Non-linear dynamics

CHAOS
The equations which we usually come
across in Physics are linear:

1. Equation of Motion of a simple

pendulum

m mg
l
2. Wave Equation

y 1 y
2 2

x 2
v t
2 2
3. Poisson Equation:

4. Schrodinger equation

2 2

V ( x) i
2m x 2
t
Is nature linear or do we choose to
solve only problems which involve
linear equations?
 It is easy to solve linear equations.
In general, it is possible to find
analytical (closed form) solutions for
most of them.
They have some useful properties
they obey the principle of
superposition.

But, in real life ,non-linearity is

lurking just around the

sin corner.
m mg
l
 With non-linearity, superposition
goes out of the window.
What are the other consequences?

Sensitive dependence on initial

conditions.
Non-periodic orbits
Chaos

Phase portrait will tell us about all

the above.
Chaos has exorcised Laplaces demon:

We may regard the present state of the universe

as the effect of its past and the cause of its
future. An intellect which at a certain moment
would know all forces that set nature in motion,
and all positions of all items of which nature is
composed, if this intellect were also vast enough
to submit these data to analysis, it would
embrace in a single formula the movements of
the greatest bodies of the universe and those of
the tiniest atom; for such an intellect nothing
would be uncertain and the future just like the
past would be present before its eyes.
Pierre Simon Laplace,A Philosophical Essay on
 In the 1960s, it was found that the
weather is chaotic The butterfly
effect
By the 1990s it was found that the
even the solar system is chaotic!
The orbits of Earth and Pluto have
been shown to be chaotic, i.e,

Hopefully it will at least stay in its

orbit!
Whew! So we have enough
time!
Phase space portrait of SHM
The phase space trajectory of a
particle cannot intersect itself
(uniqueness of solution)
Therefore, trajectories either
spiral in (damped SHM)
Spiral out (forced undamped)
Are periodic (undamped or damped
forced SHM)
Damped SHM
If the phase space has more than two
dimensions, there are other
possibilities:

The orbit in phase space can be

confined within a given volume
(conservative system) without being
periodic.

Can be chaotic.
Historically, one of the first systems in
which chaos was noticed was a system
of coupled non-linear differential
equations used to model the weather.

dx dy
( y x) x y xz
dt dt

dz
z xy
dt
 Edward Lorenz used these equations
to model the weather in the 1960s
with the parameters:
=10, = 8/3, = 28

Procedure:
1. Use Eulers method to solve these
differential equations.
2. Obtain plots for various initial
conditions. See whether the nature
of the plot varies for different initial
conditions.
3. Check for sensitivity to initial
conditions. You can do this by looking
at the coordinates of the system after
a large number of iterations (say
10000). Note down the final
coordinates of the system for a few
different initial conditions. Check
whether the system is sensitive to
initial conditions (and if so for what
range of initial conditions).
4. Also repeat the entire procedure for
different sets of the parameters,
We will begin our study of chaos by
looking at a simple equation known as
a logistic map.
The logistic map came out of attempts
to model populations (of insects etc).

Pn+1 = Pn(a-bPn)

The first term models the growth while

the second term is related to the
decrease in the population due to
overcrowding / competition

Make a change of variables.

Pn = (a/b)xn

and r = a/4

Therefore, the equation takes the

form:

xn+1 = 4rxn(1-xn)

x and 4r are restricted to the interval 0

to 1 to avoid negative values of x.
 0.5
rr=0.25
=1
0.45
x(0)=0.5
0.4

0.35

0.3

0.25

0.2

0.15

0.1

0.05

0
0 10 20 30 40 50 60 70 80 90 100
1.5
4r=2
r =2
x(0)=0.5

0.5

-0.5
0 10 20 30 40 50 60 70 80 90 100
0.64
4 r =2.5
0.62 r x(0)=0.5

0.6

0.58

0.56

0.54

0.52

0.5
0 10 20 30 40 50 60 70 80 90 100
 0.7

0.68
4r r =2.75
x(0)=0.5
0.66

0.64

0.62

0.6

0.58

0.56

0.54

0.52

0.5
0 10 20 30 40 50 60 70 80 90 100
 0.75
4r r =3
x(0)=0.5

0.7

0.65
x

0.6

0.55

0.5
0 10 20 30 40 50 60 70 80 90 100
iteration number
 0.9

4 r =3.25
x(0)=0.5
0.85
r

0.8

0.75

0.7

0.65

0.6

0.55

0.5

0.45
0 10 20 30 40 50 60 70 80 90 100
 4 r =3.25
x(0)=0.5
5
r

5
 1
4 r =3.5
0.9 r x(0)=0.5

0.8

0.7

0.6

0.5

0.4

0 10 20 30 40 50 60 70 80 90 100
 4 r=3.75; xo = 0.5
r
1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2
0 10 20 30 40 50 60 70 80 90 100
 4 r=3.9; xo = 0.5
r
1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
0 20 40 60 80 100 120 140 160 180 200
 4 r=3.95; xo = 0.5

1
r
0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
0 20 40 60 80 100 120 140 160 180 200
 4 r=3.99; xo = 0.5
r
1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1

0
0 20 40 60 80 100 120 140 160 180 200
 Logisticmapfor twodifferent initial conditions
1

0.9

0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1
4 r =3.99

0
0 5 10
r
15 20 25 30 35 40 45 50

redline:x(0)=0.6
greenline:x(0)=0.5999
 Bifurcation diagram of the logistic map
1

0.9

0.8

0.7

0.6

0.5
xn

0.4

0.3

0.2

0.1

0
0.5 1 1.5 2 2.5 3 3.5 4
4r
Fixed points and their stability:

Two fixed points for the logistic

map:

xn+1 = 4rxn(1-xn) = xn

Therefore:

xn = 0 or xn = 1-1/4r

For r < , x = 0 is the only fixed point.

Stability?
To check the stability of the fixed
points we consider small
displacements from these points:

Therefore :

xn = x* + n

xn+1 = x* + n+1
xn 1 f ( x*) n f ( x*)

x * n f ( x*)
 n 1
f ( x*)
n
For stability the R.H.S should be less
than 1.

Analyze the stability of the two fixed

points in this case.
Consider another map:

xn+1 = 2xn mod 1

Iterate it a few times and see what

happens!
0.1 This goes on till you hit the
0.2 37th iteration, after which:
0.200001
0.400002
0.4 0.800003
0.8 0.600006
0.200012
0.6 0.400024
0.2 0.800049
0.600098
0.4 0.200195
0.8 0.400391
0.800781
0.601563
0.203125
0.40625
0.8125
0.625
How do we know if a system is
chaotic?

The most important characteristic of a

chaotic system is sensitivity to initial
conditions.
The divergence of trajectories can be
described by a Lyapunov exponent.

However direct application of this

equation is difficult as:
1. The rate of separation of the
trajectories might depend on the
choice of xo.
2. The separation ceases to increase
when n is sufficiently large (x has to
be within the unit interval)
Lorenz Model:
Describes convection in Rayleigh-
Benard cell

x = measure of the fluid flow

velocity circulating around the cell
y = temperature difference between
the rising and falling fluid regions
z = temperature difference between
top and bottom (relative to the
equilibrium temperature profile)