CHAPTER
Nonlinear Oscillations
and Chaos
4.1 Introduction
The discussion of oscillators in Chapter 3 was limited to linear systems. When
pressed to divulge greater detail, however, nature insists of being nonlinear; ex-
amples are the flapping of a flag in the wind, the dripping of a leaky water
faucet, and the oscillations of a double pendulum. The techniques learned thus
far for linear systems may not be useful for nonlinear systems, but a large num-
ber of techniques have been developed for nonlinear systems, some of which we
address in this chapter. We use numerical techniques to solve some of the non-
linear equations in this chapter.
The equation of motion for the damped and driven oscillator of Chapter 3
moving in only one dimension can be written as
mE + F(X) + g(x) = h(D (4.1)
If /(3) or g(x) contains powers of # or x, respectively, higher than linear, then the
physical system is nonlinear. Complete solutions are not always available for
Equation 4.1, and sometimes special treatment is needed to solve such equa-
tions. For example, we can learn much about a physical system by considering
the deviation of the forces from linearity and by examining phase diagrams.
Such a system is the simple plane pendulum, a system that is linear only when
small oscillations are assumed.
In the beginning of the nineteenth century, the famous French mathemati-
cian Pierre Simon de Laplace espoused the view that if we knew the position and
velocities of all the particles in the universe, then we would know the future for
all time. This is the deterministic view of nature. In recent years, researchers in
1444.1 INTRODUCTION 145
many disciplines have come to realize that knowing the laws of nature is not
enough. Much of nature seems to be chaotic. In this case, we refer to determin-
istic chaos, as opposed to randomness, to be the motion of a system whose time evo-
lution has a sensitive dependence on initial conditions. The deterministic develop-
ment refers to the way a system develops from one moment to the next, where
the present system depends on the one just past in a welldetermined way
through physical laws. We are not referring to a random process in which the
present system has no causal connection to the previous one (e.g., the flipping
of a coin)
Measurements made on the state of system at a given time may not allow us
to predict the future situation even moderately far ahead, despite the fact that the
governing equations are known exactly. Deterministic chaos is always associated
with a nonlinear system; nonlinearity is a necessary condition for chaos but not a
sufficient one. Chaos occurs when a system depends in a sensitive way on its pre-
vious state. Even a tiny effect, such as a butterfly flying nearby, may be enough to
vary the conditions such that the future is entirely different than what it might
have been, notjust a tiny bit different. The advent of computers has allowed chaos
to be studied because we now have the capability of performing calculations of
the time evolution of the properties of a system that includes these tiny variations
in the initial conditions. Chaotic systems can only be solved numerically, and
there are no simple, general ways to predict when a system will exhibit chaos.
Chaotic phenomena have been uncovered in practically all areas of science
and engineering—in irregular heartbeats; the motion of planets in our solar sys-
tem; water dripping from a tap; electrical circuits; weather patterns; epidemics;
changing populations of insects, birds, and animals; and the motion of electrons
in atoms. The list goes on and on. Henri Poincaré* is generally given credit for
first recognizing the existence of chaos during his investigation of celestial me-
chanics at the end of the nineteenth century. He came to the realization that the
motion of apparently simple systems, such as the planets in our solar system, can
be extremely complicated. Although various investigators also eventually came
to understand the existence of chaos, tremendous breakthroughs did not hap-
pen until the 1970s, when computers were readily available to calculate the long-
time histories required to document the behavior.
The study of chaos has become widespread, and we will only be able to look
at the rudimentary aspects of the phenomena. Specialized textbooks on the
subject have become abundant for those desiring further study. For example,
space does not permit us to discuss the fascinating area of fractals, the compli
cated patterns that arise from chaotic processes.
“Henri Poincaré (1854-1912) was a mathematician who could also be considered a physicist and
philosopher. His career spanned the era when classical mechanies was at its height, soon to be over-
taken by relativity and quantum mechanics. He searched for precise mathematical formulas that
‘would allow him to understand the dynamic stability of systems.
‘Particularly useful books are by Baker and Gollub (Ba96), Moon (Mo92), Hilborn (Hi00), and
ogatz (S94).146 4/ NONLINEAR OSCILLATIONS AND CHAOS
4.2 Nonlinear Oscillations
Consider a potential energy of the parabolic form
L
Ux) = gh (4.2)
Then the corresponding force is
F(x) = —kx (4.3)
This is just the case of simple harmonic motion discussed in Section 3.2. Now,
suppose a particle moves in a potential well, which is some arbitrary function of
distance (as in Figure 4-1). Then, in the vicinity of the minimum of the well, we
usually approximate the potential with a parabola. Therefore, if the energy of
the particle is only slightly greater than Upin, only small amplitudes are possible
and the motion is approximately simple harmonic. If the energy is appreciably
greater than Unig, 8o that the amplitude of the motion cannot be considered
small, hen it may no longer be sufficiently accurate to make the approximation
U(s) ~ dkx® and we must deal with a nonlinear force.
In many physical situations, the deviation of the force from linearity is sym-
‘metric about the equilibrium position (which we take to be at x = 0). In such
cases, the magnitude of the force exerted on a particle is the same at —x as at x;
the direction of the force is opposite in the two cases. Therefore, in a symmetric
situation, the first correction to a linear force must be a term proportional to x*;-
hence,
F(x) hx + ex! (4.4)
where e is usually a small quantity. The potential corresponding to such a force
is
1 1
5 kx? — Text 4.5)
9 ae (4.5)
U(x)
Ug)
FIGURE 41 Arbitrary potential U(x) indicating a parabolic region where simple
harmonic motion is applicable.