roughly linearly dependent on the car's position as predicted above? Use your data to compute «'(/>max).) We have ignored driver reaction and acceleration time. 62. the nth car waits an amount of time equal to where L is front-to-front distance between cars. (A familiar experience. as sketched in Fig. Hence we expect this time to be a little too short. This theory predicts that after the light changes to green. assuming />max = 225 cars/mile) suggest that . It might be interesting to measure the waiting times at traffic lights as a function of the car's position (i. The boundary of the region in which p = />max is the characteristic emanating from x = 0 (at t = 0).326 Traffic Flow where we have used the fact that u(pmtK) = 0. The cars are still bumper to bumper in the region indicated in Fig. 72-4 on the left.

Ax. near the traffic light. then we expect the solution to this problem to be essentially Figure 72-6 Continuous model of the initial traffic density. the waiting time is or approximately % second per car. 72-6. 72-5. 72 After a Traffic Light Turns Green For each car behind the light. 72-1. the predicted waiting time is In seconds. namely the regions of roadway in which the density is either 0 or /?max. If Ax is sufficiently small. To investigate this problem we first assume that the initial traffic density was not discontinuous. see Fig.327 Sec. So far only the easiest part of the problem has been calculated. . We seemed to have utilized the method of characteristics to its total extent since the initial density consisted of only the two values shown in Fig. as the density has not been determined in the region the region in which cars actually pass through the green traffic light! Figure 72-5. but smoothly varied between p = 0 and p — pmtx in a very small distance. This is insufficient. We have predicted the density is and as shown in Fig.

If the initial traffic density is in fact discontinuous (see Fig. 72-8. where x0 is the position of the characteristic at t = 0 and is very small (we might later expect that it can be ignored). The straight line characteristics have different slopes. If Ax ^ 0. Since p ranges continuously between p = 0 and p = />max. p is constant along the characteristics which are again straight lines (you should repeat the reasons as to why) x = (dqjdp)t + *o. 72-7. Where the density is smaller. There must be characteristics which emanate close to the origin. 72-4). 72-10).The characteristics not corresponding to p = 0 or . then we will obtain the density in the "unknown" region by considering the limit of the continuous initial condition problem as Ax —> 0." The distance over which traffic changes from no cars to bumper to bumper increases as time increases. A few of these characteristics are sketched in Fig. the wave velocity diminishes. and then for denser traffic the wave velocity is negative. Figure 72-7 Space-time diagram for rapid transition from no traffic to bumper-to-bumper traffic. As density increases. the velocity dqjdp is greater (see Fig. namely between wmax and />m»x«'(/>max)> respectively. Notice that the characteristics "fan out.328 Traffic Flow equivalent to the solution in the case in which Ax = 0. The traffic "spreads out" or "expands" after the light changes from red to green. There is a value at which the wave velocity is zero (recall it is a stationary wave corresponding to the road's capacity). 72-7. p is constant along the line sketched in dotted lines in Fig. the velocity dqjdp is always between its values corresponding to p = 0 and p = /?max. the characteristics along which p = 0 and p = />mtx may be sketched in a spacetime diagram as are demonstrated in Fig.

i) the density wave velocity is known: Equation 72. characteristic.529 Sec. At the point (x.* go through x = 0 at / = 0 (this is the result of letting Ax —> 0). Thus x0 = 0 and It is as though at the discontinuity (x = 0) all traffic densities between p = 0 and p — />max are observed. the density is constant. Since dqjdp only depends on p. To obtain the density at a given x and /. often it is . 72-9. Along each Figure 72-9 Fan-shaped characteristics due to discontinuous initial data. The characteristics are called fanlike (representing an expansion wave) in the region so illustrated in Fig. we must determine which characteristic goes through that position at that time.4 must be solved for p. 72 After a Traffic Light Turns Green Figure 72-8 Method of characteristics: Ax is the initial distance over which density changes from 0 to pm»x- P = An. The observers (following constant density) then travel at different constant velocities dqjdp depending on which density they initially observe at x = 0.

dq/dp is calculated via equation 72. 72-11 equals dq/dp. dqjdp is then located on the dqfdp versus p figure and the corresponding value of p determined as illustrated in Fig. Alternatively. the slope of the straight line from the origin to the point (t. in this case.4. sometimes only a sketch of dq\dp may be known. . the density may be determined graphically as follows. Figure 72-11 Traffic density in fan-like region of characteristics: graphical technique. The traffic density can thus be estimated by finding the density on the q-p curve whose slope is the same as xjt. Given jc and t. 72-11. 72-10. An explicit example of this calculation is discussed in the next section.c) in Fig.330 Traffic Flow possible to algebraically solve for p as a function of x and t (actually. As always. 72-10. However. Figure 72-10 Determination of traffic density from density wave velocity. a function of xfi) in the region of fanlike characteristics. At a given position within the region of fanlike characteristics. as demonstrated in Fig. Given t and x. the Fundamental Diagram of Road Traffic can be used to determine graphically the density at a given position on the roadway in the region of fanlike characteristics. . as shown in Fig. Thus this straight line must have the same slope as the tangent to the flow-density (q-p) curve. we have assumed that dqjdp decreases as p increases.

. Position an observer at a traffic light. the maximum flow occurs at the light. 73. This suggests a simple experiment to measure the maximum flow. and experimentally determine how long a car has to wait after a light turns green as a function of the number of cars it is behind the light. 72. 72-10.331 Sec. A Linear VelocityDensity Relationship In order to illustrate the method of characteristics as it applies to traffic problems. 72.2. Wait until the light turns red and many cars line up. If this theory is correct (that is. Sketch the traffic density 10 minutes after the traffic light turns green. if u = «(/>)).4. Then.3. for educational reasons we will frequently find it convenient to choose a simple velocity-density relationship having the general desired features. EXERCISES 72. Compare your experimental results to the theory of this section. In the problem just discussed.5.2). simply measure the traffic flow at the light.* then *We should note that the equation expressing conservation of cars is still a nonlinear partial differential equation. 72. as sketched in Fig. If the velocity-density relationship is assumed to be linear. Show that if « = u(p) is determined by braking distance theory (see exercise 61. sketch the traffic density 10 minutes after the traffic light turns green. Sketch dqjdp as a function of x. 73 A Linear Velocity-Density Relationship The maximum flow occurs where dq/dp = 0. Stand at an intersection. x = 0. Assume that dq/dp is known as a function of />. Using the flow-density relationship of exercise 68. when the light turns green. then this measured traffic flow of cars will be constant and equal to the maximum possible for the road (the capacity of the road). 72. as soon as the light changes from red to green. Thus the density wave that is stationary (density wave velocity equals zero) indicates positions at which the flow of cars is a maximum. and stays there for all future time. then the waiting time per car after a traffic light turns green is the same as the human reaction time for braking.2. for fixed t > 0 (after the light turns green). not hampered by turns.1. Hopefully enough qualitative insight will be gained from a simple curve to justify the quantitative errors in not using an experimentally observed velocity-density curve.

which is sketched in Fig.. yielding a parabolic Fundamental Diagram of Road Traffic. In this case the traffic flow can be easily computed. 73-1. The density wave velocity.< 0 (in a simple way) (4) */?/*//> decreases as p increases (since d2qjdpz < 0. as will be shown). This has the four desired properties: (2) «(0) = «M Uf/ (1) "(An. Figure 73-1 Linear velocity-density curve.*) = 0 (3) -£. yields both positive and negative wave velocities. . For this Figure 73-2 Parabolic flow-density relationship. The maximum flow occurs when the density wave is stationary (density wave velocity equals zero).332 Traffic Flow which is sketched in Fig. 73-2. The wave velocity decreases as the density increases (i.e. dtqldp* < 0).

333 . Sec. 73 A Linear Velocity-Density Relationship

linear velocity-density curve, the density at which the traffic flow is maximized is exactly one-half the maximum density, p = pmtJ2, and the speed is similarly one-half the maximum speed, «(/>m.x/2) = wmax/2. (These values should not be taken too literally for realistic situations, as they are based on the possibly inaccurate linear density-velocity curve.) Thus the maximum traffic flow is

a quarter of the traffic flow that would occur if bumper-to-bumper traffic moved at the maximum speed. Let us suppose that the velocity is given by equation 73.1. We will solve for the traffic density after the traffic is started from a red light. That is, we will consider the initial density as before,

The density wave velocities corresponding to p = 0 and p = /?max are easy to calculate. From equation 73.3, (dq/dp)(ty = «max as before, and (dqldp)(pm^ — — "max- Thus the characteristics along which p = 0 and p = />max may be sketched on a space-time diagram, as Fig. 73-3 shows. We will explicitly

Figure 73-3

Space-time diagram for traffic light problem.

calculate the density in the fanlike region, — «maxf < x < w max f. There, the characteristics are given by

since they start from x = 0 at t = 0. For the linear velocity-density relationship, the density wave velocity is given by equation 73.3 and hence

Solving for p yields

334

Traffic Flow

For fixed time, the density is linearly dependent on x (in the region of fanlike characteristics). Note at x = 0, p = />max/2, the density corresponding to maximum flow (as shown in general in Sec. 72). Let us sketch in Fig. 73-4 the density at t = 0 and at a later time using the known positions of the boundaries of maximum and minimum traffic densities. It is seen that the density of cars spreads out.

Figure 73-4

Traffic density: before and after light turns green.

The result can be seen in a different manner. We have shown the density p stays the same moving at the density wave velocity dqjdp given by equation 73.3. Let us follow observers staying with the constant densities />m«, 3/>max/4> />m»x/2, /W/4, and 0, marked by • on the diagram in Fig. 73-5 representing the initial density. Each observer is moving at a different constant velocity. After some time (introducing an arrow showing how each observer must move), Fig. 73-6 shows that the linear dependence of the wave velocity on the density (equation 73.3) yields a linear density profile (as previously sketched from equation 73.4).

Figure 73-5.

Figure 73-6

Different traffic density wave velocities.

Let us compute the motion of an individual car starting at a distance JCD in back of this light, that is x = —x0 (at t = 0). The velocity of the car is given by the field velocity

335

Sec. 73 A Linear Velocity-Density Relationship

The car stays still until the wave, propagating the information of the change of the light, reaches the car, as illustrated in Fig. 73-7. After that time, / = x0/tfmtx, the car moves at the velocity given in the fanlike region,

Figure 73-7

Car path (while car isn't moving).

When a car behind the light starts moving its velocity is first zero and then slowly increases. Since the density is determined from equation 73.4, it follows that the car's velocity depends on both its position and on time:

To determine the trajectory of each car (i.e., the position x as a function of time f)» the solution of equation 73.5, a linear first-order nonhomogeneous ordinary differential equation, must be obtained which satisfies the initial condition that at It can be solved in many ways. One method (there are others) is to note that this equation, rewritten as

is a nonhomogeneous equidimensional equation. The method to solve this equation is analogous to the method used for the second-order equidimensional equation.* The homogeneous solution is

"One of the simplest second-order differential equations with nonconstant coefficients is the equidimensional equation (also called the Euler or Cauchy equation):

It has solutions of the form x = tr.

336

Traffic Flow

where B is an arbitrary constant (this solution is obtained either by equidimensional techniques, x = trt or by separation of variables). A particular solution is proportional to tr if the right-hand side is proportional to f (r ^ 1/2). Thus x = At is a particular solution if (by substitution)

Therefore the "undetermined" coefficient is A = Mmtx. Hence the general solution is The initial condition, equation 73.6, determines B and thus Consequently, the position of this car is determined,

The car's velocity is

From equation 73.6 the car starts moving with zero initial velocity; it slowly accelerates. Its velocity is always less than wmtx. For very large t, the car approaches maximum velocity; dx/dt —» wmtx as t —* oo, as shown in Fig. 73-8. How long does it take the car to actually pass the light ? That is, at what time is x = 0? From equation 73.7, Thus, that is 4 times longer than if the car were able to move at the maximum speed immediately. At what speed is the car going when it passes the light? We do not need to do any calculations as at the light the traffic flow is maximum, which we have shown occurs when the velocity is £ the maximum velocity, u = umtJ2. Equation 73.8 agrees with this result.

337

Sec. 73 A Linear Velocity-Density Relationship

Figure 73-8

Path of a car accelerating past a traffic signal.

If the light stays green until time T, how many cars will pass the traffic light? We have already determined that a car starting at — x0 passes the traffic light at t = 4x0/«m.x. Thus at time T, a car starting from —um^T/4 will be at the light. The number of cars contained in that distance is pmtx(umt]tTI4). (This result can be obtained in a simpler manner. We know the flow at the traffic light, the number of cars passing per hour, is wmix/>m.x/4. Thus in time T, (wm»x/>mi*/4)r cars have passed!) For a one-minute light, using />mix = 225 and «max = 40 m.p.h., the number of cars is

The graphical technique based on the flow-density curve may also be used to determine the traffic density after the light turns green, as well as to approximate each car's path. Along characteristics, the density is constant. Since the car velocity only depends on the density, it too is constant along characteristics. Thus characteristics are isoclines for the differential equation,

determining the motion of individual cars. Using an x-t diagram such that the slopes are measured in units of velocity yields Fig. 73-9. Figure 73-9 follows from the Fundamental Diagram of Road Traffic shown in Fig. 73-10. To determine car paths, small horizontal lines (indicating no motion) are sketched wherever p = />max. In addition, for example, we note that at x = 0, the density is that corresponding to the road's capacity, and a car's velocity there is marked by the dotted straight line on the Fundamental Diagram of Road Traffic. This slope is also then marked wherever the density has that value, as shown in Fig. 73-11. By connecting straight dashes (the method of isoclines, see Sec. 26), the path of a car can be estimated for this problem, as well as for those for which analytic solutions are impossible!

338

Traffic Flow

Figure 73-9.

Figure 73-10.

Figure 73-11

Car path: graphical sketch.

EXERCISES

73.1.* Assume that the traffic density is initially

Sketch the initial density. Determine and sketch the density at all later times. *In exercises 73.1-73.5 assume that « = </m«x(l — l/lmtx)-

339

Sec. 74 An Example

73.2.* Calculate the maximum acceleration of a car which starts approximately one car length behind the traffic light (i.e., x(0) = — l//?max). 73.3.* Calculate the velocity of a car at the moment it starts moving behind a light, 73.4.* Since we are assuming cars are of zero length, it is only the front of the first car that moves at the speed limit. Since the density has been determined in this section, calculate the position of the end of the first car (with length L) as a function of time (for sufficiently large time). 73.5.* Suppose at / = 0 that the density is

73.6.

Sketch the initial density. Determine and sketch the density at later times. Assume Determine the traffic density that results after an infinite line of stopped traffic is started by a red traffic light turning green. Suppose that u ^ «max(l — /;//?max). Show that a car a distance x0 behind the light takes a time x0pmtjqm^ to pass the light, where qmtK is the capacity of the road. Using the flow-density relationship of exercise 68.2, sketch the trajectory of a car starting at a given distance behind the light. At what velocity does the information that the traffic light changed from red to green travel ?

73.7. 73.8. 73.9.

74. An Example

We have shown explicitly how to use the method of characteristics to solve traffic problems which initially consist of regions of constant density. The same ideas can be utilized when the initial density varies in a prescribed way, For convenience, we again assume u(p) = w max (l — />//>max)5 *n which case the density wave velocity determines the characteristics as follows:

The characteristic starting from

340

Traffic Flow

along which the density is constant, equaling its value at / — 0, The characteristics are sketched in Fig. 74-1. We assume that the characteristics do not intersect. The more difficult case (and perhaps more interesting one) in which characteristics intersect is not discussed until Sec. 76.

Figure 74-1

Nonparallel nonintersecting characteristics.

We use the method of characteristics in two equivalent ways to determine the traffic density as a function of x and t: (1) PARAMETERIZING THE INITIAL POSITION AS A FUNCTION OF X AND t Each characteristic is labelled by its position, jc0, at / = 0. Given x and /, we try to find XQ (i.e., which characteristic goes through the point (x, t)). p is eliminated using equation 74.2 and thus equation 74.1 yields JCQ as a function of x and /, This step can not always be done explicitly as it may be impossible to solve for JCQ. For example, if

then the characteristics are determined from equation 74.1,

from which an equation like equation 74.3 cannot be explicitly obtained. Since the density at a point only depends on JCQ (i.e., on which characteristic goes through it), when equation 74.3 exists. Thus substituting equation 74.3 into equation 74.2 yields the spatial and time dependence of the traffic density, equation 74.4.

If x0 > L or x0 < 0.341 Sec. an equation results involving only jc. Since the corresponding density wave velocities are easily calculated. we obtain the two regions of constant density. . the characteristics given by equation 74.2 to determine *„ as a function of p. 74-2. and p.1. However. showing p's dependence on x and t. as sketched in Fig. it is not always possible to obtain from equation 74.1 start from a region of constant density. Again. As a specific example. 74 An Example (2) PARAMETERIZING THE INITIAL POSITION AS A FUNCTION OF THE INITIAL DENSITY An equivalent method is to first use equation 74. let us use the method of characteristics as described by both equivalent procedures (1) and (2). 74-3. following from the space-time sketches of the characteristics shown in Fig.5 is substituted into equation 74. In the region where the traffic density has not been determined as yet. assume that u(p) = «max(l — />/An»x) and Figure 74-2 Initial traffic density.2 an explicit expression for x0. /. when equation 74.

0 The characteristics which start from 0 < x0 < L satisfy equation 74.7 determines x0 as a function of x and t. the density approaches the known constants.9 . In particular. We note that as x approaches the edges of the region of varying density.8) The negative sign must be chosen above in order for 0 < x0 < L (this is seen by recalling that — «maxf < x < wmaxf + L).1. since equation 74.342 Traffic Flow Figure 74-3 Characteristics.6: admittedly a rather cumbersome expression. Equation 74. The traffic density as a function of x and t in the region corresponding to 0 < x0 < L follows by substituting equation 74.7 is a quadratic equation for x0 (more easily expressed in terms of x0 — L by noting #0 = x0 — L + L): The solution of this quadratic equation is (74. from equation 74.8 into equation 74. where and thus the equation for these characteristics is where we should remember this is valid for all XQ as long as 0 < jc0 < L. (7) *0(x.

since 0 < xa < L. both the denominator and numerator tend to zero. EXERCISES 74. which is identical to that derived by procedure (1). To determine the limit as t —* 0 of equation 74. .1.9 satisfies the initial conditions.1. Squaring this last equation yields an expression for p(x. the simplest technique is to approximate the numerator as t —> 0. t). (x0 — L)2 = L2p/pmtx or A:O = L ± L+/p/pmtli. equation 74. By substituting equation 74. f). 74 An Example Furthermore. This is not obvious since as / —» 0.6 to determine JCQ as a function of p. Assume that u(p) = w max (l . Since ^\ — t ^ 1 — fy as t —» 0. the minus sign must be used above: Note that as p varies between 0 and /?max. we should verify that equation 74.9.343 Sec.9. we begin by using equation 74.10 into equation 74. an equation for the density is obtained : This equation may be expressed as a quadratic equation for */plpmtx: Thus where the positive sign of the square root has been chosen since ^/p/pmtx > 0. x0 varies between 0 and L./?//>max) and Determine and sketch p(x. However. we see that as t —> 0 as originally specified for 0<x0<L. Alternatively.

/). In a short length of time you too are suddenly compelled to apply your brakes. 75-1.9 approaches the correct traffic density. Consider the following partial differential equation: (a) Why can't this equation model a traffic flow problem? (b) Solve this partial differential equation by the method of characteristics. Wave Propagation of Automobile Brake Lights Before the study of more complex traffic flow problems. Its velocity should be . Consider the example solved in this section. Have you ever been traveling in heavy traffic on a highway and observed that after someone applies their brake (lighting the taillight) a long distance ahead. he or she would have to travel opposite to the direction of traffic!). let us indicate a simple explanation for the fascinating wave phenomena of automobile brake lights. We model the process of applying brakes in the following way. subject to the initial conditions: 74. You will apply your brakes at a different position from where the first car applied its brakes.4.2. see Fig. The lit taillight might have traveled backwards against the traffic (if a bicyclist were to keep up to the brake light. Thus when cars are at that density taillights go on. 75. What traffic density should be approached as L —> 01 Verify that as L —> 0 equation 74. it is a reasonable approximation. Although this might not be entirely accurate. taillights indicate a path of constant density.344 Traffic Flow 74.3. Determine p(x. Assume u(p) = Kmax(l — />2//>max) and 74. Assume that a driver applies the car's brakes at some critical density pc. succeeding cars apply their brakes ? The brake lights appear to travel in your direction. Under this hypothesis. a characteristic.

then dqjdp \p=pc < 0 and the wave propatates in the direction opposite to traffic! We really are not suggesting that this is the exact mechanism by which brake lights are observed. The traffic becomes denser or compressed. 76-1. For convenience assume that p —> /?j as jc —> +00 and p —> p0 as . The solution we will obtain is called a compression wave. where 0 < pQ < pl < /?max. Congest/on Ahead Let us imagine a situation in which traffic initially becomes heavier as we go further along the road.c —> — ex. as shown in Fig. 76 Congestion Ahead Figure 75-1 Lit brake light moving backwards.. What does our mathematical model predict? First a few characteristics are sketched in the region of heavier traffic (p % /? t ) and a few in the region of lighter traffic (p ^ />0). given by If cars brake in "heavy traffic". 76. To make an analysis of this situation requires a careful experimental investigation of the conditions under which a driver of a car applies its brakes. A density wave for the heavier traffic moves at velocity dq\dp \P=PI which is less than the velocity of .345 Sec.

Figure 76-2 Characteristics intersect. 76-4). clearly it makes no sense to have three values of density at one place. that is. The one in lighter traffic moves at a constant velocity. Clearly this is impossible. as Fig. as illustrated in Fig. In this situation the prediction of "breaking" is then quite significant! . eventually the observer on the left passes the observer on the right. in shallow water) are similar to the equations for traffic density waves. in any situation in which the traffic becomes denser further along the road.. Thus we have Fig. We say the traffic density wave "breaks. Then we obtain Fig.e. At a position where an intersection occurs. The "distance" between the heavier and lighter traffic is becoming shorter. the lighter traffic density wave. 76-4. The sketched characteristics are moving forward. 76-2. faster than the one in heavier traffic. However. If we continue to apply the method of characteristics. the velocities of both density waves were assumed positive. This does not have to be so in general. Let us follow two observers. at some later time our mathematics predicts there will be three densities at some positions (for example.* The density must be a single-valued The partial differential equations describing the height of water waves near the shore (i. Something has gone wrong! To explain this difficulty the density is roughly sketched at different times based on the method of characteristics.346 Traffic Flow Figure 76-1 Heavier traffic is ahead initially. the theory predicts the density is /?„ and at the same time Pi. Eventually these two families of characteristics intersect as illustrated in the figure. characteristics will still intersect. The light traffic is catching up to the heavier traffic. Instead of the density distribution spreading out (as it does after a light turns green). 76-3 illustrates." However. Thus the method of characteristics predicts that the traffic density becomes a "multivalued" function of position. A and B. it is becoming steeper. starting at positions marked • each watching constant density.

If density and flow are continuous. 77. Discontinuous Traffic On the basis of the partial differential equation of traffic flow. . Some approximation or assumption that we used must at times be invalid. Since the method of characteristics is mathematically justified. let us briefly review the assumptions and approximations necessary in our derivation of the partial differential equation of traffic flow: 1. it is the partial differential equation itself which must not be entirely valid.347 Sec. In the next section we will resolve this difficulty presented by the method of characteristics. Assuming that average quantities (such as density and velocity) exist. we predicted the physically impossible phenomena that the traffic density becomes multivalued. function of position along the highway. we formulated the integral conservation of cars. then the integral conservation law becomes a differential conservation law. 2. 77 Discontinuous Traffic Figure 76-3 Evolution of traffic density as lighter traffic moves faster than heavier traffic. To discover what type of modification we need to make. Figure 76-4 Triple-valued traffic density as predicted by the method of characteristics.

the differential conservation law becomes a partial differential equation for the traffic density. px) rather than u = u(p). possibly beyond the present level of many readers of this text. and that this discontinuity might propagate in time xt(t). We reconstruct the derivation of conservation of cars. the assumption that the traffic density and velocity field are continuous functions of space and time. This modified assumption is briefly discussed in the exercises. We wish to continue using similar types of mathematical models involving traffic density and velocity. for example. the difficulty of multivaluedness associated with the method of characteristics disappears. u = u(p. The traffic variables remain single-valued. where necessary. cars are still not created or destroyed. flow is no longer valid at a discontinuity. a driver could respond by slowing down faster than implied by the density at the driver's position alone. the cars' velocity to depend also on the gradient of the traffic density. One or more of these assumptions must be modified. but only in regions along the highway where the traffic density has been predicted to be multivalued. < x. By postulating that the velocity is only a function of density. Hence. The resulting mathematical theory will not be difficult to understand and interpret. where we assume jc. we prefer to modify our mathematical model in a different way. Assumption (3) could be removed by allowing. dp/dx).348 Traffic Flow 3. the mathematical techniques necessary to obtain these results and to apply them to various traffic problems are relatively difficult. 77-1) and velocity field are discontinuous at some unknown position x. This would take into account the driver's ability to perceive traffic problems ahead. However. As the density becomes much larger ahead. In particular. Hence we will continue to assume that (1) is valid. Otherwise our formulation is adequate.(t) < x2: . since the resulting partial differential equation of traffic Figure 77-1 Traffic density discontinuous at x = xs(t). Furthermore. Consider the number of cars contained in the region xt < x < x2. Assume that the traffic density (as illustrated in Fig. it can be shown (although we do not) that the theory we generate concerning traffic flow with discontinuities can be related to the theory developed by using continuous traffic variables with u — u(p. We will investigate traffic flow removing. Instead. u = u(p). rather than to depend only on the traffic density. in space. let us consider assumption (2).

The rate of change of cars is only due to cars crossing at x — x^ and x — x2. Examples are the sound emitted from an explosion or the thunder resulting from lightning. When fluctuations of pressure and density are small. They are called sound waves.) A similar expression can be derived for the number of cars per hour crossing a moving boundary at x = xz. shock wave. we allow the two endpoints to move.349 Sec. There. the number of cars per hour crossing the moving boundary (the flow relative to a moving coordinate system) is (If the boundary moves with the car velocity. 66). then the pressure and density can be mathematically modeled as being discontinuous. called the position of the shock. t) has a jump-discontinuity. is introduced because of the analogous behavior which occurs in gas dynamics. then no cars pass the moving boundary. changes in pressure and density of air. for example.2 jump fThe terminology.t occurs at xs(t). propagate. Then sound is propagated at a constant speed known as the sound speed. However. and are heard (due to the sensitivity of the human ear). the equations describing sound waves can be linearized (in a manner similar to that discussed in Sec.* Since x. if the amplitudes of the fluctuations of pressure and density are not small. but instead moves with velocity dxjdt. Since the left end x = xl is not fixed. Consider the rate of change of the number of cars between x — x^t) and x — xz(t). . the result being called a shock wave. which we call a shock wave or simply a shock. If a shock wave results from exceeding the sound barrier. Thus Suppose that the jump-discontinuity. Let both the a function f(x) is said to have a ump-discontinutiy figure 77. depends on /. it is known as a sonic boom. 77 Discontinuous Traffic This integral is still well defined even if p(x.

This jump in density is denoted [/?]. Using this notation. and on the other side pz. Thus The traffic flow relative to the moving shock on one side of the shock equals the relative flow on the other side. or equivalently (since x^ = x~ and x2 — x+) Similarly the numerator is the jump in the traffic flow across the discontinuity [q]. yields The denominator is the nonzero difference in the densities on the two sides of the shock. Solving for the shock velocity. the shock velocity. then the . one on one side and one on the other side (as though the two boundaries were two bicyclists riding on a bicycle-built-for-two with the shock between the two riders). the shock velocity (or velocity of the discontinuity) is given by If on one side of the shock the density is />.350 Traffic Flow left and right boundaries of the region move exactly with the shock. pdx = 0 and therefore Jxi rxi Both ends move at the same velocity. Since no cars will be contained in the region xl <x<x2 (it being of infinitesimally small length).

the integrand p is continuous and hence we can differentiate each without any difficulties. In the next section a calculation of this kind will be undertaken. The shock condition can be derived in an alternate manner. That is why we divided the integral above into two parts.3): This formula is not valid if f(x. However. Applying this well-known formula twice.357 Sec. we have not yet explained where shocks occur and how to determine pl and p2. 77 Discontinuous Traffic shock must move at the following velocity: At points of discontinuity this shock condition replaces the use of the partial differential equation which is valid elsewhere. Imagine two fixed observers at x = x^ and x = x2. the limits of the two integrals are not constant since xt depends on time. As a review the rule for differentiating an integral with variable limits is quoted from calculus (see exercises 60. xt(t). However.2 and 60. in which case conservation of cars implies that Suppose the discontinuity occurs at some place between x = xl and x — xz. yields . It is best to divide the integral into two parts. In each integral. t) has a jump discontinuity between the two limits of integration. We will calculate the derivative of an integral whose integrand is discontinuous.

3 or 77. equation 77.3 from equation 77.3. 77-5.5 differently. If x = . Exercise 77. Then we have Fig. Consider two fixed places x = a and x = b as indicated in the figures. is infinitesimally near the shock. 77-4. the shock moves a distance Axs. However. In addition.8 derives equation 77.5 vanish. The change in the number of cars. Thus which is equivalent to the shock condition as previously derived. separated by a shock. The shock velocity is given by equations 77. suppose that a short time A/ later. x = xs. graphical Let us briefly describe the manner in which a very simple traffic shock propagates. as shown in Fig. The shock velocity can be positive or negative as illustrated in the figure. as shown in Fig. and if x — x2 is infinitesimally near jc = xs at the same time. Figure 77-3 Flow-density curve: determination of shock velocity.552 Traffic Flow This formula is valid for all time.t. Figure 77-4 Initial traffic density. is —Axs(p2 — /?. The number of cars . A^A/) — W(0).). note that the slope of the straight line connecting the points on the flow-density curve representing these two traffic situations (p ~ pl and p = p2) has exactly the value of the shock velocity. 77-3. The change in the number of cars between x = a and x = b is only explained by cars crossing at x = a and at x = b. then both integrals in equation 77. Assume that the traffic initially consists of two uniform traffic states of density pl and p2. This shock velocity can also be graphically represented on the flowdensity curve.4. Suppose that a shock occurs when lighter traffic with density Pi catches up to heavier traffic with density p2.

If u — wmsx (1 — /72//0m«xX then what is the velocity of a traffic shock separating densities p0 and p\ ? (Simplify the expression as much as possible. 77. show that the shock velocity is approximately the average of the density wave velocities associated with the two densities. A weak shock is a shock in which the shock strength (the difference in densities) is small. In the next section we will consider traffic problems in which the partial differential equation predicts multivalued solutions. [Hint: Use Taylor series methods. that crosses at x — a is Af^ = btp^p^. where v is a constant. since the traffic flow is constant there. For a weak shock. (a) What sign should v have for this expression to be physically reasonable ? . Similarly the number of cars that crosses at x = b is Aty2.353 Sec. 77. Suppose instead of u = U(p\ that a car's velocity u is 77. Show [p2] ^ [pV. Thus or Therefore the velocity of the shock.5. EXERCISES If « = «maxO — p!pmax\ then what is the velocity of a traffic shock separating densities p0 and pi ? (Simplify the expression as much as possible. 77. 77 Discontinuous Traffic Figure 77-5 Propagating shock wave: caused by two different uniform traffic densities.2. is again the same as derived before. A^/A/.1. and hence shocks must be introduced to give physically realistic solutions.3.) Show that the shock velocity is not the average of the density wave velocities associated with pQ and p\.] 77.) Show that the shock velocity is the average of the density wave velocities associated with Po and pi.4.

and Xs < X < X2.354 Traffic Flow (b) What equation now describes conservation of cars ? (c) Assume that U(p) = wmax(l — plpm^.5. Suppose that a solution exists as a density wave moving without change of shape at velocity V.) We will only analyze the traffic behind the light. see .5. Give a physical interpretation of this result. ignoring the slowing down phenomena associated with a yellow light. V. Formulate such a model. disregarding what occurs in front of the light. 78-1. [Hint: See exercise 61. Derive the equation for the shock velocity.7.6.3.5.e.] What type of initial conditions would be necessary to solve such a problem ? (b) The model in part (a) essentially introduces a delay (due to the time a driver takes to accelerate) in a driver's response to the observed density. (We assume the first car immediately stops.] Reconsider equation 77. The initial density is a constant. (a) What ordinary differential equation is satisfied by/? (b) Integrate this differential equation once. At x = 0 the traffic is stopped (i. (c) Show that the velocity of wave propagation. such that /—> p2 as x —> +00 and /—> /?t as x —> — oo. Roughly sketch this solution. Instead drivers accelerate in a manner so as to approach a desired velocity-density curve. using the appropriate partial differential equation valid for Xi < x < x. The red light is mathematically modeled in the following simple way. only if p2 > p\. modify the model in part (a). Unless drivers look far enough ahead to compensate for this response time.8. By graphical techniques show that a solution exists. [Hint: See exercise 77. and hence p = /?max for all time t > 0. 77. 78. 77. is the same as the shock velocity separating p = p\ from p = p2 (occurring if v = 0). u = 0). Uniform Traffic Stopped by a Red Light In this section we will investigate what happens to a uniform stream of moving traffic (with density p = p0) as it is suddenly halted by a red light at x — 0. (a) A more sophisticated theory of traffic flow assumes drivers do not move at a velocity determined by the density. this delay process is very unstable. p = p0..Show that called Burger's equation. as shown in Fig. Consider Burger's equation as derived in exercise 77. Accordingly. 77.

Characteristics that emanate from regions in which p = pQ move at velocity dq/dp \pt. a propagating wave demarcating the path at which densities and velocities abruptly change (i. Therefore. We will show that this difficulty is remedied by considering a shock wave. and on the other side p = y?max. Boundary and initial con- Fig.. bumper-to-bumper traffic. Let us suppose that there is a shock wave. the diagrams in Fig. The density wave associated with lower densities travels faster. Figure 78-2 ditions. are discontinuous).e. We do not . In either case there is a cross-hatched region indicating that the method of characteristics yields a multivalued solution to the partial differential equation. where p = pmax) travel at velocity dq/dp | x. 78-2.555 Sec. On one side of the shock the method of characteristics suggests the traffic density is uniform P = Po. 78-4. 78-3 indicate that characteristics intersect each other whether the initial uniform traffic is light or heavy: Figure 78-3 Stopping of uniform traffic: characteristics. 78 Uniform Traffic Stopped by a Red Light Figure 78-1 Initial traffic density. while characteristics emanating from the position of the traffic light (x = 0. the space-time diagram is of the form shown in Fig. In either light or heavy traffic.

This shock separates cars standing still from cars moving forward at velocity w(/>0). the traffic density is discontinuous. as shown in Fig. 78-5. 78-6. Hence the shock velocity is determined. yields the following equation for the shock velocity: However. Consequently. t > 0. that is Substituting the jumps in traffic flow and density. know as yet the path of the shock. = 0 at / = 0. In this case. giving a condition for this firstorder ordinary differential equation.356 Traffic Flow Figure 78-4 Unknown shock path. applying the initial condition results in the following equation for the position of the shock: The resulting space-time diagram is sketched in Fig. The theory for such a discontinuous solution implies that the path for any shock must satisfy the shock condition. If this happens. Thus the shock moves at a constant negative velocity. there is no traffic flow corresponding to bumper-to-bumper traffic). w(/>max) — 0 (that is. For any time. the shock must initiate at x. The initial position of the shock is known. then the cars must decelerate from u(p0) to .

u(p). must be modified only in regions extremely near shocks. One suspects that such a mechanism would prevent shocks. Instead of using the velocity field equations. = —N/pmt^ yields Thus the (N -f l)st car will stop at . in a more realistic mathematical model the dependence of velocity only on density. Let us show that we can explain the increase of cars in line behind the red light in another way.357 Sec. Have you ever observed sudden decelerations? Since cars cannot instantly decelerate. that is regions in which the density is changing rapidly. Suppose that each car moves at speed u(p0). We will not develop the mathematical model to include these effects although such a modification can be accomplished. and its distance from the light initially (at t = 0) is N/p0. How long will that take? Substituting X N+ . u(p0). Since its velocity is constant. consider each car individually. this theory predicts accidents at these shocks. how many cars would be forced to stop by catching up to the line of stopped cars? Consider the (N + l)st car behind the light when the light changes to red. However. In such regions a good driver presumably would observe the stopped traffic ahead and slow down before being forced to suddenly brake. this car's position is determined by solving the initial value problem: Thus When this car is forced to stop it will be N/pmtK distance from the light. At time /. 78 Uniform Traffic Stopped by a Red Light wave figure 78-5 Figure 78-6 Traffic density resulting from stopped traffic. zero instantaneously.

the coordinate x and t are reversed from their usual position). Also sketched on the diagram are lines associated with (1) density wave velocity corresponding to /> max . in which case The line of stopped vehicles moves at a constant velocity which is the same velocity as previously derived! An equivalent space-time sketch is facilitated by using the flow-density curve. and (4) shock velocity between the uniform flow p0 and bumper-to-bumper traffic/> max . To follow the building line of stopped cars. 78-7 on the flow-density curve: Figure 78-7 Fundamental Diagram of Road Traffic: various important velocities. The resulting space-time diagram is Fig. (3) car velocity corresponding to p0. 78-8 (in which. . (2) density wave velocity corresponding to /?„.358 Traffic Flow Its position when stopped is We have determined the position and time at which each car stops.3 and 78. A car starting from x = —L is marked Figure 78-8 Graphical determination of traffic density and car paths.4. as discussed before. Suppose that the uniform traffic stopped by the red light has density p0 marked in Fig. eliminate Nfrom equations 78.

For which of the preceding cases is a density shock necessary ? Briefly explain. (b) Determine and sketch the density at later times. 78. assuming that the initial traffic density is a constant p0. 78. show algebraically that the value of the shock velocity is between the velocities of the two density waves.2. Assume that u = «max(l — P/pm**) and at t = 0. Why does the density not change in time? Referring to the problem in Sec. 78 Uniform Traffic Stopped by a Red Light with a dotted line (showing the car velocity to be faster than the density wave velocity). If u = wmax(l — />//Jmax). . (c) Determine the path of a car (in space-time) which starts at x = — x0 (behind x = 0). Suppose that only a sketch of the flow-density curve is known.1. then at what velocity do cars pile up at a red traffic light. (d) Determine the path of a car (in space-time) which starts at x = x0 (ahead of x = 0). Assume that the density is known at / = 0. 78. EXERCISES 78. Assume that u = wmax(l — plpm**) and that the initial traffic density is 78. 78. Suppose that the initial traffic density is Consider the two cases.359 Sec. the traffic density is 78. (a) Sketch the initial density.6.5. p0 < p\ and pi < pQ.4.3. Determine the traffic density at later times.

Figure 79-1. 0) = p0.2. 79-1 in which initially This is an initial semi-infinite line of bumper-to-bumper traffic followed by no traffic. .12. Do exercise 78. Let us imagine the situation sketched in Fig.10.) Determine the traffic density on a semi-infinite (x > 0) highway for which the density at the entrance x = 0 is and the initial density is uniform along the highway (p(x. How would you mathematically model the yellow light? (Do not solve any problems corresponding to your model. Each car observes bumper-to-bumper traffic and hence remains standing still.9. (b) Assume that p2 < p\. assume that u(p) = wmax(l — plpmtx) and thus Pi < p0 < An«/2).. Using the flow-density relationship of exercise 68.8. There is no . 78. sketch the traffic density 10 minutes after the traffic density was 30 cars per mile for x < 0 and 150 cars per mile for x > 0.11. let us consider one more simple example. 82.8 if p0 < pi < pm^J2.e.] 79.Determine the traffic density p (for r > 0) if the initial traffic density is (a) Assume that /?2 > p\. 78. Suppose that a traffic light turned from green to yellow before turning red. 78. 78. x > 0). A Stationary Shock Wave Before studying more complex traffic problems-. The solution to this traffic problem is quite easy. what is the traffic density in front of the light? [This problem will be answered in Sec. If uniform traffic is stopped by a red light.360 Traffic Flow 78. Assume that pi is lighter traffic than p0 and both are light traffic (i. Assume that u(p) = wmtx(l — />2//>m«x).7. Sketch the density at various values of time. 78.

On these: For the third case if we assume that the linear velocity-density relationship is valid. Furthermore fanlike characteristics can be constructed emanating from the origin (as though the initial density p(x. 79-3). For all time. we obtain Fig. Figure 79-4 Characteristics (including fan-shaped characteristics). As before.361 Sec. in the region of intersecting characteristics. Figure 79-3 Intersection of characteristics. 79 A Stationary Shock Wave motion. 79-2. This problem is discussed in the same manner as for the mathematically similar initial conditions sketched in Fig. p = 0 are drawn (see Fig. there are three families of characteristics. 0) smoothly varied monotonically between p = 0 and p = /?max in a very short distance). . 79-4. Consequently. first the characteristics corresponding to p = /?mix and Figure 79-2. Thus. The characteristics clearly intersect in a V-shaped region. We will show that the method of characteristics (when modified by shock conditions) yields this same result.

Show that the solution with shocks corresponds to cutting off equal areas of the "lobes" of the triplevalued density as drawn in Fig. Show that using the method of characteristics (ignoring traffic shock waves). it will stay there. Reconsider the problem of Sec. Thus it is concluded that this shock is stationary (i.362 Traffic Flow then (see Sec. as was already known! EXERCISES 79. * Reference in Sec.e. sketched in Fig.. Whitham* has shown this to be a general result. does not move). a shock is introduced. the shock speed between bumper-tobumper traffic and no traffic is calculated: Since M(/?max) = 0.1. the density of cars is triplevalued—clearly not allowable. 79-6. the traffic density is triple-valued. . 73) the density depends linearly on position. Since it starts at x = 0. To remedy this difficulty. 78. 86. In particular. In the region of intersecting characteristics. Sketch the resulting triple-valued traffic density. Using the nonfanlike characteristics. 79-5 are some of these triple-valued densities for different values of/: Figure 79-5 Traffic density ignoring shocks.

suppose the initial density wave velocity is . The characteristics corresponding to less dense traffic move faster. we began to describe the propagation of shock waves in traffic problems.3. then shock waves are still generated. (c) A traffic shock will form at this point. (f) Describe briefly how the individual automobiles behave (do not determine their paths mathematically). t). 79. However. Suppose that 79. thus the shock waves formed immediately./2wmax. Determine the velocity of the shock. showing the shock and the characteristics necessary to determine p(x. t) before and after the shock. To be specific. In the examples considered. Find its subsequent motion.p/pmtx). 80. (b) Show that all characteristics from the interval 0 < x < L (and / = 0) intersect at the point x = L/2.363 Sec.2. (d) Sketch the x-t plane. Briefly give a physical explanation of the result. imagine a situation as drawn in Fig. The Earliest Shock In the past sections. but it takes a finite time. The initial traffic density on a road is Assume that u(p) = umtx(l . (e) Sketch p(x. we now will show that the dominant feature of traffic flow problems is that even if the traffic variables are initially smoothly spatially dependent. Let us illustrate this property by considering a situation in which the initial traffic density is locally larger somewhere along the road. t = Z. 80-1. 80 The Earliest Shock Figure 79-6. for example. (a) Sketch the initial density. the density was initially discontinuous.

Figure 80-1 Initial traffic density and initial density wave velocity.±3: Figure 80-2 First intersection of characteristics.±l.±}. ±i. in which case we sketch in Fig.±2. 364 . 80-2 the characteristics starting from x/L = 0.

We will show that even though Ax —* 0. However. Shocks are expected. see Fig.365 Sec. then no characteristics could have crossed at an earlier time. Figure 80-3 Intersection of two characteristics. If this is the first intersection. the first intersection occurs at a finite positive time. any characteristics between x = xl and x — Xi -f Ax at / = 0 will intersect one of the other two characteristics almost always before t — T: as illustrated in Fig. 80-3. 80-2 suggests that shocks do not occur immediately. the other from xl + Ax. 80-4. 80 The Earliest Shock The characteristics eventually cross due to the region in which the density is increasing. one emanating from x. The equation for the characteristics is We will analyze neighboring characteristics. Figure 80-4 An earlier intersection of characteristics. Consider the characteristic . the first shock actually occurs due to the intersection of neighboring characteristics (the limit as Ax —> 0). Suppose that the first shock occurs at / = T. The fast traffic catches up to the slower moving denser traffic. the time of the earliest shock. Instead. Thus shocks cannot first occur due to characteristics that are a finite distance Ax apart.. due to the intersection of two characteristics initially a distance Ax (not necessarily small) apart. Fig. Let us attempt to calculate' when a shock first occurs.

366 Traffic Flow emanating from x = x t at / = 0.e. the time does not approach zero. Solving for the intersection point by eliminating x yields Therefore the time at which nearly neighboring curves intersect is Consider the characteristics as paths of observers following constant density. 80-5 indicates the general behavior of these two Figure 80-5. Thus we shouldn't be surprised that as Ax —> 0. the limit as Ax —» 0 must be calculated: . and the characteristic starting from x — xl + Ax at / = 0. + Ax than at x t . To consider neighboring characteristics.. Assume that Ax > 0. 0). where X*i) = X*i»0) an<^ X*i + Ax) = X*! + Ax. how much faster the one on the left is moving). the relative velocity is also small. in which case Fig. straight line characteristics. Then this equation states that the time of intersection of the two observers is the initial distance between the observers divided by the relative velocity of the two observers (i. These characteristics intersect (in a positive time) only if that is if the density of traffic is higher at x. Although the distance in between is small.

367 Sec. when is the time t given by equation 80. d2qldp2 < 0). it follows from equation 80. and hence an alternate expression for the time when neighboring characteristics intersect is Let us determine the conditions under which neighboring characteristics actually intersect. we have marked the intersection of the "nearest sketched" characteristics (for example. indicated in Fig. in Fig. Since dq\dp only depends on p. Perhaps the initial density is such that it is locally increasing at many places along the roadway. When sketching characteristics. For all characteristics emanating in regions where dpjdx^. we must minimize the intersection time over all . > 0.e. the position where the characteristic starting at x = 0 intersects the one starting at x — —L/2).2a or 80. Thus we conclude that neighboring characteristics which emanate from regions where the density is locally increasing further along the road will always intersect.2b that only if dpjdx^ > 0 will neighboring characteristics intersect.2b positive ? Since the Fundamental Diagram of Road Traffic is always concave downwards (i.. The condition dpldxl > 0 means that the traffic density is increasing at the point where the neighboring characteristics start. To determine the first time at which an intersection (shock) occurs. 80-2. we cannot show the intersection of "exactly" neighboring characteristics. Instead. in other words. neighboring characteristics intersect. 80 The Earliest Shock Consequently from the definition of the derivative. 80-6 as ( ): Figure 80-6.

space for various times. The equation for the straight line characteristics is Xi(x.e. find the absolute minimum of t given by equation 80. From the method of characteristics.. Since the traffic density is constant along the characteristic If the initial condition is prescribed.3 by partial differentiation with respect to x (holding t fixed. In Fig. We observe from the figure that dp/dx first becomes infinite at the time of breaking (/ = tb). i.368 Traffic Flow possible neighboring characteristics. However. let us calculate dp/dx. This can be calculated by determining where These results concerning the earliest occurrence of a shock can be derived in an alternate manner. 80-7 we sketch the density as a function of Figure 80-7 Evolution of traffic density ignoring shocks. from equation 80. Thus. 74).2a. but certainly not jcj . Therefore. 0 depends in a complicated manner on x and / (see Sec. ignoring the possibilities of shocks. then p changes only as a result of changes of jc t .

Consequently. let us continue to investigate the method of characteristics. Before t = tb. 80-2). in which case traffic will thin out.2a).369 Sec. This means that one set of neighboring characteristics intersects at the same time as another set. 80-8. a shock should be introduced to avoid the triple-valued solution. Assume that u = w max (l — plpm*^(a) Show that the time of intersection of neighboring characteristics (corresponding to the collision of two observers moving with constant density) is (b) I f a t / = 0. When / = tb. the initial traffic density must be steadily decreasing. 80 The Earliest Shock Finally we obtain dp/dx becomes infinite at the same time neighboring characteristics intersect (see equation 80. but at a different position (see Fig. At t = t2 > tb there are two positions at which dp/dx — oo. However.2b imply that no shocks occur only ifdpldx: < 0 for the entire highway. dp/dx = oo at the position of breaking. the first place where neighboring characteristics intersect (see Fig. EXERCISES 80.1.2a and 80. 80-2). Equations 80. rather than shock. No shocks occur only if the initial traffic density is of the general form shown in Fig. After t — tb. (1) Sketch the initial density. traffic shocks are almost always predicted by this theory of traffic. Figure 80-8 Traffic situation such that a shock will not form. (2) Determine the time of the first shock. (3) Where does this shock first occur? . no neighboring characteristics intersect and dp/dx is never infinite.

80. sketch the characteristics corresponding to x/L — 0. 80) indicates that the characteristics are not parallel. From the sketch. show that the first shock occurs at t = 8Z. The characteristics are approximated by parallel straight lines. the exact solution obtained by the method of characteristics (see Sec.0) — p0 + f/(*i). Calculate in general dp/dt.6. 80.If the process of linearization is a valid approximation. Suppose the initial traffic density is such that at / = 0 (a) Using graph paper. 81. ±2. However. (c) From the results of this section. 66-70 we showed the traffic density at later times can be approximated by where c = dqldp\p=p<>. calculate dp/dx and dp/dt. For the problem roughly sketched in Fig.370 Traffic Flow 80. 80-1 (with p0 = 15 and pi = 45). [Hint: Since the shock velocity is approximately the average of the two density wave velocities (see exercise 77. ±f. Thus characteristics .9. 80-2. Geometrically construct the shock wave if initially the density is as sketched in Fig. ±2> ±1. 80-2. but are nearly parallel./3wmax. The density wave moves forward for light traffic and backward for heavy traffic. Show that dp/dt tends towards infinity as the time of the earliest shock is approached. when and where does the first shock occur ? Compare your answer to part (a). (b) Illustrate in your sketch the space-time region in which the method of characteristics predicts the density is triple-valued.3. Show that both tend towards infinity as the time of the earliest shock is approached. the velocity of the shock can be approximated by taking an "eyeball" estimate of the average slope!] Using the result of exercise 71. estimate the time and position of the first breaking. p(xl.3). 80. Briefly explain how we can improve the accuracy of Fig. For nearly uniform traffic.2.5. Validity of Linearization One of the first traffic problems we considered was one in which the initial traffic density is nearly uniform. 80.4. then in Sees. where the flow-density curve is given by Fig. 68-2. we will show that this linearization process is usually not valid for large times. ±3.

not all times. 0) = /7max + € f(x).3. This conclusion also frequently holds for nonlinear problems in other areas of study where a linearized analysis is utilized. (a) Why will we assume that f(x) < 0? (b) Determine the density for all times (using the linearized theory). Reconsider exercise 67. Does this correspond to your experience traveling in a car ? .3. 0) = p0 + €f(x). You may assume that u = Wmax(l — P/pmax).4. (c) Assuming that u = «max(l — /J//?max).] 81. approximately solve the differential equation of part (d). This time at which characteristics cross is quite large. Under what circumstances will the exact solution differ substantially from the linearized approximation ? 81. However. t). for large times a shock will occur not predicted by the linearized theory. if somewhere (df(x)/dx) > 0 (assuming d2g/dp2 \p=po < 0).2. (a) Assume that e > 0. Consider a situation in which traffic is initially nearly bumper to bumper. (d) What equation determines the motion of a car starting at x — L ? (e) For times that are not too large. p(x. For times less than that. [Hint: The car velocity is small and hence the car does not travel very far for times that are not too large. 81 Validity of Linearization may intersect. Reconsider exercise 81. 81. EXERCISES 81.Discuss differences between the exact answer and those predicted by the methods of Sec. calculate the velocity field. Determine p(x. A shock first forms at the smallest value of / such that Since.1. We thus conclude that linear theory is valid for limited times. Approximate the difference in velocities. show that the exact density wave velocity differs only slightly from that predicted by the linearized theory.371 Sec. If initially p(x. 67.2. no shocks occur and the slopes obtained by the method of characteristics are approximately the same as the constant value for the density wave calculated by the linear theory. O(l/e). a shock will occur at some large (positive) time. (b) Assume that e < 0. Describe the motion of a car starting at x = 0 if f(x) = — 1 + sin (x/2QL).

as illustrated in Fig.* The initial density is a constant p0. It is as though the light separates traffic of zero density from traffic of density /Jmax. "Shock Waves on the Highway. . After the light turns red. Uniform traffic density as instantaneously effected by a red This problem was analyzed by P. The shock velocity is the velocity at which a line of stopped cars increases. Assume that traffic is moving at a constant density />„. 42-51 (1956). Effect of a Red Light or an Accident In this section. I. 82-2. we will analyze a problem of practical interest.372 Traffic Flow 82. 82-1. and then traffic is stopped at a point (x = 0)/or a finite amount of time T (for example. 78). cars line up behind the light in a manner we have already investigated (Sec. Figure 82-1 Shock formed behind a red light. We will determine the effect of traffic being momentarily stopped. In this configuration the lighter traffic (p = 0) is behind heavier traffic (p = />„) and hence a shock forms between the faster moving density wave with density zero and the density wave with density Figure 82-2 light. Richards. In front of the light." Operations Research 4. the initial traffic density is also p0. see Fig. The traffic light gives a boundary condition (in front of the light) of zero traffic. because of a red light or an accident).

. The velocity at which the line continues to increase remains the same. equation 82. we assume that the light changes from red to green. the velocity of this shock on the "right" can be obtained using the shock condition: As we know. At any fixed time (before the light returns to green). this shock wave travels at the same velocity as each car (of density />0).373 Sec. 82-4. Thus if the light stayed red forever. Figure 82-3 Red light stopping traffic—method of characteristics (note two shock waves). while ahead of it p = pQ. the cars at density p0 continue to pile up at the line of stopped cars even though the light has changed (this continues to occur until the line of cars completely dissipates). Far behind the light. Traffic density after a red light stops traffic (note again two At some time T. yielding a situation in which traffic gradually accelerates through the signal. 72-73). 82 Effect of a Red Light or an Accident p = />„. Thus the position of this car is where the shock occurs. the density is as shown in Fig. Mathematically. Figure 82-4 shocks). then we obtain the characteristics sketched in Fig. Immediately near the light. it is again the shock velocity between p = p0 and p — />mtx. 82-3. since behihd this car p = 0. The last car that traveled past the traffic light before it turned red moves at velocity w(/>0).1. the problem is one already solved (see Sees.

73.1. separating the uniform density p0 from bumper-to-bumper traffic. . Thus for some time shortly after t = r. but the line is being increased at the speed . The first car after the light travels at the fastest velocity t/(0) = wmtx. and hence catches up to the traffic of density p0.374 Traffic Flow These results have been obtained for any velocity-density relationship. the cars continue to move uniformly at velocity «(/>„). moves at the velocity Far ahead of the light. there is no effect from the light. There. 82-5. In the rest of this section for convenience we will assume as discussed in Sec. we note that the density wave velocity is Furthermore the general expression for the shock velocity may be simplified as follows: or equivalently Dividing through by p2 — Pi yields a simpler expression for the velocity at which a discontinuity between densities pl and p2 propagates: An alternate interpretation of this result is discussed in exercise 77. The line of stopped cars decreases at the speed z/mtx. Using this linear relationship. the density is as sketched in Fig. Thus the "left" shock.

. tu can also be computed as the distance the lead car is behind the trail car of the uniform traffic at t = T. Which occurs first: 1. 82-5) Hence. Hence the total line of stopped cars eventually dissipates. wmax . The lead car catches up to the uniformly moving traffic at time tu. 82-5) or thus when a very long time for light traffic. 82 Effect of a Red Light or an Accident Figure 82-5 Traffic density a short time after stopped cars are restarted. (p0um^Jpmt^T:. Thus which is equivalent to equation 82.6. which is less than w max . The time it takes to dissipate the line after the light turns green is the length of stopped cars at t = r. divided by the relative velocity. the constant velocity times the length of time the light was red). (u(p^)t.375 Sec.(p0umjpmtj.5. when (following from Fig. (The dotted line indicates the density at t = t). The line of cars dissipates at time td> when (see Fig. Thus which is equivalent to equation 82. a shorter time for heavier traffic. the lead car catches up to the uniformly moving traffic. divided by the velocity at which cars dissipate. wmax — u(p0). at / = tu or 2. the line of stopped cars completely dissipates at / = / d ? We will solve for the time at which each event occurs.

For example. in which p0 > pmtx/2. vice versa. the lead car catches up first.x/2. see Fig. At maximum traffic flow (if the uniform density corresponds to the capacity of the road). the line dissipates first. back to green. Figure 82-7 Traffic density when lead car catches up at the same time as line dissipates. 82-6(b). see Fig. 82-9(a). The space-time characteristic curves are sketched for light and heavy traffic in Fig. in which p0 < /?max/2.6. After the time of each previously mentioned event. The shocks are marked in heavy lines. Eventually they catch up to the thinned out waiting line as the stopped cars have all started again. . certain cars far in back of the light do not notice that the light has changed from green to red. Figure 82-6 Traffic density: (a) in heavy traffic. it is seen that for heavy traffic. p0 = />m. a shock occurs resulting from the intersection of the fanlike characteristics (emanating from the light turning green at x = 0 at t = T) and the characteristics corresponding to the uniform density p0. both phenomena occur simultaneously. as seen in Fig.376 Traffic Flow By comparing equation 82. In addition. as depicted in Fig. 82-8(a) and (b). For light traffic. (b) in light traffic.5 to equation 82. a lead car catches up before line of stopped cars completely dissipates. These cars move at velocity u(p0). 82-7. 82-6(a).

Figure 82-8 traffic. Before the first of these two events occurs. for the "right" shock. Characteristics: preliminary sketch for both light and heavy Figure 82-9 Intersection of characteristics: (a) behind light after line dissipates. we see Fig. 377 . In both cases the shock strength (that is the difference between the two traffic densities) is decreasing! We now determine the paths of these two shocks. 82-9(b) indicates that. eventually the cars accelerating from the traffic light overtake the uniformly moving traffic of density />„. the density distributions are as previously sketched. After the "left" shock occurs. Fig. while after the "right" shock occurs we have Fig. 82-10. In either case. 82-11. (b) in front of light after lead car catches up.

378 Traffic Flow Figure 82-10.7. (see equation 73. Figure 82-11.5). is quickly solved. Using the results for fanlike regions (with a linear velocity-density relationship-see Sec. yields The velocity of this shock is not a constant. the ordinary differential equation describing the path of the shock.8. where pf and uf stand for the density and velocity in the fanlike region (either thinning or accelerating regions). Using equation 82. since it represents the moving jump-discontinuity between a constant density region and a nonconstant density region. 73) starting at / = T. The shock strength could be measured as [p] or pf — p0. It is not a uniformly moving shock. Using equation 82. pf depends on x and t. where B is the integration constant. The two different shocks ("left" and "right") correspond to different values of B. However. equation 82. .3 yields However.

More general discussions.8 Multiplying by /?max — /?„. Although both shocks are approaching the same velocity.. both shocks eventually travel forwards. yields . beyond the scope of this text. If traffic is heavy (p0 > />max/2). For this reason we determine the value of Bh the integration constant for the "left" shock. that is at To study the effect of the stopped traffic. namely at From this condition. i. from equation 82.9 both shocks go backwards.e. The initial condition for the left shock is the position and time at which a car far behind the light catches up to the last remnant of the stopped traffic. the distance between the two shocks is tending towards infinity since The initial condition for the right shock is the position and time at which the first car catches up to the uniform traffic. as t — r —> oo This asymptotic shock velocity is just the density wave velocity corresponding to p0. both shock velocities tend to the same constant. The velocity of the shock is As the time increases after the light turns green. Furthermore we have determined the rate at which the shock strengths tend to zero. show that in other situations the shock strength tends to zero: it is usually also proportional to time raised to the — £ power. 82 Effect of a Red Light or an Accident The shock strength (of either shock) tends to zero as t — r —> oo. From equation 82. while if traffic is light (p0 < pmaJ2).379 Sec. the traffic behind the light will be thoroughly investigated. interestingly enough. Br can be determined for the right shock.

back to green) has completely dissipated at this time. 82-12. its velocity is approximately constant Thus this asymptotic shock velocity is greater than (not as negative) the shock directly due to the red light. 82-13. then from equation 82. the traffic congestion caused by the momentary stoppage of traffic never disappears. or equivalently when Using the expression for Bh yields This phenomena is sketched in Fig. Figure 82-12 Shock wave behind light slows down. the traffic density at the position of the traffic light suddenly drops from pmsJ2 (the value at the origin for all time since the . For heavy traffic.8 the shock never returns to the position of the traffic light. then from equation 82.380 Traffic Flow Thus If the traffic is heavy (pQ > />mtx/2). as illustrated in Fig.9 the shock eventually reverses its motion. For large t. If the traffic is light (p0 < />max/2). the speed of the shock slows down. At this moment. but travels backwards indefinitely. The shock returns to the origin when. The accelerating line of cars (due to the light changing from green to red.

the first car that passes the light after the traffic jam clears has been traveling at the constant velocity u(p0) for time t. 82-14.cit. if p0 = $/>mtx (light but not very light traffic).. or fifteen times the length of the light. However. Since these two expressions must be equal. then the time of the elimination of the congestion is The traffic jam lasted from r to 16r. under these conditions. where Capacity ™ the capacity of the road. This time can be quite long. u(p0)t'p0.381 Sec. we see that the time for the traffic jam to clear is Therefore which is equivalent to equation 82. After the strengths of both shocks have started to diminish. As an application.10. traffic started) to the uniform density p0. An observer at the traffic light would mark this as the time when the traffic jam has finally cleared away. This figure is called an . produce a traffic jam requiring an additional hour and a quarter to clear! A simpler way to determine the time at which traffic clears is to note that the number of cars having gone through the light is (/ — T)#cap. an accident which halted traffic for five minutes would. 82 Effect of a Red Light or an Accident Figure 82-13 Shock wave behind light passes the light. a sketch of the density as a function of position would be Fig. As Richards noted. Thus the total number of cars that passed the light is the product of the initial density p0 and the initial distance u(p0)t the "first" car is behind the light.

In summary.1.1. EXERCISES 82. If the initial density is with 0 < pi < p2 < p3 < />max> then determine the density at later times. Assume that u = w max (l — /?//>max). What happens after these two shocks meet ?] . the space-time diagram for a momentary stoppage of uniform light traffic is sketched in Fig.382 traffic flow N-wave. 82-15. Calculate the shock between p^ and p2. Figure 82-15 Method of characteristics showing shocks resulting from a temporary stoppage of light traffic. [Hint: See exercise 77. Show that this shock moves faster than the shock between pi and p3. Figure 82-14 Traffic density after both line dissipates and lead car catches up.

Consider a short transition between a two-lane and three-lane highway. How long is that car delayed in passing the increased density? [Hint: Find the time it would take to travel the same distance without the increased density. 82-16. 82. Reconsider exercise 82. . Using the results of exercise 82. calculate the density per lane of traffic at later times if initially the density (per lane) is a constant p0 everywhere. Assume that the velocity-density relationship is still the same for each segment of the highway. Assume that u = «max(l — /?//>max) and that the initial traffic density is 82. as sketched in Fig. Assume u = «max(l — />//>max) and the initial traffic density is 82.3. (a) Determine the density at later times.6./>//?max). 82.6 with a three-lane highway for x < 0 and a twolane highway for x > 0.and three-lane highway is of a negligibly small distance.5.and three-lane roads respectively. 82 Effect of a Red Light or an Accident 82. Determine the density at later times.pip™*) and that 82.] Analyze the traffic flow caused by a traffic light temporarily stopping heavy traffic if u(p) = Hmax(l . Let p2 and p3 be the traffic density per lane and Figure 82-16. where pi > pQ. so that we can assume that there is a twolane highway for x < 0 and a three-lane highway for x > 0. (b) Consider a car starting from x < — a.2.4. <72 and q3 be the traffic flow per lane in the two. Assume that «(/>) = «m«x(l .383 Sec. where p\ < p0. show that What happens to cars as they go from a two-lane highway into a three-lane highway and vice versa ? Assume that the transition region between a two. (Is this reasonable ?) If the number of cars in the transition region is always small.7.5.

12. Suppose that a construction truck enters the road and travels at u(p0)/2 for a quarter of an hour before leaving the roadway.) in the three-lane highway is more than the two-lane road's total capacity (refer to exercise 82. p0 < />m«/2. 82. 82.7). how long must the light stay green in order for traffic not to back up behind the light? Redo exercise 82. one which has a bottleneck in which for some distance the road is reduced from three to two lanes: Figure 82-17.Determine the traffic density if initially where H < L/2.13.) Consider the two highways in Fig. . Consider a car a distance x = L behind the bottleneck. Calculate the resulting density as a function of time. The initial density on a single-lane highway is p0 < />m«/2 (and the corresponding velocity u(p0)).9. A moving line of traffic waiting to enter the narrower highway will develop. 82. How long is it delayed? Consider the traffic congestion caused by a slow-moving truck in the following problem. 82.11.11 is equivalent to equation 82. If the traffic light stays red for time T. 82-17.384 Traffic Flow 82. 82. Assume that u(p) = wmtx(l — plpm**) and traffic is coming to a traffic light at a light density p0. then equation 82.8. (Hints: A shock must occur.10. (a) Assume that the initial density is such that the total flow (moving at density /?0) in the three-lane highway is less than the two-lane road's total capacity. Compare highways (1) and (2) with the same initial traffic density (per lane) Po under the condition in which the total flow (moving with density /><.11 if u(p) is known but u(p) =£ Kmtx(l — plpmubAssume u(p) = wmax(l — plpmt*). (b) Assume that the initial density is such that the total flow (moving at density p0) in the three-lane highway is more than the two-lane road's total capacity.10. In particular. determine the paths of all traffic shock waves. Show that if u(p) = umtx(l — plpm**). starting at x = 0.

called 0(x. then q(p). t). We reconsider conservation of cars. and jc = x2. and x = xz.385 Sec. Exits and Entrances Let us now relax the assumptions concerning no exits or entrances on the segment of road being investigated. t) be the net number of cars per hour entering the region between x = *. It might be possible for example for ft to be negative. Exits and entrances are now allowed. where Taking a derivative with respect to x2 (and replacing x2 by x) yields the partial differential equation for conservation of cars (allowing exits and entrances). is modified by the introduction of the term ft on the right hand side representing the effect of entrances and/or exits. If u(p).) Thus conservation of cars implies It is more convenient to introduce the net number of cars per hour entering per mile of roadway. . Let /?C*15 x2. of cars between x = *. 83 Exits and Entrances 83. but also from cars entering or exiting the road between x = xl and x = x2. The rate of change Figure 83-1 Exits and entrances. (The net number is the difference between the number entering and those exiting. as Fig. 83-1 represents.3. and consequently The partial differential equation expressing conservation of cars. and x = x2 results not only from cars crossing at x = x. equation 65.

and F represents the effect of lane changing. (b) What. i. Let pt be the density in the slower lane. Hence the characteristics themselves are no longer straight lines! The density waves do not move at a constant velocity. an example will be discussed that illustrates the solution of this system. Consider a two-lane highway going in one direction which allows cars to pass. if then this conservation equation implies The curves such that dx/dt = dq/dp are again called characteristics. The expression again represents the time derivative moving with the velocity dq/dp.1. Consider a highway with one lane of traffic density /?+ and velocity u+ going in one direction. Passing is allowed enabling the average velocity to increase at the same density. The density is no longer constant in the characteristic direction. .. Formulate a model of this highway with passing.386 Traffic Flow The method of characteristics can be used to solve this equation if jff is either known or has known dependence on the traffic density.2. and p2 be the density in the faster lane. EXERCISES 83. Assume that w t = «i(/>i) and «2 = u2(pz) are different. Assume that it is possible for cars to change lanes.e. In the next section. and a second lane of density />_ and velocity M_ going in the opposite direction. (a) Show that 83. qualitative dependence on pl and p2 do you expect for F? Postulate a simple mathematical model of this. The partial differential equation has been reduced to a system of two ordinary differential equations.

Thus it is no longer appropriate to assume u(p). Suppose that the road itself is not uniform. it is necessary to know the characteristic curve as illustrated in Fig. The constant of integration p0 is the traffic density at t — 0. The density depends linearly on time along the as yet unknown characteristic curve. If the characteristic emanates from x = x0. For example. The equations for the characteristics are if The density is not constant along the characteristic. [Hint: See exercise 83.] 83. If the characteristic going through the point (x. then the .387 Sec. reduces the partial differential equation to what system of possibly nonlinear ordinary differential equations ? 84. (a) What is the formula for the traffic flow? (b) Assume cars enter at the rate of ft cars per hour per mile of roadway.2 shows that since we assume /?0 is a constant. 84-1. but the same kinds of approximations suggest u(p. Instead. What partial differential equation describes conservation of cars ? (c) The method of characteristics. x). For each characteristic curve the constant p0 may be different. Constantly Entering Cars If the net rate per mile of cars entering a highway is a constant 00. then In order to determine the density at a given position and time. Traffic problems which are more complex than those previously analyzed can be formulated. t) was curve A in Fig.3. 84-1. integrating equation 84. the road surface may vary affecting the cars' velocity. 84 Constantly Entering Cars What qualitative dependence on p+ and /?_ will u+ and «_ have? Assume {/-turns are not allowed. then the traffic density at all times must satisfy the following partial differential equation: as derived in the previous section.1.

u = «maxO — />//>m»i). However. If neighboring parabolas intersect. 84-2. Thus we obtain Fig. the density increases as traffic builds due to the constantly entering cars (assuming /?0 > 0). the characteristics are parabolas as where x0 is the initial position of the parabolic characteristic. if the characteristic was curve B. the characteristics obtained are not straight lines. we are unable to know what value p0 should be unless the characteristics are known.4. then p = fi0t + pB.388 Traffic Flow Figure 84-1 Density determined by unknown characteristic.However.3 must be solved. t) would be from equation 84. equation 84. 73).4 p = 0Qt + PA. To determine the characteristic curves. a shock forms. is valid. in this case.4. These parabolic characteristics can be sketched knowing the Figure 84-2 Constantly entering cars: parabolic characteristics. Thus from equations 84. initial distribution of traffic density. density at (x. For convenience. in which case (see Sec. in which case a shock condition is necessary. we assume that the linear velocity-density relationship. As time increases. equation 84.3 and 84. .4 is satisfied. the characteristics satisfy the following ordinary differential equation: By integrating this equation. Along each parabola.Even though the density is given by equation 84.

ft = —yp. If the initial traffic density is a constant p0.Under what conditions (if any) will a traffic shock occur? [Hint: See Sec. EXERCISES 84.] 84.. At what time is this model no longer valid? Briefly explain. . Suppose that we approximate the rate of exiting cars as being proportional to the density.1. 85 A Highway Entrance as may be derived in the exercises. Show that the neighboring parabolic characteristics of Sec. then traffic density satisfies Suppose initially that there are no cars on the road However. Thereafter each car's velocity is limited. The shock condition is the same as that which occurs without exits and entrances. From the integral conservation of cars. Show that the position of the first shock is further along the road if cars are constantly exiting than if no cars enter or exit. i. 80. What is the resulting traffic flow? We expect that the first car enters and accelerates to the maximum speed. derive the shock condition when cars are constantly entering a highway. suppose cars are entering the road (in some finite region 0 < x < xE) at a constant rate B0 per mile for all time. [Hint: See Sec. 84 first intersect at the same time as the first traffic shock occurs without entrances or exits.4.e. Suppose that cars are entering an infinite highway (—oo < x < oo) at a constant rate /?0 per mile. If we assume a linear velocity-density relationship.3. Assume that there are some exits but no entrances along a roadway.] 84.2. determine the traffic density for all later times. we will solve a problem involving traffic entering a highway. 84. 80.389 Sec. 85. A Highway Entrance In this section. Assume that u = «max(l — plpm*x).

Figure 85-1 In regions without entrances. ft is zero. while they are in the region in which ft = ft0. In the entrance region of the highway (ft = ft0).2a. In this region. Some of these parabolas start at t = 0 at x = XQ from regions of zero traffic density in which case These parabolas differ from each other by a constant translation in x. . see Fig. the characteristics are parabolas. The density is increasing in time as cars enter.390 Traffic Flow The method of characteristics implies that when In sections of the highway in front of (and behind) the region of entering cars.3a and 85. by eliminating T from equations 85. For these characteristics. the characteristics are straight. equation 85. and the characteristics are straight lines corresponding to a constant density wave velocity (equal to the maximum car velocity if there are no cars). 85-1.3b. Other parabolas emanate from x — 0 at some values of t = r at which p = 0.

If (t/mtx/>m.J4j?0) > xe. At x = XE. Figure 85-3. After leaving the region of entering cars. we obtain Fig. 85-3. then the vertex of the parabola starting from x — 0 (at t = 0) would occur outside the region of entering cars. Thus. 85-2. these parabolic characteristics become straight lines. see Fig. we will assume that this flow is less than the maximum capacity of the road. the density is the same value. in which case x = wm. it can be shown that the densities are all less than />m. The traffic density is constant along each straight line characteristic. However.x/2 and hence correspond to light traffic. the density at x = XE is where t = tE is the time that the characteristic intersects the end of the entrance region.x/>mtx/4^0. the maximum value of x occurs when Thus t = pmJ200.391 Sec. x = XE (note that tE depends on x0—see Fig. for different characteristics. that is we will show these parabolas do not turn back. In . 85-2). Since 00xE is the total flow of cars coming in the entrance. the constant value of the traffic density is different. For the parabola emanating at x = 0 at t = 0. For the characteristics emanating at / = 0 (for 0 < x < XE). Figure 85-2 Straight line characteristics bend due to entering cars. 85 A Highway Entrance These parabolas are all translations in time from the parabola corresponding to x0 = 0. Since dqfdp > 0.

for t>tt. In this region of straight line characteristics P = PotE and hence Figure 85-4 Parabolic characteristics straighten upon passing through entrance region. At XE. . After leaving the entrance region. the velocity of the characteristic is but p is the constant 00tE. the traffic becomes denser and hence moves more slowly until t = tE(x0 = 0) = t. Consequently. 85-4. the density is the same.392 Traffic Flow general Thus At x = XE.. as shown in Fig. These straight lines fan out since heavier traffic occurs after the lighter traffic.

Figure 85-5 Traffic density: (a) slowly increases. In Fig. 85-5 we sketch for fixed times the traffic density. Thus the diagram in Fig. as shown in dotted lines in Fig. 85 A Highway Entrance Hence. these characteristics for x > XE are all parallel straight lines. Traffic density continues to increase in front of the entrance . (b) continues to build. 85-4. 85-6 illustrates the traffic density for all times: Figure 85-6 region.393 Sec. (c) reaches its maximum.

. 1974. A. In addition.1. I highly recommend this book because of the excellent presentation by L. New York: John Wiley & Sons.3. control. 0) = 0 directly.. New York: Academic Press. Show that neighboring parabolic characteristics do not intersect if "maxAnax/4 > /?0*£. These two books are excellent sources of additional references mostly contained in the research literature. Linear and Nonlinear Waves. many of which are accessible to the reader of this text. New York: John Wiley & Sons. B. C.. D. and assignment. Mathematical Theories of Traffic Flow. A third book which I must recommend is WHITHAM. Probabilistic models are primarily discussed in HAIGHT. Edie which includes some of the material we have discussed. 1974. density and velocity. 85. This was written by one of the developers of deterministic traffic theory. G.] . 86. (a) Give a physical interpretation of this situation. Gazis: GAZIS. Further Reading in Traffic Flow The mathematical models we have analyzed postulate the importance of the traffic variables. generation. distribution. this book expertly presents various other applied mathematics problems involving wave motion (most from areas of physics). (c) Solve the initial value problem p(x. /?0 —* °° such that /?o*£ —> (?> a constant. A collection of four thorough review articles has been edited by D. F.394 Traffic Flow EXERCISES 85. 85. Consider the case of constantly entering cars for which umtxpmaJ4 < /?O*JE(a) Determine the traffic density for x > XE(b) When does a shock first occur ? (c) What differential equation describes the path of a shock ? (d) Determine the density everywhere before the shock occurs. Although this book only discusses traffic problems briefly. C. 0) = 0 by considering a limit of the problem analyzed in Sec. it does contain significant developments of the theory. Traffic Science. Consider a highway entrance with XE —> 0. 1963. (b) Determine the solution corresponding to p(x. [Furthermore. C. We have then pursued deterministic models. other authors in Gazis' book discuss a wide variety of traffic problems including traffic delays. 85.2.

cars) distance between. 267 length. 174-176 Andronow. 305. 13-14 Automobiles (see Cars. 354 Capacity (population) (see Carrying capacity) Capacity (road). 216. 286 centripetal. 316 Boyce. 14. 373. 260. 5-6. 22-23. A. 291. C. 162-163. 138-143 Age groups. 155-156 systems. 307-308. 43-47 Accidents. 201-202 Amplitude of oscillation. 17-18 polar coordinates. 46 dimensions. 267 Cauchy equation. maximum) Car-following models. 284 Burger's equation. 384 Boundary condition. 319331. 337. 49-50 Chaikin. 293-298 Car paths. 331 Breaking wave. 16. 33-34 Characteristic polynomial (see Characteristic roots) Characteristic roots. 314 Characteristic equation. 236 Center of mass. 312-313. 328. 344-345 Braking distance. 16. sonic. 3 Arbitrary constants. 138-143 Boom. 317. 361-371. 216. 202. 124. E. 126-128 Bicycle-built-for-two. 386-394 395 . 335. 267 spacing. 306-309. Flow (traffic). Conservation of cars. 174-176 doubling. 140 Air. Traffic) Autonomous differential equations. Density (traffic). 335 Center. Car paths. 337-347. 290-291. three steps. 46 Animal traps. 350 Binomial expansion. 242-245 Average traffic density. 114 Angular momentum. 372-382 (see also Collisions) Age distribution. 168 Cars (see also Car-following models.. 114 Chain rule. 154. 188 Average populations. 93. Velocity. 25-26 Centripetal acceleration. cars) Carrying capacity. 331 (see also Flow (traffic). 5 car.. 268-269 Banks. 244-245 Applied mathematics. 123 age dependent. 221 Characteristics. 235-236 Bottlenecks.Index Acceleration. 349 Algebraically unstable. 46 Central force.. 89 Biological models (see Populations) Birth rate. W. 346 Bumper-to-bumper traffic. 312. A. 349 Borderline cases. 35. 358-359 (see also Trajectories. 49 Angular velocity. partial derivatives. 56 Brake lights.

251 Complex eigenvalues. 155-156 constant coefficients. 330. 278 variable limit integrals. 171-185 first-order. 346.) fanlike. 112 Cycles per second. 31-33. 162. 179-180 Delay-differential equations. 278. 330. 91-97 Death rate. 281-282. 15-16. 321-322. 67 Conserved quantity. 294. 348-354 gradient. 316 Conservation of energy. 285-286 Competing species.234 dimension. 386 (see also Shocks) brake lights. 49 Conservation of cars. 348 spreads out. 12-14. 15 Damping (see also Friction) Coulomb. 334 steepens. 61-69. 287 constant. 385 integral. 272 Convergent oscillations. 281 Constant of motion. 165-176 without constant coefficients. 386-392 parabolic. 162-171 Delays: population models. 30 Newtonian. 362-363. 368. 33-35. 347348. 163-164. Conservation of energy) differential. 91-97 oscillations. 218. 335 equidimensional. 172-176 Compound interest. 93 Cauchy. 278 Conservation of angular momentum. 337 neighboring. 361-363. 183-184 Euler. 346-347 nearly uniform. 284 multivalued. 131-132 direction field. 50. 332.396 Index Characteristics (cont. 294 Delayed growth. 365-369 nonintersecting. 335 constant coefficients. 40-42 Crowding. 354 De Moivre's theorem. 196-198 polar representation. 124-125. 275. moving. 298. 31-33 force. 346 triple-valued. 91-92 (see also Energy) Conservation of number of cars (see Conservation of cars) Conservative force. 129 Differential equations (see also Delaydifferential equations. 351 Dickinson. 246 traffic models. 275-282. 196-198 Complex numbers 177. 255 Collisions. 335 first-order. 172 Density (air). 123 age dependent. 345 Congestion (traffic). 113 nonlinear oscillations. 335 autonomous. 355. 369 Density-velocity relationship (see Velocity-density relationship) Density waves. 129-131 nonlinear. 349 Density-dependent population growth. 186 Competitive exclusion. 347. 376. 347-348 local. 308. 138-143 Decaying oscillations. 333-334. 227 constant coefficient. 32 Difference equations. 170 systems. Stability) autonomous. Linearization. 361-370 isoclines. 388-392 Circular frequency. 345 shocks. 164. 29-30. 266-272. 98-99. 32. 349 stationary. 344-345 breaks. 18. 48. 349 Coulomb friction. 178-185. 393 (see also Density waves) average. Partial differential equations. 340. 94 . 302 discontinuous. 331 velocity. 129-131. 328-330. 16. 346 compression. 358 Derivative: partial. 93-94. 347-348 integral. 296-297. 40-41 delays. 151-161 Density (traffic). 345 Conservation law (see also Conservation of angular momentum. 59 Closed curves. 388. 277. Conservation of cars. 340 nonstraight. 306. 130. 304-305. 183-184 linear. 33-42. 25. 247-255 Competition. 328. 172-176 Complex roots. 152 Cubic friction. 81 Critically damped oscillations. 309-315. 234 Continuum hypothesis. 18 Circular motion. 309 (see also Capacity (road)) perturbed. 348 maximum. 302 single-valued. 216. 346-347. 126-128 Compression wave. 376-382 intersecting. 236-237 Coexistent equilibrium population. 129 second-order. 302-318 optimal. 179-180 Coordinate system. 249. Phase plane. 163-164.

C. 91 kinetic.315-318. 194-199 Elimination. 55-56 populations. 61-69 conservation.332 discontinuities. 119 (see also Population dynamks) Ecosystems (see Population. 100 spring-mass system.. 38 second-order. 298. 233 Ellipse. 70 Entrances (highway). 93. 92 . 67-69 (set also Conservation of energy. 302 pendulum. 11. 91-92 decay. 126 Fertility. 34 DiPrima. 251 Fanlike characteristics. Phase plane) Energy equation.322. 135-136. L. first-order) Fish (see Shark-fish ecosystem) Flow. 265-266. 385-394 Expansion wave. 328 Expected growth rate. 251 Exits (highway). 335 Euler's formulas. 63 work. 246. 154 (see also Carrying capacity) Equidimensional equation. 175-176 Ecology. competitive principle. 335 numerical solutions. 126 First-order difference equations (see Difference equations. method of. 350 Discrete models: one-species populations. 13. 138 Displacement from equilibrium. 187-188 linearization. 199-203 undetermined coefficients. 349 Exponential growth. 226-246 nonhomogeneous. Conservation of energy) spring-mass system. 137 particular solution.) equation. 188 elimination. 336 uniqueness theorem. vector. 33-35. 63 total. 249. 203-223 stability. maximum and minimum. 349 Flow-density curve (see Flow-density relationship) Differential equations (con/. 137 Exclusion. 40-41 Lotka-Volterra. 12-14. 16-18. 289-293. 40-41 linear. phase plane. 234-235 Energy. 82 Dimensional analysis. 161 Explosion (sound). 7 Escape velocity. 57 Displacements. 42. 106 Discontinuity. fanlike) Fecundity. 296. 132-133. models) Edie. 174-176 estimation. 77-79 potential. 193-198 phase plane. 99 dissipation. 394 Eigenvalues (matrix). 335-336 Equilibrium: isolated.. 273-276. 141. 144 Expected number of cars. 53-54 Dimensions. 16-18 Dimensionless variable. 180 Doubling time. 122-125 Discrete time. Conservation of energy) Energy integral. 328-330 (see also Characteristics. 375 Divergence theorem. 336 periodic forcing function. 141. 56 Direction field. 225 Exponential spirals. 281 Divergent oscillations. 281. 217 Extinction. systems. 187. 281 Flow (traffic). 348-354 jump-discontinuity. firstorder) homogeneous. 154. 194-199 Eigenvector. 290. 242 Discretization. 348-354 maximum. 137 Discretization time. 33-35. 130. 67 Euler equation. 63 Energy curves. 63 integral. 92 line of stopped cars. 151 Explosion (population). 234 (see also Energy. 335 systems. 294 first-order: autonomous. 273 Expected population.397 Index Energy (con/. 12-14.) isoclines. 248249 positions. 188-190 matrix methods. 12-14. 63 pendulum. 24 delays. 94 (see also Isoclines) pendulum with damping. 385-394 Environment's carrying capacity. 196 Euler's method. 125. 202 partial differential equations. 7. 191-199 (see also Differential equations. R. 23 Dissipation: energy.) first-order (con/. 167. 188. 229. 131-138. 63 (see also Energy. 349 velocity. method of. 192 equilibrium. 331-334 (see also Capacity (road)) relative to moving coordinate system. 190. 94-99 systems. first-order) First-order differential equations (see Differential equations.

220-222 straight. traffic density. 20-23 Insecticide.. 112 negative. 225 Human growth rate. 281 Focus. 309. 15-16 (see also Circular frequency. 55 Hudson Bay Company. 202 lump. 36-37 Newtonian. 277. Natural frequency) Friction. D. 394 Generation time (see Doubling time) Goldstein. 210-211. 31-33. 32 Frequency. 247 delay. 358 parabolic.. 244 Instability (see also Stability) car-following models. 39 Induction. 131-138 Growth (con/. R. 64-65 resistive. 32 spring. 279 Galileo. 124-125. gravitation. 111 predator-prey. 91-97 sufficiently small. A. 226 constant. 264. 308. 59-60 Ireland. 183 discrete-delay. 9-11 acceleration due to. 153-161. 291 Hertz. 330-332. H. 349 . 297 Instantaneous growth rate. car. 148-150 Initial conditions. 291 Hooke's law. 30 Forest ecosystem. 67 damping. 277 (see also Conservation of cars) Integrals. 332 Fundamental Theorem of Calculus. 154 Growth rate. population. 163 Fox-rabbit ecosystem. 337-338. 144 human. 264. 220222. 204-205. 229-230 saddle points. 98-99. 161 Impulse. 30 Fundamental Diagram of Road Traffic. 188. 105-107. 131-135 expected. 290-291. 151-161 exponential. 30 gravitational. 81 cubic. 124-126 Hyperbolic functions. 337 competing species.) logistic. proof by. 291 Gas dynamics. 49 Gardels. 30 restoring (see Restoring force) retardation. 165. 215. 10-11. 272 Flux. 124-126 instantaneous. 133 Haight. 15 Highways (see Traffic) Holland Tunnel. 286 Growth (see also Growth rate) constant: delay.. 229-230. 248. 215 pendulum with damping. 218 Following distances. 105. 30 Newton's law. 285. 248. 8. C. 165 delays. H. 49-50 conservative. 18 inverse-square law. 91-97 oscillations. 337 Isolated equilibrium. 126 Isoclines. 32 nonlinear oscillations. 164-165 discrete. 46 Gravity. K. 311. 348 Gravitational force. 165-169. 349. partial differential equations (see Partial differential equations) Initial value problems. 307-308. 24. 131 measured. 224. 94-99. 10-11 Greenberg. 182 discrete. 337 characteristics. 255. 115 Gradient. 7. 272 Gazis. 244-245 Free-falling body. 350 Jump-discontinuity. 351 Interest. 159. 24 Force-velocity relationship. differentiating. 164 discrete-delay. 251-252. 36 Friction coefficient. 394 Hare (see Lynx-hare ecosystem) Harmonic motion (see Simple harmonic motion) Heavy traffic.317 Herman. 226. 311-313. 46 linear damping. 122. 349 Gases. 97-98 trajectory. 352 (see also Fundamental Diagram of T _Road Traffic) Fluid dynamics. notation. 292 Force: central. 98 negligible. 4-5 potential energy. 190. 95.398 Index Flow-density relationship. 162-171 density-dependent. F. 29-42 Coulomb. 111. 131 (see also Growth rate) Integral conservation law. 126-128 Inverse-square law.. 251-252 nodes.. 183 rate (see Growth rate) zero. 210-211 spirals..

308. E. 226 Lotka-Volterra differential equations. 221 Leslie. M. 7-9. 25 frictional and restoring forces. 108-109. 135-136. 58. 153-161 (see also Logistic growth) explicit solution. 126. 157-158. 43-47 restoring and frictional forces.102. 151-161 discrete time. differential equations. 4-9. 309 (see also Capacity (road)) Orbits. 186 Neutrally stable. 189. 5 Mathematical ecology (see Populations) Matrices. 139-143. 179-180 "Jumping" rope. 251252 Nonhomogeneous differential equations. 23-24 Nodes. 233-235 (see also Stability) Linear systems. 349 Mutualism. 92-93 gravity. 222-223. 49 Linearized stability analysis.. 15 damping. 113 Newton's law. 303308 Linearization. 29-30. 349 Limit cycle. 23-24 (see also Forces) first law. 3 Momentum. 63-64 pendulum. 70 Landau. 91-97 Nonlinear pendulum (see Pendulum) Nonlinear systems. 202 (see also Pendulum. 298. 164-165 discrete. 319-320 Logistic equation. 140-141 Linear damping force. 245 Maltbus. 4-9 (see also Newton's law) Moving coordinate systems.. 5 Motion (see Newton's law. 212-215. mathematical. 165-169. 382 Odum. 155-158 Logistic growth. 242.399 Index Models. 165. J. 48 period. 137 Af-wave. 138-143 delays. 123. 335 Nonlinear oscillations: damping. 32. with. 159-161 phase plane. 226-246 Lynx-hare ecosystem. 237 Optimal traffic density. 170. 4-5 spring-mass system. 10-11. 232-233. M. 15-16 convergent.108-109. 193-199 eigenvalues. 290-291 Migration. M.. 216. 370-371 Linearized pendulum (see also Pendulum) circular frequency. Law of. 304-305. 16. 162.. 133 Merritt Parkway. 386 Left-handed spiral.. 218. 141. 9 two-mass oscillator. 168-169. R. P. 226. 301-303. 322. 139. 188-190. 247 (see also Logistic equation) delay. 100-103. 54-56. D. 225 One-species population models. Traffic) Motion. 183 Lotka. 311-312 Kepler. 272 Local conservation law. Spring-mass system) Number of cars. 168-169. 125 Mass. 286-288. phase plane. 40-42 cycles per second. 283 Lightning. 115 Light traffic. 254 . 92-93 second law. 75. 234 amplitude. 174-176. 50 Kinetic energy. 50 (see also Newton's law) Newtonian damping. 30 Linear oscillator.. 326 Linear algebra. 202 Liquids. 225. 143-151 Open curves.317 Lighthill. 115 Lane changing. 194-199 May. 91-97 friction. 59 Ordinary differential equations (see Differential equations) Oscillations. 256 Mean generation time (see Doubling time) Measured growth rate. 256 Maynard Smith. E. 130. 174-176 circular frequency. 164-165 probabilistic. 122-125 age distribution. 59 Order symbol. 3-10 third law. 162-171 density-dependent. 286-293. 14. 91-97 decaying. L. 179-180 critically damped. 183 discrete-delay. 278 (see also Conservation of cars) Local wave velocity. 274-275 (see also Conservation of cars) Numerical solutions. 77-79 spring-mass system. 245 Limited nutrients (see Logistic equation) Lincoln Tunnel. 200202. 309. 236 Newton. 156-158. 6 three dimensions. 70-75 Linear partial differential equations. J. 63 (see also Energy) pendulum. 241 Lifshitz.

155-158 nonlinear oscillations.. 77-79 stability. 206-223. 190 . 49. 242 period. 255 density. 15-16. 303-308 water waves. 188. 174-176. 156 zero. 234. 245 Phase of oscillation. 220 Polar representation. 314 phase plane. Symbiosis. 151 explosion. 104-113 energy. 188. 178 period. 14-15 revolutions per second. 51 Period (oscillation). 233 partial differential equations. 66. 15 limit cycle. 305. 91-97 overdamped. 153 Pendulum. 386 Pearl.) stability. 256 Planetary motion. 76-85 potential energy. 119-256 equilibrium. 48-49. 234-235 energy dissipation. 122 saturation level. Predator-prey. 130. C. 349 Prey (see Predator-prey) Oscillations (cont. 94-99 trajectories. 48. 51-52. 87-90 phase plane. 186. 76-85 separatrix. 175-176 pendulum. 40-42. 91-97 Hertz. 55 nonlinear. 186 Partial derivatives. 43-47 Potential energy. 178 Parabolic characteristics. 167. 70-75. 177 complex numbers. 59 Plankton (see Shark-fish ecosystem) Plant-herbivore systems. 232-233 natural frequency. 247-255 (see also Conservation of energy) ellipse. 314 Pesticide. 16 simple harmonic motion (see Simple harmonic motion) spring-mass system. 92-94 nonlinear systems. 15 nonlinear. 58. 229. Forest ecosystems. springs in. 242-243 phase. 346 Partial fractions. 242-245 carrying capacity (see Carrying capacity) coexistent. 58 damped.314 Partial differential equations. 278 chain rule. 189 dynamics. 77-79 linearized. 5 polar coordinates. 77-79 friction. 217-218. 302. 15-16 friction. 180 frequency. Fox-rabbit ecosystem. 15 Perturbation methods (see also Linearized stability analysis) linearization. 5 Poisson distribution. 12-16 underdamped. Logistic growth. 248249 expected. 256 Polar coordinates. 172-176 Populations: average. 188 spring-mass system. 70-75 linear systems. 43-47. 55 kinetic energy. 168-169 traffic density. R. 224-246. 105. 223. 68-69. 249. 50 Planetary orbits. Twospecies population models) percentage change. 272-273 Poole. Mutualism. 314-315 (see also Characteristics) characteristics. 250-255 Pielou. Three-species ecosystems. 14-15. 254 Pressure (air). 63. 100. 229-239. 186 Point mass. 332 Parallel. 280. 202 pendulum. 162 Overdamped oscillations. 302 linear. 123 rate of change. 202-223 logistic equation.400 Index Perturbation methods (cont. 229 Position. 18-20. 42-56. 234 natural position. 113-114 wild. 302. 87-90 spring-mass system.) divergent. 40-42. 161 growth (see Growth. W. 388-392 Parabolic Fundamental Diagram of Road Traffic. 70-71 Predator-prey. 152 displacement from equilibrium. 85 singular points. 159 Particular solution.. 67 (see also Energy) pendulum. 154. Growth rate) models (see Competing species. 77-79 spring-mass system. damping.92 linear oscillator. Plantherbivore systems. 27 Parsite-host pair. 14-15 Phase plane. 104-113 inverted position. 58-59 variable length. 298301. 105. 190. E. 105 VanderPol. 242-243 estimation. 157-158. 306-308 (see also Characteristics) equilibrium solution. Shark-fish ecosystem. 34-39. 336 Passing (traffic). 72-76 Periodic functions. Onespecies population models.

96-97 Retardation force. 38 Response. 224246 Sharks (see Shark-fish ecosystem) Shocks (traffic). 94-99 kinetic energy. 19. 355-357 weak. 70 Newton's law. 363-370 strength. 216-223. 100-103. 56-61. 377-379 velocity. 53 Second-order difference equations (see Difference equations. P. 7. 353 Simple harmonic motion. 7 Spring-mass system. 229 . 85 Series. 5-6 Road capacity (see Capacity (road)) Rope. F. 188190. 18 Spring force. 126-128 Scaling. 205-211. 349 Sound barrier. 223. 81 energy integral. J. 349 Speed (see also Velocity) sound. 221 right-handed. 200-202 (see also Stability) Stable age distribution. 219 stable. phase plane.. velocity) earliest.215 Stable spiral. 156-158 Stable. 9-11 Stability. 70 force (see Spring force) friction. 26 Reed.. 208-209 Reaction time. E. 201 equilibrium. 15-16. 218 Steady-state car-following models (see Car-following models) Stochastic models (see Probabilistic models) Stoker. 7. 81. 372 Right-handed spiral. 216. 217 Singular points. 246 Rarefactive wave. 27 Shallow water waves. 12-16. 237 exponential. 158 Smoots. 168-169. 394 Probability. 186 Principle of competitive exclusion. 154. 141 Stable equilibrium solutions (see Stability) Stable node. 9-11 isoclines. 12. 222-223. 6-42. 139 Symbiosis. 218. 48-49. 349. 19 stiffness. 153 Relative flow.. 185-186. 324 Ray. 291 Rabbits (see Fox-rabbit ecosystem) Random births. 234 Coulomb friction. 9-11 horizontal. 11 Spring constant. 143-151 (see also Probabilistic? models) Poisson distribution. 219 Rigid bodies. 23-28 vertical. I. 350-354 wave. 126 Reproductive rate. 102. 284 Spirals. 349 Sound waves. 70-75. springs in. 16 Richards. 272-273. 8 dimension. 100. 98-99. 213. 272-273 Propagation (wave). 349 Sound speed. 21 Restoring force. 232. 108-109. 156 Savings bank.251 (see also Linearized stability analysis) diagram. second-order) Second-order differential equations (see Differential equations. 218 Spring (see also Spring-mass system) firmness. 200202. 24 experimental. 58. 6-8 oscillations. 346 Shark-fish ecosystem. 218 unstable. 217 left-handed. 372-384. 25. 306 (see also Density waves. second-order) Separatrix. 388-389 condition (see Shocks. Waves) Queens-Midtown Tunnel. 108-109. 64-65 neutrally stable.401 Index Smith. 123 Resistive force. 236 pendulum. 30 Resonance. 17 Sonic boom. 58-59 populations. 246 traffic. 71 period. 143-151. 296-297 Rest. 102. 199-203. 58-59 perturbation methods. 143-151 Random processes. 236. 114 Survivorship rate. 349 Reproductive capacity. 32 Revolutions per second. 251 Probabilistic models: population. 94-99 potential energy. 4. 70-71 spring constant (see Spring constant) spring force (see Spring force) two-mass oscillator. 72-74 phase plane. 38 Response time. J. 283. 121.349 Speed limit. 331 Reduced mass. 21-22. 311-312 Saddle points. 251 Saturation level. 188. 293 braking. 112 gravity. 110.

13. 51 two variables. 324 shallow water. 328 "jumping" rope. 283. 89. 101. 344-345 breaking. 306 (see also Density waves) discontinuous. 317 jam. 302 error. 189. 381 light. 163 Tunnels. 209-211. 5 angular. 382 rarefactive. 227 Underdamped oscillations. 257-394 accidents. 32 Three-species ecosystems. 185-186 (see also Competing species. 362. 345 conservation of cars (see Conservation of cars) density (see Density (traffic) ) density waves (see Density waves) experiments. 311-312 N-wave. 154 Systems (see also Differential equations. 260 density waves. differential equations. 5 flow. 43-47 Velocity. 218 Van der Pol oscillator. 68. 323-338 stopped. 282-289. 294-296 linear. 260-263 Verhulst. 63 (see also Energy) Traffic. 349 Time delays (see Delays) Total energy. 206 (see also Stability) Unstable equilibrium solutions (see Stability) Unstable node. 286-289 flow (see Flow (traffic)) heavy.349 speed. 229 Zero population growth. 133 remainder. 331 velocity. 34-39. 354-359 velocity (see Velocity. 213. 113-114 Vectors. 291 (see also Lincoln Tunnel) Two-mass oscillator. 346 density. 252-254 Yield. 309. 153 Volterra. 187191 Taillights (see Brake lights) Taylor series. 349 (see also Shocks. 52 exponential. 354-360 yellow. 46 dimensions. waves) sound. 348 car. 372-382 car-following models (see Car-following models) characteristics (see Characteristics) congestion. 227. B. 127-128 Zero population. 171. 251 Traps. 331-339. 317 nearly uniform.402 Index Units.105 Undetermined coefficients. velocity) waves (see Density waves) Velocity-density relationship. 16. 55. G. 348-354 maximum. 82 . 353 Whitham. 263 Terminal velocity. 32 traffic. 16. 37.. systems) delay-differential equations. 336 Uniqueness theorem. 281 polar coordinates. 254 Thunder. Symbiosis) difference equations. 17 escape. 360 Trajectories. 326-327 Water waves. 302-318 shocks (see Shocks (traffic)) started. 319-320 (see also Density waves) water. 394 Witt. 244-245 Tree populations. 335 (see also Car paths) mass. 11 straight line. 114 Work. A. 57. 43-47 terminal. 323 Temperature. 306 stationary. 294 first-order differential equations. 323-339 red. 17 Unstable. 312-313. 346 Weak shocks. Predator-prey. 63 Yeast. A. 374 Velocity field. 206-223 (see also Phase plane) cars. 23-28 Two-species population models.. 346 Waves (see also Density waves) brake lights. 154. 200-202. 67 polar coordinates.215 Unstable spiral. 226 Waiting times at light. 127. local. 337 green. 283 shocks. traffic) Traffic light. 350-354 (see also Shocks. 309. 346 shocks. 306 (see also Density waves) expansion.