Coupled Pendulums

Apparatus:
Two pendulums, a helical spring, Meter scale, stop watch
Purpose of Experiment:
To study the different oscillating modes of coupled pendulums.
Basic Methodology:
Periods of oscillations for various initial conditions are observed. These are then used to verify
the relations between frequencies of these oscillations.

I Theory
Simple harmonic motion arises in a very general manner in many situations in Physics and
Engineering. The response of a system to small displacements from the equilibrium configuration
can usually be described in terms of simple harmonic oscillations of individual oscillators making
up the system. However, the oscillators will not have independent motion but are generally
coupled to other oscillators. Think, for example, of the passage of vibrations through a solid.
A solid can be thought of as being composed of a lattice of atoms connected to each other by
springs. The motion of each individual atom is coupled to that of its neighboring atoms.
A system of coupled oscillators exhibits some typical motion called the normal modes
of vibration. In general, the individual oscillators may have different natural frequencies of
vibration.In a normal mode, the motion of the system is such that all the individual
oscillators vibrate with the same frequency–called the normal mode frequency–and
with definite phase relations between the individual motions.
If a system has n degrees of freedom (i.e. has
n coupled oscillators), then there will be n normal
l
modes of the system. A general motion of the sys-
tem can be described as a superposition of normal
θA θB
mode vibrations. If a single oscillator is excited,
then eventually the energy gets transferred to all
the modes. L
In this experiment we will study some of the
above features in the simple case of two coupled
oscillators. The system studied in the experiment
consists of two identical rigid pendulums, A and B.
xA xB
A linear spring couples the oscillations of the two
pendulums. A schematic diagram of the system is Figure 1: Coupled Pendulums: Setup
given in Figure 1.
The equations of motion of the two physical pendulums are easily obtained. Let L be
the length of each pendulum (from point of suspension). Let θA and θB be the angular dis-
placements, and xA and xB the linear displacements of the two pendulums respectively. The
compression of the spring will be (xA − xB )l/L where l is the distance between the point of
suspension and the point where the spring is attached. If k is the spring constant and I the

1
2 PHYSICS LABORATORY MANUAL

moment of inertia, the rotational equation for pendulum A will thus be

d2 θA l
I = −mgLCM sin θA − k(xA − xB ) l cos θA , (1)
dt2 L
where the first term on the right is the restoring torque due to gravity (LCM being the distance
between the point of suspension and the position of the centre of mass of pendulum A) while
the second term is that due to the spring coupling. Assuming the mass attached to pendulum
A to be sufficiently heavy we can equate LCM and L. We also consider small displacements θA ,
so that sin θA ≈ θA and cos θA ≈ 1.
Substituting θA = xA /L and using the above approximations, we obtain the following equa-
tion of motion for the linear displacement xA :

d2 xA mgL l2
= − x A − k(xA − x B ) . (2)
dt2 I I
Likewise the equation for xB is

d2 xB mgL l2
= − x B + k(x A − x B ) . (3)
dt2 I I
Equations (2) and (3) are coupled, i.e. the equation for xA involves xB and vice-versa.
Without the coupling, i.e. in the absence of the spring, xA and xB would be independent
p
oscillations with the natural frequency ω02 = mgL/I.
It is easy to find uncoupled equations describing the normal modes of the system. Define
the variables
x1 = xA + xB ; x2 = xA − xB . (4)

Adding and subtracting eqs. (2) and (3), we obtain equations for the variables x1 and x2 as

d2 x1 mgL
= − x1 , (5)
dt2 I
d2 x2 mgL kl2
= − x2 − 2 x2 . (6)
dt2 I I
These equations for x1 and x2 are uncoupled. The variables x1 and x2 describe independent
oscillations and are the two normal modes of the system. The general solution to these equations
will be
x1 = A1 cos(ω1 t + φ1 ); x2 (t) = A2 cos(ω2 t + φ2 ), (7)

(A1 , A2 being the amplitudes of the two modes and φ1 , φ2 arbitrary phases). The corresponding
the natural frequencies are the normal mode frequencies:
r
2kl2 p
ω1 = ω0 ; ω2 = ω02 + = ω0 1 + 2Ω2 , (8)
I
2
where ω0 is the natural frequency of each uncoupled pendulum and Ω2 = klI .
Normal Mode Oscillations:
It is instructive to visualize the motion of the coupled system in these normal modes. In
the first normal mode, x1 (t) 6= 0, but x2 (t) = 0 at all times. Now the individual motions of
pendulums A and B are
1 1
xA (t) = (x1 (t) + x2 (t)) , xB (t) = (x1 (t) − x2 (t)) . (9)
2 2
 3

1st Normal Mode 2nd Normal Mode

Figure 2: The two Normal modes

In the first normal mode,

A1
xA (t) = cos(ω1 t + φ1 ) = xB (t). (10)
2
Both pendulums move in phase with the same amplitude and with frequency ω1 .
In the second normal mode, x1 (t) = 0 for all times and x2 (t) 6= 0 the individual motions
are
A1
xA (t) = cos(ω2 t + φ2 ) = −xB (t) (11)
2
In this mode the displacements of the pendulums are always opposite. Their motions have the
same amplitude and frequency (ω2 ) but with a relative phase difference of π. Figure 2 shows
the motions in the normal modes.
Beat Mode:
A general motion of the coupled pendulums will be a superposition of the motions of the
two normal modes:
1
xA (t) = [A1 cos(ω1 t + φ1 ) + A2 cos(ω2 t + φ2 )]
2
1
xB (t) = [A1 cos(ω1 t + φ1 ) − A2 cos(ω2 t + φ2 )] (12)
2

For a given initial condition the unknown constants (two amplitudes and two phases) can
be solved for. Consider the case where the pendulum A is given a displacement A at t = 0 and
released from rest while B remains at its equilibrium position at t = 0. The constants can be
solved for (see Exercise 4) to give the subsequent motions of the pendulums to be
   
ω2 − ω1 ω2 + ω1
xA (t) = A cos t cos t
2 2
   
ω2 − ω1 ω2 + ω1
xB (t) = A sin t sin t (13)
2 2

The motions of the pendulums A and B exhibit a typical beat phenomenon. The motion can
be understood as oscillations with an angular frequency ω3 = ω2 +ω
2
1
and a sinusoidally varying
amplitude A(t) = A cos(ω2 −ω1 )t/2, whose amplitude shows a sinusoidal oscillation with angular
frequency ωb = ω2 −ω
2 . This mode is called Beat mode and ωb is called Beat frequency. As
1

examples, Figure 3 shows plots of x(t) = sin(2πt) sin(50πt) and x(t) = cos(2πt)cos(50πt) vs. t.
4 PHYSICS LABORATORY MANUAL

Figure 3: Beat phenomenon. Left: x(t) = sin(2πt) sin(50πt), Right: x(t) = cos(2πt)cos(50πt)

II Set-up and Procedure:

1. Suspend the two pendulums from the same height. Attach the coupling spring at a distance
l from the pivot. Use the scale to measure the initial amplitude of each pendulum in each
of the following cases

(a) Normal mode 1: both have same amplitude and phase,

(b) Normal mode 2: they have same amplitude but opposite phase,
(c) Beat mode: one fixed and other given an initial amplitude.

(The initial conditions are explained in the theory part).

2. In each mode, note down the time period of oscillations of any ONE pendulum and
calculate the respective angular frequencies.

3. Repeat the experiment for different coupling length (say l = 20 cm, 35 cm,50 cm, 65 cm,
80 cm). Find the natural frequencies of the pendulums separately using date for each each
mode, as given below.

• In-phase mode (normal mode 1):

ω1 = ω0 , (14)

• Opposite phase mode (normal mode 2):

kl 2
ω22 = ω02 + 2Ω2 , where Ω = . (15)
I
So a graph between ω2 and l2 will be a straight line with y intercept equal to ω02 .
• Beat mode: for weak coupling (ω0 < Ω) we have

Ω2
ω3 ≈ ω0 + , (16)
2ω0
Ω2
ωb ≈ . (17)
2ω0
 5

III Exercises and Viva Questions

1. What are the normal mode oscillations of a system? How many normal modes will a
system possess?

2. Infer the normal mode frequencies for the coupled pendulum by directly considering the
motion in the two modes as shown in Figure 2.

3. Qualitatively explain why the first normal mode frequency is independent of the position
of the spring while the second normal mode frequency increases with l, the distance of the
spring from the point of support.

4. For the case where pendulum A is lifted and released from rest derive the unknown
constants A1 , A2 , φ1 and φ2 in equation (11) to obtain the solution equation (12).

5. Explain the effect of damping on the motion. Redraw Figure 3 qualitatively if damping
is present.

6. List all the approximations made in the theory of the coupled pendulum treated in the
theory as against the actual apparatus used and estimate the error introduced. Also,
consider possible sources of random errors while conducting the experiment.

7. In normal mode2 you might have seen an oscillating amplitude instead of a steady or
slowly decaying one. Why this is happening? How to avoid this behavior?

8. Explain what happens if the pendulums are coupled by a weighted string with tension τ ,
instead of the spring.

References:
1. Vibrations and Waves, A.P. French, Arnold-Heinemann, New Delhi, 1972.

2. The elements of Physics, I.S. Grant and W.R. Phillips, Oxford University Press, Oxford,
2001.