20.

Parametric Resonance
Michael Fowler

Introduction
(Following Landau para 27)

A one-dimensional simple harmonic oscillator, a mass on a spring,

d
( mx ) + kx =
0
dt

has two parameters, m and k . For some systems, the parameters can be changed externally (an
example being the length of a pendulum if at the top end the string goes over a pulley).

We are interested here in the systems response to some externally imposed periodic variation of its
parameters, and in particular well be looking at resonant response, meaning large response to a small
imposed variation.

Note first that imposed variation in the mass term is easily dealt with, by simply redefining the time
dt
variable to d = dt / m ( t ) , meaning = m ( t ). Then

d dx 1 d 1 dx 1 d 2 x
= m = m ,
dt dt m d m d m d 2

d 2x
and the equation of motion becomes + m ( ) kx =
0.
d 2

This means we can always transform the equation so all the parametric variation is in the spring
constant, so well just analyze the equation

d 2x
2
+ 2 (t ) x =
0.
dt

Furthermore, since were looking for resonance phenomena, we will only consider a small parametric
variation at a single frequency, that is, well take

2 ( t ) = 02 (1 + h cos t ) ,

where h 1 , and h is positive (a trivial requirementjust setting the time origin).

(Note: We prefer where Landau uses , which is often used for a resonance width these days.)
 2

We have now a driven oscillator:

d 2x
+ 02 x =
02 xh cos t.
dt 2

How does this differ from our previous analysis of a driven oscillator? In a very important way!

The amplitude x is a factor in the driving force.

For one thing, this means that if the oscillator is initially at rest, it stays that way, in contrast to an
ordinary externally driven oscillator. But if the amplitude increases, so does the driving force. This can
lead to an exponential increase in amplitude, unlike the linear increase we found earlier with an external
driver. (Of course, in a real system, friction and nonlinear potential terms will limit the growth.)

What frequencies will prove important in driving the oscillator to large amplitude? It responds best, of
course, to its natural frequency 0 . But if it is in fact already oscillating at that frequency, then the
driving force, including the factor of x , is proportional to

cos 0t cos=
t 1
2 cos ( 0 ) t + 12 cos ( + 0 ) t ,

with no component at the natural frequency 0 for a general .

The simplest way to get resonance is to take =20 . Can we understand this physically? Yes. Imagine
a mass oscillating backwards and forwards on a spring, and the spring force increases just after those
points where the mass is furthest away from equilibrium, so it gets an extra tug inwards twice a cycle.
This will feed in energy. (You can drive a swing this way.) In contrast, if you drive at the natural
frequency, giving little push inwards just after it begins to swing inwards from one side, then youll be
giving it a little push outwards just after it begins to swing back from the other side. Of course, if you
push only from one side, like swinging a swing, this worksbut it isnt a single frequency force, the next
harmonic is doing most of the work.

Resonance near Double the Natural Frequency

From the above argument, the place to look for resonance is close to =20 . Landau takes

x + 02 1 + h cos ( 20 + ) t x =
0

and, bearing in mind that were looking for oscillations close to the natural frequency, puts in

=x a ( t ) cos (0 + 12 ) t + b ( t ) sin (0 + 12 ) t ,

with a ( t ) , b ( t ) slowly varying.

 3

Its important to realize that this is an approximate approach. It neglects nonresonant frequencies
which must be present in small amounts, for example

cos (0 + 12 ) t cos ( 20 +=
)t 1
2 cos 3 (0 + 12 ) t + 12 cos (0 + 12 ) t

and the 3 (0 + 12 ) term is thrown away.

And, since the assumption is that a ( t ) , b ( t ) are slowly varying, their second derivatives are dropped
too, leaving just

x= 2a ( t ) 0 sin (0 + 12 ) t a ( t ) (02 + 0 ) cos (0 + 12 ) t
+2b ( t ) 0 cos (0 + 12 ) t b ( t ) (02 + 0 ) sin (0 + 12 ) t.

This must equal

02 1 + h cos ( 20 + ) t a ( t ) cos (0 + 12 ) t + b ( t ) sin (0 + 12 ) t .

Keeping only the resonant terms, we take cos (0 + 12 ) t cos ( 20 + =

)t 1
2 cos (0 + 12 ) t and
sin (0 + 12 ) t cos ( 20 + ) t =
12 sin (0 + 12 ) t , so this expression becomes

02 1 + h cos ( 20 + ) t a ( t ) cos (0 + 12 ) t + b ( t ) sin (0 + 12 ) t

02 a ( t ) cos (0 + 12 ) t + b ( t ) sin (0 + 12 ) t + 12 ha ( t ) cos (0 + 12 ) t 12 hb ( t ) sin (0 + 12 ) t
=

The equation becomes:

x= 2a ( t ) 0 sin (0 + 12 ) t a ( t ) (02 + 0 ) cos (0 + 12 ) t

+ 2b ( t ) 0 cos (0 + 12 ) t b ( t ) (02 + 0 ) sin (0 + 12 ) t
02 a ( t ) cos (0 + 12 ) t + b ( t ) sin (0 + 12 ) t + 12 ha ( t ) cos (0 + 12 ) t 12 hb ( t ) sin (0 + 12 ) t
=
=

The zeroth-order terms cancel between the two sides, leaving

2a ( t ) 0 sin (0 + 12 ) t a ( t ) 0 cos (0 + 12 ) t + 2b ( t ) 0 cos (0 + 12 ) t b ( t ) 0 sin (0 + 12 ) t
02 12 ha ( t ) cos (0 + 12 ) t 12 hb ( t ) sin (0 + 12 ) t
=

Collecting the terms in sin (0 + 12 ) t , cos (0 + 12 ) t :

( )
( 2a + b + 12 h0b ) 0 sin (0 + 12 ) t + 2b ( t ) a + 12 h0 a 0 cos (0 + 12 ) t =
0.
 4

The sine and cosine cant cancel each other, so the two coefficients must both be identically zero. This
gives two first order differential equations for the functions a ( t ) , b ( t ) , and we look for exponentially
to a ( t ) ae
increasing functions, proportional= = st
, b ( t ) be st , which will be solutions provided

sa + 12 ( + 12 h 0 ) b =
0,
1
2 ( 1
2 h 0 ) a sb =
0.

The amplitude growth rate is therefore

=s2 1 ( 1 h ) 2 2 .
4
2 0

Parametric resonance will take place if s is real, that is, if

12 h 0 < < 12 h 0 ,

a band of width h 0 about 2 0 .

Example: Pendulum Driven at near Double the Natural Frequency

=
A simple pendulum of length , mass m is attached to a point which oscillates vertically y a cos t .
Measuring y downwards, the pendulum position is

x sin =
= , y a cos t + cos .

The Lagrangian

=
L 1
2 m ( x 2 + y 2 ) + mg cos

(
m ( 2 cos 2 ) 2 + 12 m a sin t + sin )
2
= 1
2 + mg cos
d
= 1
2 m 22 ma sin t cos + a 2 2 sin 2 t + mg cos
dt

The purely time-dependent term will not affect the equations of motion, so we drop it, and since the
equations are not affected by adding a total derivative to the Lagrangian, we can integrate the second
d
term by parts (meaning were dropping a term ( ma sin t cos ) ) to get
dt

=
L 1
2 m 22 + ma 2 cos t cos + mg cos .

(Weve also dropped the term mga cos t from the potential energy termit has no or
dependence, so will not affect the equations of motion.)
 5

The equation for small oscillations is

+ 02 1 + ( 4a / ) cos ( 20 + ) t = 0, 02= g / .

Comparing this with

x + 02 1 + h cos ( 20 + ) t x =
0

we see that 4a / h , so the parametric resonance range around 20 = 2 g / is of width

1
2 h0 = 2a g / 3 .